US007820809B2

(12) UnitedO States Patent (10) Patent No.: US 7820,809 B2 KhVOrOVa et al. (45) Date of Patent: *Oct. 26, 2010

(54) FUNCTIONAL AND HYPERFUNCTIONAL 2004.0029275 A1 2/2004 Brown et al. SRNA DIRECTED AGAINST BCL-2 2004.00541.55 A1 3, 2004 Woolf 2004, OO63654 A1 4/2004 Davis et al. (75) Inventors: Anastasia Khvorova, Boulder, CO 2004/0101857 A1 5, 2004 Ward (US); Angela Reynolds, Littleton, CO 2004/O180357 A1 9, 2004 Reich et al. (US); Devin Leake, Denver, CO (US); 2004/O192629 A1 9, 2004 Xu et al. William Marshall, Boulder, CO (US); 2004/0204380 A1 10, 2004 Ackermann Stephen Scaringe, Lafayette, CO (US); 2004/0219671 A1 1 1/2004 McSwiggen Steven Read, Denver, CO (US) 2004/024.8296 A1 12/2004 Beresford s s 2004/0248299 A1 12/2004 Jayasena et al. (73) Assignee: Dharmacon, Inc., Lafayette, CO (US) 2004/0259247 A1 12, 2004 Tuschlet al. 2005/OO48529 A1 3/2005 McSwiggen

(*) Notice: Subject to any disclaimer, the term of this 2.99. A. 3. S.CSW1ggen past list: G adjusted under 35 2005/0176025 A1* 8/2005 McSwiggen et al...... 435/6 M YW- y U days. 2005/0181382 A1 8/2005 Zamore et al. 2005, 0186586 A1 8, 2005 Zamore et al. This patent is Subject to a terminal dis- 2005/0227935 A1 10/2005 McSwiggen Ca10. 2005/0239731 A1 10/2005 McSwiggen et al. 2005.0245475 A1 11/2005 Khvorova (21) Appl. No.: 11/083,784 2006/0286575 A1 12/2006 Farrell 2007/0031844 A1 2/2007 Khvorova et al. (22) Filed: Mar 18, 2005 2007,0254.850 A1 11/2007 Lieberman (65) Prior Publication Data FOREIGN PATENT DOCUMENTS US 2005/0245475A1 Nov. 3, 2005 WO WO980.0532 1, 1998 Related U.S. Application Data WO WOOO2O645 4/2000 (62) P.- - - of application No. 10/714,333, filed on Nov. WO WOOO76497WOOO21559 12/2000 s WO WOO244321 6, 2002 (60) Provisional application No. 60/502,050, filed on Sep. WO WOO3O35869 5, 2003 10, 2003. WO WOO3O35870 5, 2003 (51) Int. Cl. C7H 2L/02 (2006.01) (Continued) C7H 2L/04 (2006.01) OTHER PUBLICATIONS C12O 1/68 (2006.01) A61 K3 L/70 (2006.01) Holen et al., Nucleic Acids Research, Apr. 15, 2002, 30: 1757-1766.* Ho et al., Nucleic Acids Research, 1996, 24: 1901-1907.* (52) U.S. Cl...... soa5.s: Elbashir, et al. (2001) Functional Anatomy of siRNAS for Mediating • us Efficient RNAi in Drosophila melanogaster Embryo Lysate. The (58) Field of Classification Search ...... None EMBO Journal, v.20(23):6877-6888.* See application file for complete search history. Futami, et al. (2002) Induction of Apoptosis in Hela Cells With siRNA Expression Vector Targeted Against Bcl-2. Nucleic Acids (56) References Cited Resarch Supplement No. 2: 251-2.* Promega (2003) siRNA Target Designer-Version 1.1, Selentgene U6 U.S. PATENT DOCUMENTS Cassette RNA Interference System. httP://www.promega.com/ 6,001,992 A 12/1999 Ackermann siRNADesigner/program.* 6,111,086 A 8/2000 Scaringe 6,172.216 B1 1/2001 Bennett et al. (Continued) 2. R 3. Winstein Primary Examiner Richard Schnizer 6965.025 B2 11/2005 Gaarde et al. Assistant Examiner Jennifer Pitrak 6994,979 B2 2/2006 Reed et al. (74) Attorney, Agent, or Firm—Kalow & Springut, LLP: 7,022,831 B1 4/2006 Reed Scott D. Locke, Esq. 7,022,837 B2 4/2006 Harding 7,157,570 B2 1/2007 Yun (57) ABSTRACT 2002fOO81578 A1 6, 2002 Plowman et al. 3. h A. 2.2. RS,letal Efficient sequence specific gene silencing is possible through SC Ca. the use of siRNA technology. By selecting particular siRNAs 388. A. g ity etal by rationale design, one can maximize the generation of an 2003/0105051 A1 6, 2003 MSY effective gene silencing reagent, as well as methods for 2003/0143732 A1 7/2003 Fosnaugh silencing genes. 2003. O157030 A1 8, 2003 Davis et al. 2003/0228597 A1 12, 2003 COWSert 9 Claims, 33 Drawing Sheets US 7820,809 B2 Page 2

FOREIGN PATENT DOCUMENTS Tuschl, et al. (2001) The siRNA User Guide. Max Planck Institute for Biophysical Chemistry, pp. 1, 3 and 5. http://www.mpibpc.gwdg.de/ WO WOO3,064625 8, 2003 abteilungen/100/105/siRNAuserguide.pdf. WO WO 03/070969 8, 2003 Tuschl, Expanding small RNA interference (May 2002), Nature WO WOO3070897 8, 2003 Biotechnology, vol. 20, pp. 446-448. WO WOO3072704 9, 2003 Tuschl et al. (2003) The siRNA User Guide. 6 pages. WO WO2004.031237 4/2004 Vankayalapati, et al. (2003) Targeting Aurora2 Kinase in WO WO2004O46324 6, 2004 Oncogenesis: A Structural Bioinformatics Approach to Target Vali WO WO2004O485.11 6, 2004 dation and Rational Drug Design. Molecular Cancer Therapeutics, WO WO2004O901.05 10, 2004 W.2:283:94. WO WO2005 OO1043 1, 2005 Walton, et al. (1999) Prediction of Antisense Oligonucleotide Bind WO WO2005078095 8, 2005 ing Affinity to a Structured RNA Target. Biotechnol. Bioengineering, WO WO2005089.224 9, 2005 v.65(1): 1-9. (Abstract Only). WO WO2O05117991 12/2005 Wang, et al., (2006) Dexamethasone Represses Signaling through the WO WO, 2006O15389 2, 2006 Mammalian Target of Rapamycin in Muscle Cells by Enhancing WO WO2O06110813 10, 2006 Expression of REDD1. The Journal of Biological Chemistry vol. 281, No. 51, pp. 39128-39134. OTHER PUBLICATIONS Yao et al. (2000) Nature Cell Biology 2:484-491. Yuan, siRNA Selection Server: an automated siRNA oligonucleotide Amarzguioui et al., Secondary structure prediction and in vitro acces prediction server (2004). sibility of Mrna as tools in the selection of target sites for ribozymes. Zender et al., siRNA based strategies for inhibition of apoptotic 2000 Oxford University Press, Nucleic Acids Research, 2000, vol. pathways in vivo-analytical and therapeutic implications, Jan. 28, No. 2141 13-4124. 2004, vol. 9, pp. 51-54. Kasif et al. A computational framework for optimal masking in the Zhang, Physical and Functional Interationic between Myeloid Cell synthesis of oligonucleotide microarrays. 2002 Oxford University Leukemia 1 Protein (MCL1) ad Fortilin, J. Biol. Chem. 37430-37438 Press, Nucleic Acids Research, 2002, vol. 30 No. 20 e106. (2002). Vickers, et al., Efficient Reduction of Target RNAs by Small Inter Abeliovich, et al. (2000) Mice Lacking Alph-Synuclein Display fering RNA and RNase H-dependent Antisense Agents, Feb. 2003, Functional Deficits in the Nigrostriatal Dopamine System. Neuron The Journal of Biological Chemistry, vol. 278, pp. 7108-7118. v.25:239-52. Zhang et al., Physical and Functional Interaction between Myeloid Amarzguioui Metal, Tolerance for mutations and chemical modifi Cell Leukemia 1 Protein (MCL1) and Fortilin, 2002, The Journal of cations in a siRNA Nucleic Acids Research vol. 31, No. 2, Jan. 15, Biological Chemistry, vol. 277, pp. 37340-37438. 2003 pp. 589-595, XP002270887 ISSN: 0305-1048. “siRNA Target Finder” results for the entry of the nucleotide Bass, B.L., (2001) The Short Answer, Nature V.411:428-29. sequence of GeneBank Accession No. NM 021960. http://www. Brown et al. (1997) Identification and cDNA Cloning of a Novel ambion.com/techlib misc? siRNA finder.html, accessed on Mar. 6, Mammalian C2 Domain-Containing Phosphoinositide 3-kinase, 2008 in U.S. Appl. No. 1 1/633,404. HsC2-PI3K. Biochemical and Biophysical Research Communica Reynolds, et al. Rational siRNA design for RNA Interference, Nature tions. 233, 537-544. Biotechnology, Mar. 2004, vol. 22, No. 3, pp. 326-330. Brummelkarna et al. A System for Stable Expression of Short Ui-Tei, et al., Guidelines for the selection of highly effective siRNA Intereferring RNAs in Mammalian Cells, Science vol. 296 pp. 550 sequences for mammalian and chick RNA interference, Nucleic 553 (2002). Acids Research, Feb. 2004, vol. 32, No. 3, pp. 936-948. Cahill et al. (1999) Genomics. 58:181-187. Tusterman, et al., Dicers at RISC: The Mechanism of RNAi, Cell, Caplen et al. Specific inhibition of gene expression by Small double Apr. 2004, vol. 117, pp. 1-4. stranded RNAs in invertebrate systems, Proc. Natl. Acad. Sci USA Brusca, John S. PCT International Search Report, USPTO, Sep. 24. vol. 98, pp. 9742-9747 (2001). 2007. Caplen, dsRNA-mediated gene silencing in cultured Drosophila Epps-Ford, Janet L., PCT International Search Report, USPTO, Feb. cells: a tissue culture model for analysis of RNA interference. (2000). 25, 2005. Caplen, RNAi as a gene therapy approach (2003). Elbashir, et al., Analysis of Gene Function in Somatic mammalian cells using small interfering RNAS, 2000, vol. 26, pp. 199-213. Chalk, siRNA db: a database of siRNA sequences (2004). Khvorova, et al., Functional siRNAs, and miRNAs Exhibit Strand Chen et al. (2005) TSSK5, A Novel Member of the Testis-Specific Bias, Cell, Oct. 2003, vol. 115, pp. 209-216. Serine/Threonin Kinase Family, Phosphorylates CREB at Ser-122. Shietal. (2001) Gremlin negatively Modulates BMP-4 Induction of and Stimulates the CRE-CREB Responsive Pathway. Biochemical Embryonic Mouse Lung Branching Morphogenesis. Am J Physiol and Biophysical Research Communications. 333: 742-749. Lung Cell Mol Physiol, 280, pp. L1030-L1039. Cherry, Michael J. (1995) Computer Manipulation of DNA and Pro siDesign Center for “gene name: Src.” http://www.dharmacon.com/ tein Sequences. Current Protocols in Molecular Biology, 7.7.1-7.7. DesignCenter/DesignCenterPage.aspx, accessed on May 13, 2008. 23. siRNA Design for RNA Interference (RNAi) Experiments, http:// Chi, Genomewide view of gene silencing by Small interfering RNAs web.archive.org/web/20010101000000-20021231235.959/http:// (2003). www.ambion.com/techlib/misc/siRNA design.html. Accessed on Chinault, Overlap Hybridization Screening: Isolation and Character Mar. 6, 2008. ization of Overlapping DNA Framgments Surroungin the leu2 Gene Sorensen, et al. (2003) Gene Silencing by Systemic Delivery of on Yeast Chromosome III (1979), 5(2):111-26. Synthetic siRNAs in Adult Mice, J. Mol. Biol. 327, 761-766. Daniel et al., Specific association of Type I with Ran GTPase in Tan et al. (2006) Functional Cooperation Between FACT and MCM lipopolysacchardie-mediated differentiation, Oncogene 2001, vol. Helicase Facilitates Initiation of Chromatin DNA Replication. The 20:2618-2625. EMBO Journal, vol. 25, pp. 3975-3985. Dodelet et al. (2000) Eph Receptors and Ephrin Ligands: Truss, HuSide—the Human siRNA database: an open access data embryogenesis to tumorigenesis. Oncogene 19:5614-5619. base for published functional siRNA sequences and technical details Domin et al.(1997) Cloning of Human Phosphoinositide 3-Kinase of efficient transfer into recipient cells, Nucleic Acids Research vol. with a C2 Domain that Displays Reduced Sensitivity to the Inhibitor 33, pp. D108-D111 (2005). Wortmannin. Biochem. J., v.325: 139-47. Tsuji, et al. (2006) Essential Role of Phosphoorylation of MCM2 by Dottori, et al. (1998) EphA4 (Sek1) receptor tyrosine kinase is Cdc?/DbfA in the Initiation of DNA Replication in Mammalian Cells. required for the development of the corticospinal tract, Proc. Natl. Molecular Biology of the Cell, vol. 17, pp. 4459-4472. Acad. Sci., vol. 95, pp. 13248-13253, Neurobiology. US 7820,809 B2 Page 3

Ekholm, et al.(2004) Deregulation of Cyclin E in Human Cells Inter Murphy, et al. (2000) Synucleins are Developmentally Expressed, feres with Prereplication Complex Assembly. The Journal of Cell and Alpha-Synuclein Regulates the Size of the Presynaptic Vesicular Biology, vol. 165, pp. 789-800. Pool in Primary Hippocampal Neurons. The Journal of Elbashir et al. (2001) Duplexes of 21-Nucleotide RNAS Mediate Neuroscience, vol. 2009):3214-20. RNA Interference in Cultured Mammalian Cells. Nature vol. 411, pp. Naito, silDirect: highly effective, target specific siRNA design soft 494-498. ware for mammalian RNA interference, Nucleic Acids Research vol. El Touny, et al. (2006) Identification of Both Myt-I and Wee-I as 32, W124-129 (2004). Necessary Mediators of the P21-Independent Inactivation of the Cdc NCBI Sequence Viewer v2.0 for NM-004438, http://www.ncbi.nlm. 2/CyclinBI complex and Growth Inhibition of TRAMP Cancer Cells nih.gov/entrez/viewer.fcgi?db-nuccore&id=45439363, accessed on by Genistein. The Prostate, vol. 66:1542-1555. Jul. 3, 2008. European Patent Office, Office Action dated Nov. 15, 2007. Olie et al. A Novel Antisense Oligonucleotide Targeting Survivin Far, R-K, The activity of siRNA in mammalian cells is related to Expression Induces Apoptosis and Sensitizes Lung Cancer Cells to structural target accessibility: a comparison with antisense Chemotherapy, Cancer Research 2000, vol. 60: pp. 2805-2809. oligonucleotides (2003). Oishi, et al., Identification and Characterization of PKN Beta, a Feng, et al. (2003) Inhibiting the Expression of DNA Replication Novel Isoform of Protein Kinase PKN: Expression and Arachiodonic Initiation Proteins Induces Apoptosis in Human Cancer Cells. Cancer Acid Dependency Are Different from those of PKN alpha, 1999 Research, vol. 63, pp. 7356-7364. Biochemical and Biophysical Research Communication, 261, pp. Fox, et al. (1995) clDNA Cloning and Tissue Distribution of Five 808-814. Human EPH-like Receptor Protein-tyrosine kinases, Oncogene; Pan, et al. (2005) Calmodulin-dependent protein kinase IV regulates 10(5): 897-905. nuclear export of Cabinl during T-cell activation. The EMBO Jour Hammond et al., Post-transcriptional gene silencing by double nal, 24, 2104-21 13. stranded RNA, Nature Reviews, 2001, vol. 2, 110-119) MacMilan Promega siRNA Target Designer-Version 1.51, http://www. Magazines Ltd. promega.com/siRNADesigner/program/default.asp. Accessed Jun. Hao, et al. (2004) Expression Analysis of the Human Testis-specific 24, 2008. Serine/Threonine Kinase (TSSK) Homologues. ATSSK Member is Rabert et al. (1998) A Tetrodextrin-Resistant Voltage-gated Sodium Present in the Equatorial Segment of Human Sperm. Molecular Channel from Human Dorsal Root Ganglia, hPN3/SCN10A. Pain, Human Reproduction, vol. 10, No. 6 pp. 433-444. 78, pp. 107-114. Harborth et al., (2001) J. Cell. Sci. 114:4557-4565). Ross, et al. (2001) Inhibition of Kirsten-ras Expression in Human Human Neuropeptide Y Blast results, http://www.ncbi.nlm.nih.gov/ Colorectal Cancer Using Rationally Selected Kirsten-ras Antisen blast, accessed on Mar. 18, 2008. Oligonucleotides, Molecular Cancer Therapeutics vol. 1, 29-41. Iliakis et al. (1990) Induction and Repair of DNA Strand Breaks in Semizarov, Specificity of short interfereing RNA determined through Radiation-resistant Cells Obtained by Transformation of Primary Rat gene expression signatures, Proceedings of the National Academy of Embryo Cells with the Oncogenes H-ras and v-myc, Cancer Sciences USA vol. 100, pp. 6347-6352 (2003). Research, 50, pp. 6575-6579. Betrand et al., Comparison of Antisense Oligonucleotides and Jackson, et al., Expression profiling reveals off-target gene regulation siRNAs in cell culture and in vivo, 2002 Biotechnical and Biophysi by RNAi, Nature Biotechnology, vol. 21, pp. 635-637 (2003). cal Research Communication, 296 1000-1004. Kretscmer-Kazemei. The Activity of siRNA in mammalian cells is Tuschl et al., Expanding RNA Interference, (May 2002) Nature related to structural target accessibility: a comparison with antisense Biotechnology, vol. 20, pp. 446-448. oligonucleoitdes (2003). Futami, et al. (2002) Induction of Apoptosis in Hela Cells with siRNA Kumar, High-Throughput Selection of Effective RNAi Probes for Expression Vector Targeted Against BCL-2. Nucleic Acids Research Gene Silencing (2003). Supplement No. 2:251-252. Laitala et al., Inhibition of Bone Resorption in Vitro by Antisense Promega (2003) siRNA Target Designer-Version 1.1, SilentGene U6 RNA and DNAMolecules Targeted against Carbonic Anhydrase II or Cassette RNA Interference System. http://www.promega.com/ Two Subuntis of Vacuolar H + ATPase, Journal of Clinical Investiga siRNADesigner/program. tion 1994, vol. 93, pp. 2311-2318. Marathietal. RAD1, A Human Structural Homologofthe Schizosac Lapidot-Lifson, et al. (1992) Cloning and Antisense charomyces pombe RAD1 Cell Cycle Checkpoint Gene, 1998, Oligodeoxynucleotide Inhibition of a Human Homolog of cdc2 Genomics, vol. 54, pp. 344-347. required in Hematopoiesis, Proc. Natl. Acad. Sci., vol. 89, pp. 579 Kalra, et al., Central Administration of Antisense 583. oligodeoxynucleotides to neuropeptide Y (NPY) mRNA reveals the Levenkova, Gene specific siRNA selector, Bioinofrmatics vol. 20. critical role of newly synthesized NPY in regulation of LHRH pp. 430-432 (2004). release, 1995, Regulatory Peptides, vol. 59, pp. 215-220. Lindgren, et al. (2002) Contribution of Known and Unknown Sus Boutla et al., Short 5'-phosporylated double-stranded RNAs induce ceptibility Genes to Early-Onset Diabetes in Scandinavia, Diabetes RNA interference in Drosophila, 2001, Current Biology, vol. 11, pp. vol. 51, 1609-1617. 1776-1780. Lu, et al., The Human AQP4 gene: Definition of the locus encoding NCBI Sequence Viewer v2.0 for NM-000905, http://www.ncbi.nlm. two water channel polypeptides in brain, Proc. Natl Acad. Sci vol.93, nih.gov/entrez/viewer, accessed on Mar. 18, 2008. pp. 10908-10912 (Oct. 1996). SiRNA Converter, http://web.archive.org/web/2002010 1 Miller, V M et al., “Allele-specific silencing of dominant disease 20021231re ?http://www.ambion.com/techlib/misc/ genes' Proceedings of the National Acadamy of Sciences of USA, siRNA finder.html. vol. 100, No. 12, Jun. 10, 2003, pp. 7195-7200, XP002276730. siRNA Design for RNA Interference (RNAi) Experiments, http:// Miyagashi et al. (2003) Comparison of the Suppressive Effects of web.archive.org/web/20010101000000-20021231235.959/http:// Antisense Oligonucleotides and siRNAs Directed Against the Same www.ambion.com/techlib/misc? siRNA design.html. Targets in Mammalian Cells. Antisense and Nucleic Acid Drug Development 13:1-7. * cited by examiner U.S. Patent Oct. 26, 2010 Sheet 1 of 33 US 7820,809 B2

Figure 1

C. 3'5'-p IITp-5 3' --> siRNA or miRNA intermediate Reversible dsRNA-RISC complex formation

5 stable end 5' unstable end Non-processive unwinding 3. Processive unwinding

Irreversible ssRNA-RISC complex formation No RISC activation U.S. Patent Oct. 26, 2010 Sheet 2 of 33 US 7820,809 B2

r MM S

so i

SO &O. SO. 3O. 3O. SO. SCd w we was (Onuoo %) uoissendix buef afuel

U.S. Patent Oct. 26, 2010 Sheet 4 of 33 US 7820,809 B2

(%) ?auerbarasseroteucipund i U.S. Patent Oct. 26, 2010 Sheet 5 Of 33 US 7820,809 B2

IOIUOOJO BuOueIIS 96 U.S. Patent Oct. 26, 2010 Sheet 6 of 33 US 7820,809 B2

d d {d d N o o d rt ve (otiuoo %) uoss3.dx

s 2 (9 s e i t

Co O Cd d cN co c s we v ve (ouoo 9%) uossadx

U.S. Patent Oct. 26, 2010 Sheet 8 of 33 US 7820,809 B2

6|9||Z08/7€.k uopisodSw,9 U.S. Patent Oct. 26, 2010 Sheet 9 Of 33 US 7820,809 B2

1.6

1.4

1.2

1

0.8

0.6

0.2 O all SiRNA 100 nM

FIGURE 7

U.S. Patent Oct. 26, 2010 Sheet 11 of 33 US 7820,809 B2

NS -08 13%,

il F 1gure 8b

22222222222222222222 S ZZZZZZZZZZZZZ

3

~~~~§§~~~~~~ ...... …...…...….....SN@§§ZZZZZZZZZZZZZ??Ž? SENSE STRAND MODIFEED

IGF1R 3 RS 22222222222222

غg???????????????????

EGFROM WFILER1ABIO-PUBLICJSCHELTERNRNANDHARMACON2003 JULYWGF1R CHEM MOD1 MAT CLUSTERNG PARAMETERS, ALGORTHM = AHC TYPE - GENE SIMILARITY MEASURE = CORRELATION-BASED, METR CTYPE = , LINKA > O2 AND?ººººººººººº%***************~? P-VALUE< 0.01 IN 1 ORMORE EXPERIMENTS AND WITH 1 ORMORE GENE S N6EXPERIMENTS SELECTED FROM23661 GENESIN6EXPERMENTS

U.S. Patent Oct. 26, 2010 Sheet 12 of 33 US 7820,809 B2

3

i

s

. O O ve fuouaS 9, U.S. Patent Oct. 26, 2010 Sheet 13 of 33 US 7820,809 B2

8 3 S S o ruota) dix to 4 dx U.S. Patent Oct. 26, 2010 Sheet 14 of 33 US 7820,809 B2

00I•

·*,

|--~~~~|

---+--~~~~!)+---+--i-taa --~~~~~~~~~~~~~!) U.S. Patent Oct. 26, 2010 Sheet 15 Of 33 US 7820,809 B2

OUOO OO y 8, E 2. Onuoo OO 5 - Z (D

Onuoo OO

2. K OJUOO |OO y E 9 2. - is w Onuoo s OO O 2 s 9.a 9, D Oro) 5 S OO C D O y Z E. Sp () () t D OLOO S

S; god O C is S OUOO 9. Ood o - É 1. Z - (s On OO OO gD E E 2 3 Z ö OquoC) S y SS c is e ( )

CN O OO CO. v. CN w ve (uuo 96) dx. U.S. Patent Oct. 26, 2010 Sheet 16 of 33 US 7820,809 B2

ynowyOQVOnDVnVOVOOÐÍVNCHIS pyynynnwOOnVOVnÐVVOZVNCHIS yQyovyop?OOVVOVOOV00£VNCHIS nyoynowyOnyÐnÐVnVOVVVNCHIS nnnovOnOOQOnOVOVVOQSVNCHIS yOnnnonOnOOOOOOQVOÐ9VNRIKS nnny.pnn?nÐQOQOOOOOÐ0LVNCHIS VOynowyOQVOnDVnVOVO8VNCHIS OynowyOnyonÐVQVOVOO6VNCHIS pnmnovOnOÐQOQOVOVVO0IVNCHIS zlogdo)JosaouonbòS£I?Ináà U.S. Patent Oct. 26, 2010 Sheet 17 Of 33 US 7820,809 B2

MO R

O

qesyNYJ?speuô?sepK?euo?euO?Áquwop»3ouxiZ-108 uo?e?u00uOOINu00||

0Z). 00|| 08 09 07 0Z 0 OuOO % U.S. Patent Oct. 26, 2010 Sheet 18 of 33 US 7820,809 B2

() T s E 2 -F n r 5 mm 5 = C SM s - s se -

O O O O O O O V N O CO O s GN V V y uOSseudx % U.S. Patent Oct. 26, 2010 Sheet 19 Of 33 US 7820,809 B2

Hs vu O 2 P O . f s

Cd Co Co C Co Co O CN S2 do CO s en uosseudx % U.S. Patent Oct. 26, 2010 Sheet 20 of 33 US 7820,809 B2

O O O O O O O cN O CO CO v cN re- ra uoisseudx % U.S. Patent Oct. 26, 2010 Sheet 21 of 33 US 7820,809 B2

O C ed O O O O ca Cd od O s cN uosseudx % U.S. Patent Oct. 26, 2010 Sheet 22 of 33 US 7820,809 B2

xxas& e

88s

3.

P

8

&

r &

E. aNX:

38sixSSX

&

8.

E.

8.33 8x8

sts

2.

& 8 8 & 9 & O uosseudx % U.S. Patent Oct. 26, 2010 Sheet 23 of 33 US 7820,809 B2

s 8 8 & 9 & c uOSseudx % U.S. Patent Oct. 26, 2010 Sheet 24 of 33 US 7820,809 B2

O O O O s 8 CO O s S uossedX % U.S. Patent Oct. 26, 2010 Sheet 25 Of 33 US 7820,809 B2

& x 8: s

s e E3 x: s

O O O O O O CN O OO CO V CN O ve V uoissedX % Oct. 26, 2010 Sheet 26 of 33

uOSSoIdx % U.S. Patent Oct. 26, 2010 Sheet 27 Of 33 US 7820,809 B2

CD (VD s s O) 9. O GS & s 8

r CS CS Cs c Olu OOJO % U.S. Patent Oct. 26, 2010 Sheet 28 of 33 US 7820,809 B2

Oulu OO OUSel 9 CIT WUOO G CIT WUOO 7 CI WUOO . 9 Cl WUOO a CI WUOO . C. WUOO . A87 WUOO 008 WUOO 082 WUOO A. WUOO . 826 WUOO 28 WU00 99/ WUOO 90. WUOO

O), UOO JO % U.S. Patent Oct. 26, 2010 Sheet 29 of 33 US 7820,809 B2

TONOO

L3 U CD O - s Z U Y1

cN v

TONOOO % U.S. Patent Oct. 26, 2010 Sheet 30 of 33 US 7820,809 B2

?InáHZZ

(30U (0%) ge Jogezletiou U.S. Patent US 7820,809 B2

- P4

f

G - :WNI??S U.S. Patent Oct. 26, 2010 Sheet 32 of 33 US 7820,809 B2

Multitargeting of siRNA S 1.2 t S 1 a 0.8

28 o0.4 tSFSApool of S1 0.2 Sp O O - 5(t 2. 3P 8 (t FIG. 24 U.S. Patent Oct. 26, 2010 Sheet 33 Of 33 US 7820,809 B2

OO

qesyNYJ?speuô?sepK?euo?euOL.KquMopX[0oux,Z-108 uo?eu?ueouooWd002

OuOO % US 7,820,809 B2 1. 2 FUNCTIONAL AND HYPERFUNCTIONAL et al., The dsRNA Binding Protein RDE-4 Interacts with RDE SRNA DIRECTED AGAINST BCL-2 1, DCR-1 and a DexII-box Helicase to Direct RNAi in C. elegans, Cell 2002, June 28; 109(7):861-71; Ketting et al., CROSS REFERENCE TO RELATED Dicer Functions in RNA Interference and in Synthesis of APPLICATIONS Small RNA Involved in Developmental Timing in C. elegans; Martinez et al., Single-Stranded Antisense siRNAs Guide Tar This application is a divisional application of U.S. patent get RNA Cleavage in RNAi, Cell 2002, September. 6: 110(5): application Ser. No. 10/714,333, filed Nov. 14, 2003 entitled 563; Hutvagner & Zamore. A microRNA in a multiple-turn “Functional and Hyperfunctional siRNA.” which claims the over RNAi enzyme complex, Science 2002, 297:2056. benefit of U.S. Provisional Application Ser. No. 60/502,050, 10 From a mechanistic perspective, introduction of long filed Sep. 10, 2003, entitled “Methods for Selecting siRNA.” double stranded RNA into plants and invertebrate cells is the entire disclosures of which are hereby incorporated by broken down into siRNA by a Type III endonuclease known reference into the present disclosure. as Dicer. Sharp, RNA interference 2001, Genes Dev. 2001, 15:485. Dicer, a ribonuclease-II-like enzyme, processes the REFERENCE TO TABLES SUBMITTED IN 15 dsRNA into 19-23 base pair short interfering RNAs with ELECTRONICFORM characteristic two base 3' overhangs. Bernstein, Caudy, Ham mond, & Hannon, Role for a bidentate ribonuclease in the Applicants hereby incorporate by reference the material initiation step of RNA interference, Nature 2001, 409:363. Submitted previously, in duplicate on the compact disks The siRNAs are then incorporated into an RNA-induced labeled PATENT APPLICATION NO. 60/426,137, DISK1 silencing complex (RISC) where one or more helicases of 1, COPY 1 of 2: PATENT APPLICATION NO. 60/426, unwind the siRNA duplex, enabling the complementary anti 137, DISK1 of 1, COPY 2 of 2; which copies are identical, in sense Strand to guide target recognition. Nykanen, Haley, & files entitled Table 12.txt, date of creation Jun. 26, 2003, Zamore, ATP requirements and small interfering RNA struc with a size of 31,045 kb; Table 13.txt, date of creation Nov. ture in the RNA interference pathway, Cell 2001, 107:309. 13, 2003, with a size of 78,451 kb: Table 14.txt, date of 25 Upon binding to the appropriate target mRNA, one or more creation Nov. 13, 2003, with a size of 454 kb; and Table endonucleases within the RISC cleaves the target to induce 15.txt date of creation Nov. 13, 2003, with a size of 1,690 kb. silencing. Elbashir, Lendeckel, & Tuschl, RNA interference is mediated by 21- and 22-nucleotide RNAs, Genes Dev 2001, FIELD OF INVENTION 15:188, FIG.1. 30 The interference effect can be long lasting and may be The present invention relates to RNA interference detectable after many cell divisions. Moreover, RNAi exhib (“RNAi). its sequence specificity. Kisielow, M. et al. (2002) Isoform specific knockdown and expression of adaptor protein ShcA BACKGROUND OF THE INVENTION using small interfering RNA. J. of Biochemistry 363: 1-5. 35 Thus, the RNAi machinery can specifically knock down one Relatively recently, researchers observed that double type of transcript, while not affecting closely related mRNA. stranded RNA (“dsRNA) could be used to inhibit protein These properties make siRNA a potentially valuable tool for expression. This ability to silence a gene has broad potential inhibiting gene expression and studying gene function and for treating human diseases, and many researchers and com drug target validation. Moreover, siRNAs are potentially use mercial entities are currently investing considerable 40 ful as therapeutic agents against: (I) diseases that are caused resources in developing therapies based on this technology. by over-expression or misexpression of genes; and (2) dis Double Stranded RNA induced gene silencing can occur on eases brought about by expression of genes that contain muta at least three different levels: (i) transcription inactivation, tions. which refers to RNA guided DNA or histone methylation; (ii) Successful siRNA-dependent gene silencing depends on a siRNA induced mRNA degradation; and (iii) mRNA induced 45 number of factors. One of the most contentious issues in transcriptional attenuation. RNAi is the question of the necessity of siRNA design, i.e., It is generally considered that the major mechanism of considering the sequence of the siRNA used. Early work in C. RNA induced silencing (RNA interference, or RNAi) in elegans and plants circumvented the issue of design by intro mammalian cells is mRNA degradation. Initial attempts to ducing long dsRNA (see, for instance, Fire, A. et al. (1998) use RNAi in mammalian cells focused on the use of long 50 Nature 391:806-811). In this primitive organism, long strands of dsRNA. However, these attempts to induce RNAi dsRNA molecules are cleaved into siRNA by Dicer, thus met with limited success, due in part to the induction of the generating a diverse population of duplexes that can poten interferon response, which results in a general, as opposed to tially cover the entire transcript. While some fraction of these a target-specific, inhibition of protein synthesis. Thus, long molecules are non-functional (i.e. induce little or no silenc dsRNA is not a viable option for RNAi in mammalian sys 55 ing) one or more have the potential to be highly functional, temS. thereby silencing the gene of interest and alleviating the need More recently it has been shown that when short (18-30 bp) for siRNA design. Unfortunately, due to the interferon RNA duplexes are introduced into mammalian cells in cul response, this same approach is unavailable for mammalian ture, sequence-specific inhibition of target mRNA can be systems. While this effect can be circumvented by bypassing realized without inducing an interferon response. Certain of 60 the Dicer cleavage step and directly introducing siRNA, this these short dsRNAs, referred to as small inhibitory RNAs tactic carries with it the risk that the chosen siRNA sequence (“siRNAs), can act catalytically at sub-molar concentrations may be non-functional or semi-functional. to cleave greater than 95% of the target mRNA in the cell. A A number of researches have expressed the view that description of the mechanisms for siRNA activity, as well as siRNA design is not a crucial element of RNAi. On the other some of its applications are described in Provost et al., Ribo 65 hand, others in the field have begun to explore the possibility nuclease Activity and RNA Binding of Recombinant Human that RNAi can be made more efficient by paying attention to Dicer, E.M.B.O.J., 2002 Nov. 1; 21 (21): 5864-5874; Tabara the design of the siRNA. Unfortunately, none of the reported US 7,820,809 B2 3 4 methods have provided a satisfactory scheme for reliably Us=1 if U is the base at position 5 on the sense strand, selecting siRNA with acceptable levels of functionality. otherwise its value is 0; or Accordingly, there is a need to develop rational criteria by which to select siRNA with an acceptable level of function ality, and to identify siRNA that have this improved level of 5 functionality, as well as to identify siRNAs that are hyper functional.

SUMMARY OF THE INVENTION 10 The present invention is directed to increasing the effi ciency of RNAi, particularly in mammalian systems. Accord ingly, the present invention provides kits, siRNAS and meth ods for increasing siRNA efficacy. According to one embodiment, the present invention pro 15 vides a kit for gene silencing, wherein said kit is comprised of wherein a pool of at least two siRNA duplexes, each of which is A=1 if A is the base at position 1 of the sense strand, comprised of a sequence that is complementary to a portion of otherwise its value is 0; the sequence of one or more target messenger RNA. A=1 if A is the base at position 2 of the sense strand, According to a second embodiment, the present invention otherwise its value is 0; provides a method for optimizing RNA interference by using A=1 if A is the base at position 3 of the sense strand, one or more siRNAS that are optimized according to a for otherwise its value is 0; mula (or algorithm) selected from: A 1 if A is the base at position 4 of the sense Strand, Relative functionality of siRNA=-(GC/3)+(AUso)- 25 otherwise its value is 0; (Tm2o.o. )*3-(G1)*3-(Co)+(Ag)*2+(A)+ As=1 if A is the base at position 5 of the sense strand, (Ulo)+(A14)-(US)-(A11) Formula I otherwise its value is 0; A=1 if A is the base at position 6 of the sense strand, Relative functionality of siRNA=-(GC/3)-(A.U.s.o.) otherwise its value is 0; *3-(G1)*3-(Co)+(Ag)*2+(A) Formula II 30 A, -1 if A is the base at position 7 of the sense strand, otherwise its value is 0; Relative functionality of siRNA=-(GC/3)+(AUS)- Ao 1 if A is the base at position 10 of the sense strand, (Tm2o.o. )*3 Formula III otherwise its value is 0; Relative functionality of siRNA=-GC/2+(AUso)/2- A=1 if A is the base at position II of the sense strand, (Tm2)*2-(G)*3-(Co)+(A)*2+(A)+ 35 otherwise its value is 0; (Ulo)+(A14)-(US)-(A11) Formula IV A=1 if A is the base at position 13 of the sense strand, otherwise its value is 0; Relative functionality of siRNA=-(G)*3-(Co)+ A=1 if A is the base at position 19 of the sense strand, (Ag)*2+(A)+(Ulo)+(A)-(US)-(A11) Formula V otherwise if another base is present or the sense strand is 40 only 18 base pairs in length, its value is 0; Relative functionality of siRNA=-(G)*3-(Co)+ C=1 if C is the base at position 3 of the sense strand, (Ag)*2+(A) Formula VI otherwise its value is 0; C=1 if C is the base at position 4 of the sense strand, Relative functionality of siRNA=-(GC/2)+(AUso), 2-(Tn-os)*1-(G1)*3-(Co)+(Ag)*3+(A)*3+ otherwise its value is 0; 45 Cs=1 if C is the base at position 5 of the sense strand, (Ulo)/2+(A)/2-(Us), 2-(A11)/2 Formula VII otherwise its value is 0; wherein in Formulas I-VII: C=1 if C is the base at position 6 of the sense strand, otherwise its value is 0; Tmos =1 if the Tm is greater than 20° C.; C7–1 if C is the base at position 7 of the sense strand, A-1 if A is the base at position 19 on the sense Strand, 50 otherwise its value is 0; otherwise its value is 0; C=1 if C is the base at position 9 of the sense strand, AU so 0-5 depending on the number of A or U bases on otherwise its value is 0; the sense strand at positions 15-19; C, 1 if C is the base at position 17 of the sense strand, G=1 if G is the base at position 13 on the sense strand, otherwise its value is 0; otherwise its value is 0; 55 Cs-1 if C is the base at position 18 of the sense strand, C=1 if C is the base at position 19 of the sense strand, otherwise its value is 0; otherwise its value is 0; Co-1 if C is the base at position 19 of the sense strand, otherwise if another base is present or the sense strand is GC=the number of G and C bases in the entire sense strand; only 18 base pairs in length, its value is 0; A=1 if A is the base at position 3 on the sense strand, 60 G=1 if 6 is the base at position 1 on the sense strand, otherwise its value is 0; otherwise its value is 0; A=1 if A is the base at position 11 on the sense strand, G=1 if G is the base at position 2 of the sense strand, otherwise its value is 0; otherwise its value is 0; A 1 if A is the base at position 14 on the sense Strand, Gs=1 if G is the base at position 8 on the sense strand, otherwise its value is 0; 65 otherwise its value is 0; U=1 if U is the base at position 10 on the sense strand, Go-1 if G is the base at position 10 on the sense strand, otherwise its value is 0; otherwise its value is 0; US 7,820,809 B2 5 6 G=1 if G is the base at position 13 on the sense strand, According to this embodiment, preferably the set of siR otherwise its value is 0; NAs comprises at least 90 siRNAs from at least one gene, G=1 if G is the base at position 19 of the sense strand, more preferably at least 180 siRNAs from at least two differ otherwise if another base is present or the sense strand is ent genes, and most preferably at least 270 and 360 siRNAs only 18 base pairs in length, its value is 0; from at least three and four different genes, respectively. U=1 if U is the base at position 1 on the sense Strand, Additionally, in step (d) the determination is made with pref otherwise its value is 0; erably at least two, more preferably at least three, even more U=1 if U is the base at position 2 on the sense Strand, preferably at least four, and most preferably all of the vari otherwise its value is 0; ables. The resulting algorithm is not target sequence specific. U=1 if U is the base at position 3 on the sense Strand, 10 In a fifth embodiment, the present invention provides ratio otherwise its value is 0; nally designed siRNAs identified using the formulas above. U-1 if U is the base at position 4 on the sense Strand, In a sixth embodiment, the present invention is directed to otherwise its value is 0; hyperfunctional siRNA. U-1 if U is the base at position 7 on the sense Strand, For a better understanding of the present invention together otherwise its value is 0; 15 with other and further advantages and embodiments, refer U=1 if U is the base at position 9 on the sense Strand, ence is made to the following description taken in conjunction otherwise its value is 0; with the examples, the scope of which is set forth in the U=1 if U is the base at position 10 on the sense strand, appended claims. otherwise its value is 0; Us=1 if U is the base at position 15 on the sense strand, BRIEF DESCRIPTION OF THE FIGURES otherwise its value is 0; U-1 if U is the base at position 16 on the sense strand, FIG. 1 shows a model for siRNA-RISC interactions. RISC otherwise its value is 0; has the ability to interact with either end of the siRNA or U7–1 if U is the base at position 17 on the sense strand, miRNA molecule. Following binding, the duplex is 25 unwound, and the relevant target is identified, cleaved, and otherwise its value is 0; released. Us=1 if U is the base at position 18 on the sense strand, FIG. 2 is a representation of the functionality of two hun otherwise its value is 0; dred and seventy siRNA duplexes that were generated to GCso the number of G and C bases within positions target human cyclophilin, human diazepam-binding inhibitor 15-19 of the sense strand or within positions 15-18 if the 30 (DB), and firefly luciferase. sense Strand is only 18 base pairs in length; FIG. 3a is a representation of the silencing effect of 30 GC, the number of G and C bases in the sense strand; siRNAs in three different cells lines, HEK293, DU145, and Tm=100 if the targeting site contains an inverted repeat Hela. FIG. 3b shows the frequency of different functional longer than 4 base pairs, otherwise its value is 0; and groups (>95% silencing (black), >80% silencing (gray), X=the number of times that the same nucleotide repeats 35 >50% silencing (dark gray), and <50% silencing (white)) four or more times in a row. based on GC content. In cases where a given bar is absent According to a third embodiment, the present invention is from a particular GC percentage, no siRNA were identified directed to a kit comprised of at least one siRNA that contains for that particular group. FIG. 3c shows the frequency of a sequence that is optimized according to one of the formulas different functional groups based on melting temperature above. Preferably the kit contains at least two optimized 40 (Tm). Again, each group has four different divisions: >95% siRNA, each of which comprises a duplex, wherein one (black), >80% (gray), >50% (dark gray), and <50% (white) Strand of each duplex comprises at least eighteen contiguous silencing. bases that are complementary to a region of a target messen FIG. 4 is a representation of a statistical analysis that ger RNA. For mammalian systems, the siRNA preferably revealed correlations between silencing and five sequence comprises between 18 and 30 nucleotide base pairs. 45 related properties of siRNA: (A) an A at position 19 of the The ability to use the above algorithms, which are not sense Strand, (B) an A at position 3 of the sense strand, (C) a sequence or species specific, allows for the cost-effective Uat position 10 of the sense strand, (D) a base other than Gat selection of optimized siRNAS for specific target sequences. position 13 of the sense strand, and (E) a base other than Cat Accordingly, there will be both greater efficiency and reli position 19 of the sense strand. All variables were correlated ability in the use of siRNA technologies. 50 with siRNA silencing of firefly luciferase and human cyclo According to a fourth embodiment, the present invention philin. SiRNAs satisfying the criterion are grouped on the left provides a method for developing an siRNA algorithm for (Selected) while those that do not, are grouped on the right selecting functional and hyperfunctional siRNAS for a given (Eliminated). Y-axis is “96 Silencing of Control. Each posi sequence. The method comprises: tion on the X-axis represents a unique siRNA. (a) selecting a set of siRNAs; 55 FIGS.5A and 5 B are representations of firefly luciferase (b) measuring the gene silencing ability of each siRNA and cyclophilin siRNA panels Sorted according to function from said set; ality and predicted values using Formula VIII. The siRNA (c) determining the relative functionality of each siRNA; found within the circle represent those that have Formula VIII (d) determining the amount of improved functionality by values (SMARTscoresTM) above Zero. SiRNA outside the the presence or absence of at least one variable selected 60 indicated area have calculated Formula VIII values that are from the group consisting of the total GC content, melt below zero. Y-axis is “Expression (% Control). Each posi ing temperature of the siRNA, GC content at positions tion on the X-axis represents a unique siRNA. 15-19, the presence or absence of a particular nucleotide FIG. 6A is a representation of the average internal stability at a particular position and the number of times that the profile (AISP) derived from 270 siRNAs taken from three same nucleotide repeats within a given sequence; and 65 separate genes (cyclophilin B, DBI and firefly luciferase). (e) developing an algorithm using the information of step Graphs represent AISP values of highly functional, func (d). tional, and non-functional siRNA. FIG. 6B is a comparison US 7,820,809 B2 7 8 between the AISP of naturally derived GFP siRNA (filled 29, 31,33,35, 37,39, 41,43, 45, 47,49,51,53,55, 57, 59, 61, squares) and the AISP of siRNA from cyclophilin B, DBI, and 63, 65, 67, 69, 71, 73, 75, 77, 79, 81, 83, 85, 87, 89, and luciferase having >90% silencing properties (no fill) for the Plasmid. antisense strand. “DG' is the symbol for AG, free energy. FIGS. 16A and 16B are histograms demonstrating the inhi FIG. 7 is a histogram showing the differences in duplex bition of target gene expression by pools of 2 (16A) and 3 functionality upon introduction of basepair mismatches. The (16B) siRNA duplexes taken from the walk described in FIG. X-axis shows the mismatch introduced inot the siRNA and 15. The Y-axis in each represents the percent expression rela the position it is introduced (e.g., 8C-> A reveals that position 8 (which normally has a C) has been changed to an A). The tive to control. The X-axis in each represents the position of Y-axis is “96 Silencing (Normalized to Control).' 10 the first siRNA in paired pools, or trios of siRNAs. For FIG. 8a is histogram that shows the effects of 5'sense and instance, the first paired pool contains siRNAs 1 and 3. The antisense strand modification with 2'-O-methylation on func second paired pool contains siRNAs3 and 5. Pool 3 (of paired tionality. FIG. 8b is an expression profile showing a compari pools) contains siRNAS 5 and 7, and so on. For each of 16A son of sense strand off-target effects for IGF1R-3 and 2'-O- and 16B, the X-axis from left to right reads 1,3,5,7,9,11, 13, methyl IGF1R-3. Sense strand off-targets (lower white box) 15 15, 17, 19, 21, 23, 25, 27, 29, 31, 33,35, 37, 39, 41, 43, 45,47, are not induced when the 5' end of the sense strand is modified 49,51,53,55,57, 59, 61,63, 65, 67, 69,71, 73,75, 77,79, 81, with 2'-O-methyl groups (top white box). 83, 85, 87, 89, and Plasmid. FIG. 9 shows a graph of SMARTscoresTM versus RNAi FIGS. 17A and 17B are histograms demonstrating the inhi silencing values for more than 360 siRNA directed against 30 bition of target gene expression by pools of 4 (17A) and 5 different genes. SiRNA to the right of the vertical bar repre (17B) siRNA duplexes. The Y-axis in each represents the sent those siRNA that have desirable SMARTscoresTM. percent expression relative to control. The X-axis in each FIGS. 10A-E compare the RNAi of five different genes represents the position of the first siRNA in each pool. For (SEAP DBI, PLK, Firefly Luciferase, and Renila Luciferase) each of 17A and 17B, the X-axis from left to right reads 1, 3, by varying numbers of randomly selected siRNA and four 5, 7,9,11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31,33,35, 37,39, rationally designed (SMART-selected) siRNA chosen using 25 41, 43,45, 47,49,51,53,55, 57, 59, 61, 63, 65, 67, 69, 71,73, the algorithm described in Formula VIII. In addition, RNAi 75, 77, 79, 81, 83, 85, 87 89 and Plasmid. induced by a pool of the four SMART-selected siRNA is FIGS. 18A and 18B are histograms demonstrating the inhi reported at two different concentrations (100 and 400 nM). bition of target gene expression by siRNAs that are ten (18A) 10F is a comparison between a pool of randomly selected and twenty (18B) base pairs apart. The Y-axis represents the EGFR siRNA (Pool 1) and a pool of SMART selected EGFR 30 percent expression relative to a control. The X-axis represents siRNA (Pool 2). Pool 1, S1-S4 and Pool 2S1-S4 represent the the position of the first siRNA in each pool. For each of 18A individual members that made up each respective pool. Note and 18B, the X-axis from left to right reads 1,3,5,7,9,11, 13, that numbers for random siRNAs represent the position of the 15, 17, 19, 21, 23, 25, 27, 29, 31, 33,35, 37, 39, 41, 43, 45,47, 5' end of the sense strand of the duplex. The Y-axis represents 49,51,53,55,57, 59, 61,63, 65, 67, 69,71, 73,75, 77,79, 81, the 96 expression of the control(s). The X-axis is the percent 35 83, 85, 87, 89, and Plasmid. expression of the control. FIG. 19 shows that pools of siRNAs (dark gray bar) work FIG. 11 shows the Western blot results from cells treated as well (or better) than the best siRNA in the pool (light gray with siRNA directed against twelve different genes involved bar). The Y-axis represents the percent expression relative to in the clathrin-dependent endocytosis pathway (CHC, DynII, a control. The X-axis represents the position of the first CALM, CLCa, CLCb, Eps15, Eps15R, Rab5a, Rab5b, 40 siRNA in each pool. The X-axis from left to right reads 1, 3, Rab5c, B2 subunit of AP-2 and EEA.1). SiRNA were selected 5, 7,9,11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31,33,35, 37,39, using Formula VIII. “Pool represents a mixture of duplexes 41, 43,45, 47,49,51,53,55, 57, 59, 61, 63, 65, 67, 69, 71,73, 1-4. Total concentration of each siRNA in the pool is 25 nM. 75, 77, 79, 81, 83, 85, 87, 89, and Plasmid. Total concentration=4x25=100 nM. 45 FIG. 20 shows that the combination of several semifunc FIG. 12 is a representation of the gene silencing capabili tional siRNAs (dark gray) result in a significant improvement ties of rationally-selected siRNA directed against ten differ of gene expression inhibition over individual (semi-func ent genes (human and mouse cyclophilin, C-myc, human tional; light gray) siRNA. The Y-axis represents the percent lamin A/C, QB (ubiquinol-cytochrome c reductase core pro expression relative to a control. tein I). MEK1 and MEK2, ATE1 (arginyl-tRNA protein trans 50 FIG.21A, 21 Band 21C show both pools (Library, Lib) and ferase), GAPDH, and Eg5). The Y-axis is the percent expres individual siRNAs in inhibition of gene expression of Beta sion of the control. Numbers 1,2,3 and 4 represent individual Galactosidase, Renilla Luciferase and SEAP (alkaline phos rationally selected siRNA. “Pool represents a mixture of the phatase). Numbers on the X-axis indicate the position of the four individual siRNA. 5'-most nucleotide of the sense strand of the duplex. The FIG. 13 is the sequence of the top ten Bcl2 siRNAs as 55 Y-axis represents the percent expression of each gene relative determined by Formula VIII. Sequences are listed 5' to 3'. to a control. Libraries contain siRNAs that begin at the fol FIG. 14 is the knockdown by the top ten Bcl2 siRNAs at lowing nucleotides: FIG. 21B: Seap: Lib 1: 206, 766, 812, 100 nM concentrations. The Y-axis represents the amount of 923, Lib 2: 1117, 1280, 1300, 1487, Lib 3: 206,766,812,923, expression relative to the non-specific (ns) and transfection 1117, 1280, 1300, 1487, Lib 4: 206, 812, 1117, 1300, Lib 5: mixture control. 60 766,923, 1280, 1487, Lib 6: 206, 1487; FIG. 21A: Bgal: Lib FIG. 15 represents a functional walk where siRNA begin 1 (denoted as I on the figure): 979, 1339, 2029, 2590, Lib 2 ning on every other base pair of a region of the luciferase gene (denoted as II on the figure): 1087, 1783, 2399, 3257, Lib 3 are tested for the ability to silence the luciferase gene. The (denoted as III on the figure): 979, 1783, 2590, 3257, Lib 4 Y-axis represents the percent expression relative to a control. (denoted as IV on the figure): 979, 1087, 1339, 1783, 2029, The X-axis represents the position of each individual siRNA. 65 2399, 2590, 3257, Lib 5 (denoted as V on the figure): 979, Reading from left to right across the X-axis, the position 1087, 1339, 1783, Lib 6 (denoted as VI on the figure): 2029, designations are 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 2399, 2590, 3257; FIG. 21C: Renilla: Lib 1: 174, 300, 432, US 7,820,809 B2 9 10 568, Lib 2:592,633, 729,867, Lib 3: 174,300,432,568,592, refers to decreases in gene product expression that results 633, 729,867, Lib 4: 174,432,592,729, Lib 5:300,568,633, from RNA interference (RNAi), a defined, though partially 867, Lib 6:592,568. characterized pathway whereby small inhibitory RNA FIG.22 shows the results of an EGFR and TfnR internal (siRNA) act in concert with host proteins (e.g. the RNA ization assay when single gene knockdowns are performed. induced silencing complex, RISC) to degrade messenger The Y-axis represents percent internalization relative to con RNA (mRNA) in a sequence-dependent fashion. The level of trol. gene silencing can be measured by a variety of means, includ FIG. 23 shows the results of an EGFR and TfnR internal ing, but not limited to, measurement of transcript levels by ization assay when multiple genes are knocked down (e.g. Northern Blot Analysis, B-DNA techniques, transcription Rab5a, b, c). The Y-axis represents the percent internalization 10 sensitive reporter constructs, expression profiling (e.g. DNA relative to control. chips), and related technologies. Alternatively, the level of FIG. 24 shows the simultaneous knockdown of four differ ent genes. SiRNAs directed against G6PD, GAPDH, PLK, silencing can be measured by assessing the level of the pro and UBQ were simultaneously introduced into cells. Twenty tein encoded by a specific gene. This can be accomplished by four hours later, cultures were harvested and assayed for performing a number of studies including Western Analysis, mRNA target levels for each of the four genes. A comparison 15 measuring the levels of expression of a reporter protein that is made between cells transfected with individual siRNAs vs. has e.g. fluorescent properties (e.g. GFP) or enzymatic activ a pool of siRNAS directed against all four genes. ity (e.g. alkaline phosphatases), or several other procedures. FIG. 25 shows the functionality often siRNAs at 0.3 nM Transfection concentrations. The term “transfection” refers to a process by which agents DETAILED DESCRIPTION are introduced into a cell. The list of agents that can be transfected is large and includes, but is not limited to, siRNA, sense and/or anti-sense sequences, DNA encoding one or Definitions more genes and organized into an expression plasmid, pro Unless stated otherwise, the following terms and phrases teins, protein fragments, and more. There are multiple meth have the meanings provided below: 25 ods for transfecting agents into a cell including, but not lim siRNA ited to, electroporation, calcium phosphate-based transfections, DEAE-dextran-based transfections, lipid The term “siRNA” refers to small inhibitory RNA duplexes based transfections, molecular conjugate-based transfections that induce the RNA interference (RNAi) pathway. These (e.g. polylysine-DNA conjugates), microinjection and others. molecules can vary in length (generally between 18-30 base 30 pairS) and contain varying degrees of complementarity to Target their target mRNA in the antisense strand. Some, but not all, siRNA have unpaired overhanging bases on the 5' or 3' end of The term “target” is used in a variety of different forms the sense Strand and/or the antisense Strand. The term throughout this document and is defined by the context in “siRNA’ includes duplexes of two separate strands, as well as which it is used. “Target mRNA refers to a messenger RNA single strands that can form hairpin structures comprising a 35 to which a given siRNA can be directed against. “Target duplex region. sequence' and “target site' refer to a sequence within the SiRNA may be divided into five (5) groups (non-func mRNA to which the sense strand of an siRNA shows varying tional, semi-functional, functional, highly functional, and degrees of homology and the antisense strand exhibits vary hyper-functional) based on the level or degree of silencing ing degrees of complementarity. The term “siRNA target can that they induce in cultured cell lines. As used herein, these 40 refer to the gene, mRNA, or protein against which an siRNA definitions are based on a set of conditions where the siRNA is directed. Similarly “target silencing can refer to the state is transfected into said cell line at a concentration of 100 nM of a gene, or the corresponding mRNA or protein. and the level of silencing is tested at a time of roughly 24 Off-Target Silencing and Off-Target Interference hours after transfection, and not exceeding 72 hours after The phrases “off-target silencing and “off-target interfer transfection. In this context, “non-functional siRNA are 45 ence are defined as degradation of mRNA other than the defined as those siRNA that induce less than 50% (<50%) intended target mRNA due to overlapping and/or partial target silencing. "Semi-functional siRNA’ induce 50-79% homology with secondary mRNA messages. target silencing. "Functional siRNA are molecules that induce 80-95% gene silencing. “Highly-functional siRNA SMARTScoreTM are molecules that induce greater than 95% gene silencing. 50 The term “SMARTscoreTM refers to a number determined “Hyperfunctional siRNA are a special class of molecules. by applying any of the Formulas I-Formula IX to a given For purposes of this document, hyperfunctional siRNA are siRNA sequence. The term “SMART-selected” or “rationally defined as those molecules that: (1) induce greater than 95% selected' or “rational selection” refers to siRNA that have silencing of a specific target when they are transfected at been selected on the basis of their SMARTscoresTM. Subnanomolar concentrations (i.e., less than one nanomolar); and/or (2) induce functional (or better) levels of silencing for 55 Complementary greater than 96 hours. These relative functionalities (though The term “complementary” refers to the ability of poly not intended to be absolutes) may be used to compare siRNAs nucleotides to form base pairs with one another. Base pairs to a particular target for applications such as functional are typically formed by hydrogen bonds between nucleotide genomics, target identification and therapeutics. units in antiparallel polynucleotide Strands. Complementary miRNA 60 polynucleotide strands can base pair in the Watson-Crick The term “miRNA refers to microRNA. manner (e.g., A to T. A to U. C to G), or in any other manner that allows for the formation of duplexes. As persons skilled Gene Silencing in the art are aware, when using RNA as opposed to DNA, The phrase “gene silencing refers to a process by which uracil rather than thymine is the base that is considered to be the expression of a specific gene product is lessened or attenu 65 complementary to adenosine. However, when a U is denoted ated. Gene silencing can take place by a variety of pathways. in the context of the present invention, the ability to substitute Unless specified otherwise, as used herin, gene silencing a T is implied, unless otherwise stated. US 7,820,809 B2 11 12 Perfect complementarity or 100% complementarity refers Nucleotide to the situation in which each nucleotide unit of one poly The term “nucleotide' refers to a ribonucleotide or a deox nucleotide strand can hydrogen bond with a nucleotide unit of yribonucleotide or modified form thereof, as well as an ana a second polynucleotide Strand. Less than perfect comple log thereof. Nucleotides include species that comprise mentarity refers to the situation in which some, but not all, purines, e.g., adenine, hypoxanthine, guanine, and their nucleotide units of two strands can hydrogen bond with each derivatives and analogs, as well as pyrimidines, e.g., cytosine, other. For example, for two 20-mers, if only two base pairs on uracil, thymine, and their derivatives and analogs. each Strand can hydrogen bond with each other, the poly Nucleotide analogs include nucleotides having modifica nucleotide strands exhibit 10% complementarity. In the same tions in the chemical structure of the base, Sugar and/or phos 10 phate, including, but not limited to, 5-position pyrimidine example, if 18 base pairs on each Strand can hydrogen bond modifications, 8-position purine modifications, modifica with each other, the polynucleotide strands exhibit 90% tions at cytosine exocyclic amines, and Substitution of complementarity. “Substantial complementarity” refers to 5-bromo-uracil; and 2'-position Sugar modifications, includ polynucleotide strands exhibiting 79% or greater comple ing but not limited to, Sugar-modified ribonucleotides in mentarity, excluding regions of the polynucleotide strands, 15 which the 2'-OH is replaced by a group such as an H, OR, R, Such as overhangs, that are selected so as to be noncomple halo, SH, SR, NH, NHR, NR, or CN, wherein R is an alkyl mentary. (“Substantial similarity” refers to polynucleotide moiety. Nucleotide analogs are also meant to include nucle strands exhibiting 79% or greater similarity, excluding otides with bases such as inosine, queuosine, Xanthine, Sugars regions of the polynucleotide Strands, such as overhangs, that Such as 2'-methyl ribose, non-natural phosphodiester link are selected so as not to be similar.) Thus, for example, two ages Such as methylphosphonates, phosphorothioates and polynucleotides of 29 nucleotide units each, wherein each peptides. comprises a di-dT at the 3' terminus such that the duplex Modified bases refer to nucleotide bases such as, for region spans 27 bases, and wherein 26 of the 27 bases of the example, adenine, guanine, cytosine, thymine, uracil, Xan duplex region on each Strand are complementary, are Substan thine, inosine, and queuosine that have been modified by the tially complementary since they are 96.3% complementary 25 replacement or addition of one or more atoms or groups. when excluding the di-dT overhangs. Some examples of types of modifications that can comprise nucleotides that are modified with respect to the base moieties Deoxynucleotide include but are not limited to, alkylated, halogenated, thi The term “deoxynucleotide' refers to a nucleotide or poly olated, aminated, amidated, or acetylated bases, individually nucleotide lacking a hydroxyl group (OH group) at the 2' 30 or in combination. More specific examples include, for and/or 3' position of a Sugar moiety. Instead, it has a hydrogen example, 5-propynyluridine, 5-propynylcytidine, 6-methy bonded to the 2' and/or 3' carbon. Within an RNA molecule ladenine, 6-methylguanine, N.N.-dimethyladenine, 2-propy that comprises one or more deoxynucleotides, “deoxynucle ladenine, 2-propylguanine, 2-aminoadenine, 1-methyli otide' refers to the lack of an OH group at the 2'position of the nosine, 3-methyluridine, 5-methylcytidine, 5-methyluridine Sugar moiety, having instead a hydrogen bonded directly to 35 and other nucleotides having a modification at the 5 position, the 2" carbon. 5-(2-amino)propyl uridine, 5-halocytidine, 5-halouridine, 4-acetylcytidine, 1-methyladenosine, 2-methyladenosine, Deoxyribonucleotide 3-methylcytidine, 6-methyluridine, 2-methylguanosine, The terms “deoxyribonucleotide' and “DNA” refer to a 7-methylguanosine, 2,2-dimethylguanosine, 5-methylami nucleotide or polynucleotide comprising at least one Sugar 40 noethyluridine, 5-methyloxyuridine, deaZanucleotides Such moiety that has an H, rather than an OH, at its 2' and/or as 7-deazaadenosine, 6-azouridine, 6-azocytidine, 6-aZothy 3'position. midine, 5-methyl-2-thiouridine, other thio bases such as 2-thiouridine and 4-thiouridine and 2-thiocytidine, dihydrou Substantially Similar ridine, pseudouridine, queuosine, archaeosine, naphthyl and The phrase “substantially similar refers to a similarity of 45 Substituted naphthyl groups, any O- and N-alkylated purines at least 90% with respect to the identity of the bases of the and pyrimidines such as N6-methyladenosine, 5-methylcar Sequence. bonylmethyluridine, uridine 5-oxyacetic acid, pyridine-4- one, pyridine-2-one, phenyl and modified phenyl groups such Duplex Region as aminophenol or 2.4.6-trimethoxy benzene, modified The phrase “duplex region” refers to the region in two 50 cytosines that act as G-clamp nucleotides, 8-substituted complementary or Substantially complementary polynucle adenines and guanines, 5-substituted uracils and thymines, otides that form base pairs with one another, either by Wat azapyrimidines, carboxyhydroxyalkyl nucleotides, carboxy son-Crick base pairing or any other manner that allows for a alkylaminoalkyl nucleotides, and alkylcarbonylalkylated stabilized duplex between polynucleotide strands that are nucleotides. Modified nucleotides also include those nucle complementary or Substantially complementary. For 55 otides that are modified with respect to the Sugar moiety, as example, a polynucleotide strand having 21 nucleotide units well as nucleotides having Sugars or analogs thereof that are can base pair with another polynucleotide of 21 nucleotide not ribosyl. For example, the Sugar moieties may be, or be units, yet only 19 bases on each Strand are complementary or based on, mannoses, arabinoses, glucopyranoses, galactopy Substantially complementary, such that the “duplex region' ranoses, 4'-thioribose, and other Sugars, heterocycles, or car has 19 base pairs. The remaining bases may, for example, 60 bocycles. exist as 5' and 3' overhangs. Further, within the duplex region, The term nucleotide is also meant to include what are 100% complementarity is not required; substantial comple known in the art as universal bases. By way of example, mentarity is allowable within a duplex region. Substantial universal bases include but are not limited to 3-nitropyrrole, complementarity refers to 79% or greater complementarity. 5-nitroindole, or nebularine. The term “nucleotide' is also For example, a mismatch in a duplex region consisting of 19 65 meant to include the N3' to P5' phosphoramidate, resulting base pairs results in 94.7% complementarity, rendering the from the substitution of a ribosyl 3' oxygen with an amine duplex region Substantially complementary. group. US 7,820,809 B2 13 14 Further, the term nucleotide also includes those species ability of identifying a functional or hyperfunctional siRNA. that have a detectable label, such as for example a radioactive In one method, rationally designed siRNA can be identified or fluorescent moiety, or mass label attached to the nucle by maximizing one or more of the following criteria: otide. 1. A low GC content, preferably between about 30-52%. Polynucleotide 2. At least 2, preferably at least 3A or U bases at positions The term “polynucleotide' refers to polymers of nucle 15-19 of the siRNA on the sense strand. otides, and includes but is not limited to DNA, RNA, DNA/ 3. An A base at position 19 of the sense strand. RNA hybrids including polynucleotide chains of regularly 4. An A base at position 3 of the sense strand. and/or irregularly alternating deoxyribosyl moieties and ribo 10 5. AU base at position 10 of the sense strand. syl moieties (i.e., wherein alternate nucleotide units have an 6. An A base at position 14 of the sense Strand. OH, then and—H, then an OH, then an—H, and so on at the 2 position of a Sugar moiety), and modifications of these 7. A base other than Cat position 19 of the sense strand. kinds of polynucleotides, wherein the attachment of various 8. A base other than G at position 13 of the sense strand. entities or moieties to the nucleotide units at any position are 9. A Tm, which refers to the character of the internal repeat included. 15 that results in inter- or intramolecular structures for one strand of the duplex, that is preferably not stable at Polyribonucleotide greater than 50°C., more preferably not stable at greater The term “polyribonucleotide' refers to a polynucleotide than 37° C., even more preferably not stable at greater comprising two or more modified or unmodified ribonucle than 30° C. and most preferably not stable at greater than otides and/or their analogs. The term “polyribonucleotide' is 200 C. used interchangeably with the term "oligoribonucleotide.” 10. A base other than U at position 5 of the sense strand. Ribonucleotide and Ribonucleic Acid 11. A base other than A at position 11 of the sense strand. The term “ribonucleotide' and the phrase “ribonucleic Criteria 5, 6, 10 and 11 are minor criteria, but are nonethe acid (RNA), refer to a modified or unmodified nucleotide or 25 less desirable. Accordingly, preferably an siRNA will satisfy polynucleotide comprising at least one ribonucleotide unit. A as many of the aforementioned criteria as possible, more ribonucleotide unit comprises an hydroxyl group attached to preferably at least 1-4 and 7-9, and most preferably all of the the 2' position of a ribosyl moiety that has a nitrogenous base criteria attached in N-glycosidic linkage at the 1' position of a ribosyl With respect to the criteria, GC content, as well as a high moiety, and a moiety that either allows for linkage to another 30 number of AU in positions 15-19, may be important for ease nucleotide or precludes linkage. ment of the unwinding of double stranded siRNA duplex. Duplex unwinding has been shown to be crucial for siRNA DETAILED DESCRIPTION OF THE INVENTION functionality in vivo. With respect to criterion 9, the internal structure is mea The present invention is directed to improving the effi 35 Sured in terms of the melting temperature of the single strand ciency of gene silencing by siRNA. Through the inclusion of of siRNA, which is the temperature at which 50% of the multiple siRNA sequences that are targeted to a particular molecules will become denatured. With respect to criteria 2-8 gene and/or selecting an siRNA sequence based on certain and 10-11, the positions refer to sequence positions on the defined criteria, improved efficiency may be achieved. sense strand, which is the strand that is identical to the mRNA. The present invention will now be described in connection 40 with preferred embodiments. These embodiments are pre In one preferred embodiment, at least criteria 1 and 8 are sented in order to aid in an understanding of the present satisfied. In another preferred embodiment, at least criteria 7 invention and are not intended, and should not be construed, and 8 are satisfied. In still another preferred embodiment, at to limit the invention in any way. All alternatives, modifica least criteria 1, 8 and 9 are satisfied. It should be noted that all of the aforementioned criteria tions and equivalents that may become apparent to those of 45 ordinary skill upon reading this disclosure are included regarding sequence position specifics are with respect to the 5' within the spirit and scope of the present invention. end of the sense Strand. Reference is made to the sense strand, Furthermore, this disclosure is not a primer on RNA inter because most databases contain information that describes ference. Basic concepts known to persons skilled in the art the information of the mRNA. Because according to the have not been set forth in detail. 50 present invention a chain can be from 18 to 30 bases in length, Optimizing siRNA and the aforementioned criteria assumes a chain 19 base pairs According to one embodiment, the present invention pro in length, it is important to keep the aforementioned criteria vides a method for improving the effectiveness of gene silenc applicable to the correct bases. ing for use to silence a particular gene through the selection of When there are only 18 bases, the base pair that is not an optimal siRNA. An siRNA selected according to this 55 present is the base pair that is located at the 3' of the sense method may be used individually, or in conjunction with the strand. When there are twenty to thirty bases present, then first embodiment, i.e., with one or more other siRNAs, each of additional bases are added at the 5' end of the sense chain and which may or may not be selected by this criteria in order to occupy positions 1 to 11. Accordingly, with respect to SEQ. maximize their efficiency. ID NO. 0001. NNANANNNNUCNAANNNNA and SEQ. The degree to which it is possible to select an siRNA for a 60 ID NO. 0028. GUCNNANANNNNUCNAANNNNA, both given mRNA that maximizes these criteria will depend on the would have A at position 3, A at position 5, U at position 10, sequence of the mRNA itself. However, the selection criteria C at position 11, A and position 13, A and position 14 and A will be independent of the target sequence. According to this at position 19. However, SEQ. ID NO. 0028 would also have method, an siRNA is selected for a given gene by using a C at position -1, U at position -2 and G at position -3. rational design. That said, rational design can be described in 65 For a 19 base pair siRNA, an optimal sequence of one of the a variety of ways. Rational design is, in simplest terms, the strands may be represented below, where N is any base, A, C, application of a proven set of criteria that enhance the prob G, or U: US 7,820,809 B2 15 16 Relative functionality of siRNA=-(GC/2)+(AUso), 2-(Tn-os)*1-(G1)*3-(Co)+(Ag)*3+(A)*3+ SEO. ID NO. OOO1. INNANANNNNUCNAANNNNA (Ulo)/2+(A1), 2-(Us), 2-(A11)/2 Formula VII

SEO. ID NO. OOO2. INNANANNNNUGNAANNNNA In Formulas I-VII: 5 SEO. ID NO. OOO3. INNANANNNNUUNAANNNNA wherein A-1 if A is the base at position 19 on the sense strand, otherwise its value is 0, SEO. ID NO. OOO4. INNANANNNNUCNCANNNNA AU so 0-5 depending on the number of A or U bases on SEO. ID NO. OOO5. INNANANNNNUGNCANNNNA the sense strand at positions 15-19; 10 G=1 if G is the base at position 13 on the sense strand, SEO. ID NO. OOO6. INNANANNNNUUNCANNNNA otherwise its value is 0; Co-1 if C is the base at position 19 of the sense strand, SEO. ID NO. OOO7. INNANANNNNUCNUANNNNA otherwise its value is 0; SEO. ID NO. OOO8. . INNANANNNNUGNUANNNNA GC=the number of G and C bases in the entire sense strand; 15 Tmo. 1 if the Tm is greater than 20° C.; SEO. ID NO. OOO9. INNANANNNNUUNUANNNNA A=1 if A is the base at position 3 on the sense strand, SEO. ID NO. OO10. INNANCNNNNUCNAANNNNA otherwise its value is 0; U-1 if U is the base at position 10 on the sense strand, SEO. ID NO. OO11. INNANCNNNNUGNAANNNNA otherwise its value is 0; A 1 if A is the base at position 14 on the sense strand, SEO. ID NO. OO12. INNANCNNNNUUNAANNNNA otherwise its value is 0; SEO. ID NO. OO13. INNANCNNNNUCNCANNNNA Us=1 if U is the base at position 5 on the sense strand, otherwise its value is 0; and SEO. ID NO. OO14. INNANCNNNNUGNCANNNNA A=1 if A is the base at position 11 of the sense strand, 25 otherwise its value is 0. SEO. ID NO. OO15. INNANCNNNNUUNCANNNNA Formulas I-VII provide relative information regarding SEO. ID NO. OO16. INNANCNNNNUCNUANNNNA functionality. When the values for two sequences are com pared for a given formula, the relative functionality is ascer SEO. ID NO. OO17. INNANCNNNNUGNUANNNNA tained; a higher positive number indicates a greater function SEO. ID NO. OO18. INNANCNNNNUUNUANNNNA 30 ality. For example, in many applications a value of 5 or greater is beneficial. SEQ. ID NO. OO19. INNANGNNNNUCNAANNNNA Additionally, in many applications, more than one of these formulas would provide useful information as to the relative SEO. ID NO. OO2O. INNANGNNNNUGNAANNNNA functionality of potential siRNA sequences. However, it is beneficial to have more than one type of formula, because not SEO. ID NO. OO21. INNANGNNNNUUNAANNNNA 35 every formula will be able to help to differentiate among SEO. ID NO. OO22. INNANGNNNNUCNCANNNNA potential siRNA sequences. For example, in particularly high GC mRNAs, formulas that take that parameter into account SEO. ID NO. OO23. INNANGNNNNUGNCANNNNA would not be useful and application of formulas that lack GC SEO. ID NO. OO24. INNANGNNNNUUNCANNNNA elements (e.g., formulas V and VI) might provide greater 40 insights into duplex functionality. Similarly, formula II might SEO. ID NO. OO25. INNANGNNNNUCNUANNNNA by used in situations where hairpin structures are not observed in duplexes, and formula IV might be applicable for SEO. ID NO. OO26. INNANGNNNNUGNUANNNNA sequences that have higher AU content. Thus, one may con SEO. ID NO. OO27. INNANGNNNNNU.NUANNNNA sider a particular sequence in light of more than one or even 45 all of these algorithms to obtain the best differentiation In one embodiment, the sequence used as an siRNA is among sequences. In some instances, application of a given selected by choosing the siRNA that score highest according algorithim may identify an unusuSally large number of poten to one of the following seven algorithms that are represented tial siRNA sequences, and in those cases, it may be appropri by Formulas I-VII: ate to re-analyze that sequence with a second algorithm that 50 is, for instance, more stringent. Alternatively, it is conceivable Relative functionality of siRNA=-(GC/3)+(AUso)- that analysis of a sequence with a given formula yields no (Tm2)*3-(G)*3-(Co)+(A)*2+(A)+ acceptable siRNA sequences (i.e. low SMARTscoresTM). In (Ulo)+(A14)-(US)-(A11) Formula I this instance, it may be appropriate to re-analyze that sequences with a second algorithm that is, for instance, less Relative functionality of siRNA=-(GC/3)-(AUso) stringent. Instill other instances, analysis of a single sequence 55 with two separate formulas may give rise to conflicting results (i.e. one formula generates a set of siRNA with high SMART Relative functionality of siRNA=-(GC/3)+(AUso)- ScoresTM while the other formula identifies a set of siRNA (Tm2o.o. )*3 Formula III with low SMARTscoresTM). In these instances, it may be necessary to determine which weighted factor(s) (e.g. GC 60 content) are contributing to the discrepancy and assessing the sequence to decide whether these factors should or should not (Ulo)+(A14)-(US)-(A11) Formula IV be included. Alternatively, the sequence could be analyzed by a third, fourth, or fifth algorithm to identify a set of rationally Relative functionality of siRNA=-(G)*3-(Co)+ designed siRNA. 65 The above-referenced criteria are particularly advanta Relative functionality of siRNA=-(G)*3-(Co)+ geous when used in combination with pooling techniques as (Ag)*2+(A) Formula VI depicted in Table I: US 7,820,809 B2 18

TABLE I Functional Probability Oligos Pools (GC15)-30-X Formula IX: Criteria >95% -80% <70% -95% c-80% <70% wherein Current 33.0 SO.O 23.0 79.5 97.3 O.3 A=1 if A is the base at position 1 of the sense strand, New SO.O 88.5 8.0 93.8 99.98 O.OOS otherwise its value is 0; (GC) 28.0 S8.9 36.0 72.8 97.1 1.6 10 A=1 if A is the base at position 2 of the sense strand, otherwise its value is 0; The term "current refers to Tuschl's conventional siRNA A=1 if A is the base at position 3 of the sense strand, parameters (Elbashir, S. M. et al. (2002) “Analysis of gene otherwise its value is 0; function in Somatic mammalian cells using Small interfering 15 A 1 if A is the base at position 4 of the sense strand, RNAs Methods 26: 199-213). “New” refers to the design otherwise its value is 0; parameters described in Formulas I-VII. “GC refers to cri As 1 if A is the base at position 5 of the sense strand, teria that select siRNA solely on the basis of GC content. otherwise its value is 0; As Table I indicates, when more functional siRNA A=1 if A is the base at position 6 of the sense strand, duplexes are chosen, siRNAs that produce <70% silencing otherwise its value is 0; drops from 23% to 8% and the number of siRNA duplexes A, -1 if A is the base at position 7 of the sense strand, that produce >80% silencing rises from 50% to 88.5%. Fur otherwise its value is 0; ther, of the siRNA duplexes with >80% silencing, a larger A-1 if A is the base at position 10 of the sense strand, portion of these siRNAs actually silence >95% of the target otherwise its value is 0; expression (the new criteria increases the portion from 33% to 25 A=1 if A is the base at position 11 of the sense strand, 50%). Using this new criteria in pooled siRNAs, shows that, otherwise its value is 0; with pooling, the amount of silencing >95% increases from A=1 if A is the base at position 13 of the sense strand, 79.5% to 93.8% and essentially eliminates any siRNA pool otherwise its value is 0; from silencing less than 70%. A-1 if A is the base at position 19 of the sense strand, Table II similarly shows the particularly beneficial results 30 otherwise if another base is present or the sense strand is of pooling in combination with the aforementioned criteria. only 18 base pairs in length, its value is 0; However, Table II, which takes into account each of the afore C=1 if C is the base at position 3 of the sense strand, mentioned variables, demonstrates even a greater degree of otherwise its value is 0; improvement in functionality. C=1 if C is the base at position 4 of the sense strand, 35 otherwise its value is 0; TABLE II C=1 if C is the base at position 5 of the sense strand, otherwise its value is 0; Functional Probability C=1 if C is the base at position 6 of the sense strand, Oligos Pools otherwise its value is 0; 40 C7–1 if C is the base at position 7 of the sense strand, Func- Non- Func- Non otherwise its value is 0; tional Average functional tional Average functional C=1 if C is the base at position 9 of the sense strand, Random 2O 40 50 67 97 3 otherwise its value is 0; Criteria 1 52 99 O.1 97 93 O.OO40 C7–1 if C is the base at position 17 of the sense strand, Criteria 4 89 99 O.1 99 99 OOOOO otherwise its value is 0; 45 Cs-1 if C is the base at position 18 of the sense strand, otherwise its value is 0; The terms “functional.” “Average and “Non-functional C-1 if C is the base at position 19 of the sense strand, refer to siRNA that exhibit >80%, >50%, and <50% function otherwise if another base is present or the sense strand is ality, respectively. Criteria 1 and 4 refer to specific criteria 50 only 18 base pairs in length, its value is 0; described above. G=1 if G is the base at position 1 on the sense strand, The above-described algorithms may be used with or with otherwise its value is 0; out a computer program that allows for the inputting of the G=1 if G is the base at position 2 of the sense strand, sequence of the mRNA and automatically outputs the optimal otherwise its value is 0; siRNA. The computer program may, for example, be acces 55 Gs-1 if G is the base at position 8 on the sense strand, sible from a local terminal or personal computer, over an otherwise its value is 0; internal network or over the Internet. Go-1 if G is the base at position 10 on the sense strand, In addition to the formulas above, more detailed algorithms otherwise its value is 0; may be used for selecting siRNA. Preferably, at least one G=1 if G is the base at position 13 on the sense strand, RNA duplex of between 18 and 30 base pairs is selected such 60 otherwise its value is 0; that it is optimized according a formula selected from: G-1 if G is the base at position 19 of the sense strand, otherwise if another base is present or the sense strand is only 18 base pairs in length, its value is 0; U=1 if U is the base at position 1 on the sense strand, 65 otherwise its value is 0; U=1 if U is the base at position 2 on the sense strand, (GCIS-19)-30 X; and Formula VIII: otherwise its value is 0; US 7,820,809 B2 19 20 U=1 if U is the base at position 3 on the sense strand, tors extended this observation to siRNAs for which function otherwise its value is 0; ality had been assessed in tissue culture. U=1 if U is the base at position 4 on the sense strand, With respect to the parameter GCs, a value of 0-5 will otherwise its value is 0; be ascribed depending on the number of G or C bases at U7–1 if U is the base at position 7 on the sense strand, positions 15 to 19. If there are only 18 base pairs, the value is otherwise its value is 0; between 0 and 4. U=1 if U is the base at position 9 on the sense strand, With respect to the criterion GC content, a number from otherwise its value is 0; 0-30 will be ascribed, which correlates to the total number of U-1 if U is the base at position 10 on the sense strand, G and C nucleotides on the sense strand, excluding over otherwise its value is 0; 10 hangs. Without wishing to be bound by any one theory, it is Us=1 if U is the base at position 15 on the sense strand, postulated that the significance of the GC content (as well as otherwise its value is 0; AU content at positions 15-19, which is a parameter for U=1 if U is the base at position 16 on the sense strand, formulas III-VII) relates to the easement of the unwinding of otherwise its value is 0; a double-stranded siRNA duplex. Duplex unwinding is U7–1 if U is the base at position 17 on the sense strand, 15 believed to be crucial for siRNA functionality in vivo and otherwise its value is 0; overall low internal stability, especially low internal stability Us=1 if U is the base at position 18 on the sense strand, of the first unwound base pair is believed to be important to otherwise its value is 0; maintain sufficient processivity of RISC complex-induced GCso the number of G and C bases within positions duplex unwinding. If the duplex has 19 base pairs, those at 15-19 of the sense strand, or within positions 15-18 if the positions 15-19 on the sense strand will unwind first if the sense Strand is only 18 base pairs in length; molecule exhibits a sufficiently low internal stability at that GC, the number of G and C bases in the sense strand; position. As persons skilled in the art are aware, RISC is a Tm=100 if the siRNA oligo has the internal repeat longer complex of approximately twelve proteins; Dicer is one, but then 4 base pairs, otherwise its value is 0; and not the only, helicase within this complex. Accordingly, X=the number of times that the same nucleotide repeats 25 although the GC parameters are believed to relate to activity four or more times in a row. with Dicer, they are also important for activity with other The above formulas VIII and IX, as well as formulas I-VII, RISC proteins. provide methods for selecting siRNA in order to increase the efficiency of gene silencing. A Subset of variables of any of The value of the parameter Tm is 0 when there are no the formulas may be used, though when fewer variables are 30 internal repeats longer than (or equal to) four base pairs used, the optimization hierarchy becomes less reliable. present in the siRNA duplex; otherwise the value is 1. Thus With respect to the variables of the above-referenced for for example, if the sequence ACGUACGU, or any other four mulas, a single letter of A or C or G or U followed by a nucleotide (or more) palindrome exists within the structure, subscript refers to a binary condition. The binary condition is the value will be one (1). Alternatively if the structure ACG that either the particular base is present at that particular 35 GACG, or any other 3 nucleotide (or less) palindrome exists, position (wherein the value is “1”) or the base is not present the value will be zero (0). (wherein the value is “0”). Because position 19 is optional, The variable “X” refers to the number of times that the i.e. there might be only 18 base pairs, when there are only 18 same nucleotide occurs contiguously in a stretch of four or base pairs, any base with a subscript of 19 in the formulas more units. If there are, for example, four contiguous AS in above would have a Zero value for that parameter. Before or 40 one part of the sequence and elsewhere in the sequence four after each variable is a number followed by *, which indicates contiguous Cs, X-2. Further, if there are two separate con that the value of the variable is to be multiplied or weighed by tiguous stretches of four of the same nucleotides or eight or that number. more of the same nucleotides in a row, then X=2. However, X The numbers preceding the variables A, or G, or C, or U in does not increase for five, six or seven contiguous nucle Formulas VIII and IX (or after the variables in Formula I-VII) 45 otides. were determined by comparing the difference in the fre Again, when applying Formula VIII or Formula IX to a quency of individual bases at different positions in functional given mRNA, (the “target RNA or “target molecule'), one siRNA and total siRNA. Specifically, the frequency in which may use a computer program to evaluate the criteria for every a given base was observed at a particular position in func sequence of 18-30 base pairs or only sequences of a fixed tional groups was compared with the frequency that that same 50 length, e.g., 19 base pairs. Preferably the computer program is base was observed in the total, randomly selected siRNA set. designed such that it provides a report ranking of all of the If the absolute value of the difference between the functional potential siRNAs between 18 and 30 base pairs, ranked and total values was found to be greater than 6%, that param according to which sequences generate the highest value. A eter was included in the equation. Thus for instance, if the higher value refers to a more efficient siRNA for a particular frequency of finding a “G” at position 13 (G) is found to be 55 target gene. The computer program that may be used, may be 6% in a given functional group, and the frequency of G 3 in developed in any computer language that is known to be the total population of siRNAs is 20%, the difference between useful for scoring nucleotide sequences, or it may be devel the two values is 6%-20%-- 14%. As the absolute value is oped with the assistance of commercially available product greater than six (6), this factor (-14) is included in the equa such as Microsoft's product net. Additionally, rather than run tion. Thus in Formula VIII, in cases where the siRNA under 60 every sequence through one and/or anotherformula, one may study has a G in position 13, the accrued value is (-14)* compare a Subset of the sequences, which may be desirable if (1)=-14. In contrast, when a base other than G is found at for example only a Subset are available. For instance, it may position 13, the accrued value is (-14)*(0)=0. be desirable to first perform a BLAST (Basic Local Align When developing a means to optimize siRNAs, the inven ment Search Tool) search and to identify sequences that have tors observed that a bias toward low internal thermodynamic 65 no homology to other targets. Alternatively, it may be desir stability of the duplex at the 5'-antisense (AS) end is charac able to scan the sequence and to identify regions of moderate teristic of naturally occurring miRNA precursors. The inven GC context, then perform relevant calculations using one of US 7,820,809 B2 21 22 the above-described formulas on these regions. These calcu The siRNA selected according to the aforementioned cri lations can be done manually or with the aid of a computer. teria or one of the aforementioned algorithms are also, for As with Formulas I-VII, either Formula VIII or Formula IX example, useful in the simultaneous screening and functional may be used for a given mRNA target sequence. However, it analysis of multiple genes and gene families using high is possible that according to one or the other formula more throughput strategies, as well as in direct gene Suppression or than one siRNA will have the same value. Accordingly, it is silencing. beneficial to have a second formula by which to differentiate Development of the Algorithms sequences. Formula IX was derived in a similar fashion as To identify siRNA sequence features that promote func Formula VIII, yet used a larger data set and thus yields tionality and to quantify the importance of certain currently sequences with higher statistical correlations to highly func 10 accepted conventional factors—such as G/C content and tar tional duplexes. The sequence that has the highest value get site accessibility—the inventors synthesized an siRNA ascribed to it may be referred to as a “first optimized duplex.” panel consisting of 270 siRNAS targeting three genes, Human The sequence that has the second highest value ascribed to it Cyclophilin, Firefly Luciferase, and Human DBI. In all three may be referred to as a “second optimized duplex. Similarly, cases, siRNAS were directed against specific regions of each the sequences that have the third and fourth highest values 15 gene. For Human Cyclophilin and Firefly Luciferase, ninety ascribed to them may be referred to as a third optimized siRNAs were directed against a 199 bp segment of each duplex and a fourth optimized duplex, respectively. When respective mRNA. For DBI, 90 siRNAs were directed against more than one sequence has the same value, each of them a smaller, 109 base pair region of the mRNA. The sequences may, for example, be referred to as first optimized duplex to which the siRNAs were directed are provided below. sequences or co-first optimized duplexes. It should be noted that in certain sequences, “t is present. SiRNA sequences identified using Formula VIII are con This is because many databases contain information in this tained within the enclosed compact disks. The data included manner. However, the t denotes auracil residue in mRNA and on the enclosed compact disks is described more fully below. siRNA. Any algorithm will, unless otherwise specified, pro The sequences identified by Formula VIII that are disclosed in cess at in a sequence as a u. the compacts disks may be used in gene silencing applica 25 tions. Human Cyclophilin: 193-390, M60857 It should be noted that for Formulas VIII and IX all of the aforementioned criteria are identified as positions on the SEO. ID NO. 29: sense strand when oriented in the 5' to 3’ direction as they are gttcCaaaaa.cagtggata attttgttggccttagct acaggagagaaagg identified in connection with Formulas I-VII unless otherwise 30 specified. atttggctacaaaaa.ca.gcaaattic catcgtgtaatcaagg actt catga Formulas I-IX, may be used to selector to evaluate one, or more than one, siRNA in order to optimize silencing. Prefer to Cagggcggagactitcaccaggggagatggcacaggaggaaagagcatc ably, at least two optimized siRNAs that have been selected tacggtgagcgct tcc.ccgatgagaact tcaaactgaag cact acgggcc according to at least one of these formulas are used to silence 35 a gene, more preferably at least three and most preferably at tggctggg least four. The siRNAs may be used individually or together in a pool or kit. Further, they may be applied to a cell simul Firefly Luciferase: 1434-1631, U47298 (pGL3, Promega) taneously or separately. Preferably, the at least two siRNAs are applied simultaneously. Pools are particularly beneficial 40 for many research applications. However, for therapeutics, it SEO. ID NO. 3 O: may be more desirable to employ a single hyperfunctional tgaact tcc.cgcc.gc.cgttgttgttittggagcacggaaagacgatgacgg siRNA as described elsewhere in this application. aaaaagagat.cgtggattacgtc.gc.cagt caagtaacaa.ccgcgaaaaag When planning to conduct gene silencing, and it is neces sary to choose between two or more siRNAs, one should do so 45 ttgcgcggaggagttgttgtttgttggacgaagtaccgaaaggtott accgg by comparing the relative values when the siRNA are sub jected to one of the formulas above. In generala higher scored aaaact cacgcaagaaaaat cagagagat.cct catalaaggccaagaagg siRNA should be used. DBI, NM 020548 (202-310) (every position) Useful applications include, but are not limited to, target validation, gene functional analysis, research and drug dis 50 covery, gene therapy and therapeutics. Methods for using SEO. ID NO. OO31: siRNA in these applications are well known to persons of skill acgggcaaggccalagtgggatgcctggaatgagctgaaagggact tccala in the art. ggaagatgcc atgaaagdulacat Caacaaagtagaagagctaaagaaaaa Because the ability of siRNA to function is dependent on the sequence of the RNA and not the species into which it is 55 at acggg introduced, the present invention is applicable across abroad range of species, including but not limited to all mammalian A list of the siRNAs appears in Table IV (see Examples species, such as humans, dogs, horses, cats, cows, mice, ham Section, Example I) sters, chimpanzees and gorillas, as well as other species and The set of duplexes was analyzed to identify correlations organisms such as bacteria, viruses, insects, plants and C. 60 between siRNA functionality and other biophysical or ther elegans. modynamic properties. When the siRNA panel was analyzed The present invention is also applicable for use for silenc in functional and non-functional Subgroups, certain nucle ing a broad range of genes, including but not limited to the otides were much more abundant at certain positions in func roughly 45,000 genes of a human genome, and has particular tional or non-functional groups. More specifically, the fre relevance in cases where those genes are associated with 65 quency of each nucleotide at each position in highly diseases such as diabetes, Alzheimer's, cancer, as well as all functional siRNA duplexes was compared with that of non genes in the genomes of the aforementioned organisms. functional duplexes in order to assess the preference for or US 7,820,809 B2 23 24 againstany given nucleotide at every position. These analyses modynamic flexibility of the duplex 5'-end (positions 1-5; S) were used to determine important criteria to be included in the did not appear to correlate appreciably with silencing siRNA algorithms (Formulas VIII and IX). potency, while that of the 3'-end (positions 15-19; S) corre The data set was also analyzed for distinguishing biophysi lated with efficient silencing. No duplexes lacking A/Ubp in cal properties of siRNAS in the functional group. Such as positions 15-19 were functional. The presence of one A/Ubp optimal percent of GC content, propensity for internal struc in this region conferred some degree of functionality, but the tures and regional thermodynamic stability. Of the presented presence of three or more A/Us was preferable and therefore criteria, several are involved in duplex recognition, RISC defined as Criterion 11. When applied to the test panel, only activation/duplex unwinding, and target cleavage catalysis. a marginal increase in the probability of functional siRNA The original data set that was the source of the statistically 10 derived criteria is shown in FIG. 2. Additionally, this figure selection was achieved: a 1.8% and 2.3% increase for F50 and shows that random selection yields siRNA duplexes with F95 duplexes, respectively (Table III). unpredictable and widely varying silencing potencies as mea The complementary strands of siRNAs that contain inter sured in tissue culture using HEK293 cells. In the figure, nal repeats or palindromes may form internal fold-back struc duplexes are plotted Such that each X-axis tick-mark repre 15 tures. These hairpin-like structures exist in equilibrium with sents an individual siRNA, with each subsequent siRNA dif the duplexed form effectively reducing the concentration of fering in target position by two nucleotides for Human Cyclo functional duplexes. The propensity to form internal hairpins philin and Firefly Luciferase, and by one nucleotide for and their relative stability can be estimated by predicted melt Human DBI. Furthermore, the y-axis denotes the level of ing temperatures. HighTm reflects a tendency to form hairpin target expression remaining after transfection of the duplex structures. Lower Tm values indicate a lesser tendency to into cells and Subsequent silencing of the target. form hairpins. When the functional classes of siRNAs were SiRNA identified and optimized in this document work sorted by T. (FIG. 3c), the following trends were identified: equally well in a wide range of cell types. FIG.3a shows the duplexes lacking stable internal repeats were the most potent evaluation of thirty siRNAs targeting the DBI gene in three silencers (no F95 duplex with predicted hairpin structure cell lines derived from different tissues. Each DBI siRNA 25 T>60° C.). In contrast, about 60% of the duplexes in the displays very similar functionality in HEK293 (ATCC, CRL groups having internal hairpins with calculated T. Values 1573, human embryonic kidney), HeLa (ATCC, CCL-2, cer less than 20° C. were F80. Thus, the stability of internal vical epithelial adenocarcinoma) and DU145 (HTB-81, pros repeats is inversely proportional to the silencing effect and tate) cells as deterimined by the B-DNA assay. Thus, siRNA defines Criterion III (predicted hairpin structure Tes200C). functionality is determined by the primary sequence of the 30 Sequence-Based Determinants of SiRNA Functionality siRNA and not by the intracellular environment. Addition ally, it should be noted that although the present invention When the siRNA panel was sorted into functional and provides for a determination of the functionality of siRNA for non-functional groups, the frequency of a specific nucleotide a given target, the same siRNA may silence more than one at each position in a functional siRNA duplex was compared gene. For example, the complementary sequence of the 35 with that of a nonfunctional duplex in order to assess the silencing siRNA may be present in more than one gene. preference for or against a certain nucleotide. FIG. 4 shows Accordingly, in these circumstances, it may be desirable not the results of these queries and the Subsequent resorting of the to use the siRNA with highest SMARTscoreTM. In such cir data set (from FIG. 2). The data is separated into two sets: cumstances, it may be desirable to use the siRNA with the those duplexes that meet the criteria, a specific nucleotide in next highest SMARTscoreTM. 40 a certain position grouped on the left (Selected) and those To determine the relevance of G/C content in siRNA func that do not grouped on the right (Eliminated). The duplexes tion, the G/C content of each duplex in the panel was calcu are further sorted from most functional to least functional lated and the functional classes of siRNAs (57% GC and Table III show quantitative analysis for the following five content) compared to the 36%-52% group. The group with sequence-related properties of siRNA: (A) an A at position 19 extremely low GC content (26% or less) contained a higher 50 of the sense strand; (B) an A at position 3 of the sense strand; proportion of non-functional siRNAS and no highly-func (C) a Uat position 10 of the sense strand; (D) a base other than tional siRNAs. The G/C content range of 30%-52% was Gat position 13 of the sense strand; and (E) a base other than therefore selected as Criterion I for siRNA functionality, con Cat position 19 of the sense strand. sistent with the observation that a G/C range 30%-70% pro When the siRNAs in the panel were evaluated for the motes efficient RNAi targeting. Application of this criterion 55 presence of an A at position 19 of the sense strand, the per alone provided only a marginal increase in the probability of centage of non-functional duplexes decreased from 20% to selecting functional siRNAs from the panel: selection of F50 11.8%, and the percentage of F95 duplexes increased from and F95 siRNAs was improved by 3.6% and 2.2%, respec 21.7% to 29.4% (Table III). Thus, the presence of an A in this tively. The siRNA panel presented here permitted a more position defined Criterion IV. systematic analysis and quantification of the importance of 60 Another sequence-related property correlated with silenc this criterion than that used previously. ing was the presence of an A in position 3 of the sense Strand A relative measure of local internal stability is the A/U base (FIG. 4b). Of the siRNAs with A3, 34.4% were F95, com pair (bp) content; therefore, the frequency of A/U bp was pared with 21.7% randomly selected siRNAs. The presence determined for each of the five terminal positions of the of a U base in position 10 of the sense strand exhibited an even duplex (5' sense (S)/5' antisense (AS)) of all siRNAs in the 65 greater impact (FIG.4c). Of the duplexes in this group, 41.7% panel. Duplexes were then categorized by the number of A/U were F95. These properties became criteria V and VI, respec bp in positions 1-5 and 15-19 of the sense strand. The ther tively. US 7,820,809 B2 25 26 Two negative sequence-related criteria that were identified some functional siRNAs but included all non-functional siR also appear on FIG. 4. The absence of a Gat position 13 of the NAS. A graphical representation of this selection is shown in sense Strand, conferred a marginal increase in selecting func FIG.S. tional duplexes (FIG. 4d). Similarly, lack of a Cat position 19 The methods for obtaining the seven criteria embodied in of the sense strand also correlated with functionality (FIG. Table III are illustrative of the results of the process used to 4e). Thus, among functional duplexes, position 19 was most develop the information for Formulas VIII and IX. Thus simi likely occupied by A, and rarely occupied by C. These rules lar techniques were used to establish the other variables and their multipliers. As described above, basic statistical meth were defined as criteria VII and VIII, respectively. ods were use to determine the relative values for these mul Application of each criterion individually provided mar 10 tipliers. ginal but statistically significant increases in the probability To determine the value for “Improvement over Random' of selecting a potent siRNA. Although the results were infor the difference in the frequency of a given attribute (e.g. GC mative, the inventors sought to maximize potency and there content, base preference) at a particular position is deter fore consider multiple criteria or parameters. Optimization is mined between individual functional groups (e.g. F95 group contained a “U” at this 25 position 41.7% of the time. In contrast, the total group of 270 Improvement siRNAs had a “U” at this position 21.7% of the time, thus the Criterion % Functional over Random improvement over random is calculated to be 20% (or 41.7%- I. 30%-52% GC content 95% silencing; (ii) 80-95% silencing; and (iii) less than eF8O 75.0% 18.9% 50% silencing. Sequences exhibiting 51-84% silencing were eF95 29.4% 7.70% 40 V. An Abase at position 3 95% gene silencing, see FIG. 6a, Formula IX. Each siRNA was then assigned a score (referred >F95) have a low internal stability (AISP of position 19=~- to as a SMARTscoreTM) according to the values derived from 7.6 kcal/mol). In contrast low-efficiency siRNA (i.e. those the formulas. Duplexes that scored higher than 0 or 20, for exhibiting less than 50% silencing, 95% silencing group, siRNAs determining which end of the molecule enters the RISC com that exhibit poor silencing activity have weak AISP values 10 plex. In contrast, the high and low AISP values observed in (-7.6, -7.5, and -7.5 kcal/mol for positions 1, 2 and 3 respec the central regions of the molecule may be critical for siRNA tively). target mRNA interactions and product release, respectively. Overall the profiles of both strong and weak siRNAs form If the AISP values described above accurately define the distinct sinusoidal shapes that are roughly 180° out-of-phase thermodynamic parameters of strong siRNA, it would be with each other. While these thermodynamic descriptions 15 expected that similar patterns would be observed in strong define the archetypal profile of a strong siRNA, it will likely siRNA isolated from nature. Natural siRNAs existina harsh, be the case that neither the AG values given for key positions RNase-rich environment and it can be hypothesized that only in the profile or the absolute position of the profile along the those siRNA that exhibit heightened affinity for RISC (i.e. Y-axis (i.e. the AG-axis) are absolutes. Profiles that are shifted siRNA that exhibit an average internal stability profile similar upward or downward (i.e. having on an average, higher or to those observed in strong siRNA) would survive in an intra lower values at every position) but retain the relative shape cellular environment. This hypothesis was tested using GFP and position of the profile along the X-axis can beforeseen as specific siRNA isolated from N. benthamiana. Llave et al. being equally effective as the model profile described here. (2002) Endogenous and Silencing-Associated Small RNAs in Moreover, it is likely that siRNA that have strong or even Plants, The Plant Cell 14, 1605-1619, introduced long stronger gene-specific silencing effects might have exagger 25 double-stranded GFP-encoding RNA into plants and subse ated AG values (either higher or lower) at key positions. Thus, quently re-isolated GFP-specific siRNA from the tissues. The for instance, it is possible that the 5'-most position of the sense AISP of fifty-nine of these GFP-siRNA were determined, strand (position 19) could have AG values of 7.4 kcal/mol or averaged, and Subsequently plotted alongside the AISP pro lower and still be a strong siRNA if, for instance, a G-C->G- file obtained from the cyclophilin B/DBI/luciferase siRNA T/U mismatch were substituted at position 19 and altered 30 having >90% silencing properties (FIG. 6b). Comparison of duplex stability. Similarly, position 12 and position 7 could the two groups show that profiles are nearly identical. This have values above 8.3 kcal/mol and below 7.7 kcal/mole, finding validates the information provided by the internal respectively, without abating the silencing effectiveness of stability profiles and demonstrates that: (1) the profile iden the molecule. Thus, for instance, at position 12, a stabilizing tified by analysis of the cyclophilin B/DBI/luciferase siRNAs chemical modification (e.g. a chemical modification of the 2' 35 are not gene specific; and (2) AISP values can be used to position of the Sugar backbone) could be added that increases search for strong siRNAS in a variety of species. the average internal stability at that position. Similarly, at Both chemical modifications and base-pair mismatches position 7, mismatches similar to those described previously can be incorporated into siRNA to alter the duplex's AISP and could be introduced that would lower the AG values at that functionality. For instance, introduction of mismatches at position. 40 positions 1 or 2 of the sense strand destabilized the 5'end of Lastly, it is important to note that while functional and the sense Strand and increases the functionality of the mol non-functional siRNA were originally defined as those mol ecule (see Luc, FIG. 7). Similarly, addition of 2'-O-methyl ecules having specific silencing properties, both broader or groups to positions 1 and 2 of the sense Strand can also alter more limiting parameters can be used to define these mol the AISP and (as a result) increase both the functionality of ecules. As used herein, unless otherwise specified, “non 45 the molecule and eliminate off-target effects that results from functional siRNA are defined as those siRNA that induce sense strand homology with the unrelated targets (FIGS. 8a, less than 50% (<50%) target silencing, “semi-functional 8b). siRNA’ induce 50-79% target silencing, “functional siRNA'. are molecules that induce 80-95% gene silencing, and Rationale for Criteria in a Biological Context “highly-functional siRNA are molecules that induce great 50 The fate of siRNA in the RNAi pathway may be described than 95% gene silencing. These definitions are not intended to in 5 major steps: (1) duplex recognition and pre-RISC com be rigid and can vary depending upon the design and needs of plex formation; (2) ATP-dependent duplex unwinding/strand the application. For instance, it is possible that a researcher selection and RISC activation; (3) mRNA target identifica attempting to map a gene to a chromosome using a functional tion; (4) mRNA cleavage, and (5) product release (FIG. 1). assay, may identify an siRNA that reduces gene activity by 55 Given the level of nucleic acid-protein interactions at each only 30%. While this level of gene silencing may be “non step, siRNA functionality is likely influenced by specific functional for e.g. therapeutic needs, it is sufficient for gene biophysical and molecular properties that promote efficient mapping purposes and is, under these uses and conditions, interactions within the context of the multi-component com “functional.” For these reasons, functional siRNA can be plexes. Indeed, the systematic analysis of the siRNA test set defined as those molecules having greater than 10%. 20%, 60 identified multiple factors that correlate well with function 30%, 40%, 50%, 60%, 70%, 80%, or 90% silencing capabili ality. When combined into a single algorithm, they proved to ties at 100 nM transfection conditions. Similarly, depending be very effective in selecting active siRNAs. upon the needs of the study and/or application, non-func The factors described here may also be predictive of key tional and semi-functional siRNA can be defined as having functional associations important for each step in RNAi. For different parameters. For instance, semi-functional siRNA 65 example, the potential formation of internal hairpin structures can be defined as being those molecules that induce 20%, correlated negatively with siRNA functionality. Complemen 30%, 40%, 50%, 60%, or 70% silencing at 100 nM transfec tary strands with stable internal repeats are more likely to US 7,820,809 B2 29 30 exist as stable hairpins thus decreasing the effective concen will be 19 base pairs in length, and one strand of each of the tration of the functional duplex form. This Suggests that the siRNAs will be 100% complementary to a portion of the duplex is the preferred conformation for initial pre-RISC target mRNA. association. Indeed, although single complementary Strands By increasing the number of siRNAs directed to a particu can induce gene silencing, the effective concentration 5 lar target using a pool or kit, one is able both to increase the required is at least two orders of magnitude higher than that of likelihood that at least one siRNA with satisfactory function the duplex form. ality will be included, as well as to benefit from additive or synergistic effects. Further, when two or more siRNAs siRNA-pre-RISC complex formation is followed by an directed against a single gene do not have satisfactory levels ATP-dependent duplex unwinding step and “activation' of 10 of functionality alone, if combined, they may satisfactorily the RISC. The siRNA functionality was shown to correlate promote degradation of the target messenger RNA and Suc with overall low internal stability of the duplex and low cessfully inhibit translation. By including multiple siRNAs in internal stability of the 3' sense end (or differential internal the system, not only is the probability of silencing increased, stability of the 3' sense compare to the 5' sense strand), which but the economics of operation are also improved when com may reflect strand selection and entry into the RISC. Overall 15 pared to adding different siRNAs sequentially. This effect is duplex stability and low internal stability at the 3' end of the contrary to the conventional wisdom that the concurrent use sense strand were also correlated with siRNA functionality. of multiple siRNA will negatively impact gene silencing (e.g. Interestingly, siRNAs with very high and very low overall Holen, T. et al. (2003) “Similar behavior of single strand and stability profiles correlate strongly with non-functional double strand siRNAS Suggests they act through a common duplexes. One interpretation is that high internal stability RNAi pathway.” NAR31: 2401-21407). prevents efficient unwinding while very low stability reduces In fact, when two siRNAs were pooled together, 54% of the siRNA target affinity and subsequent mRNA cleavage by the pools of two siRNAs induced more than 95% gene silencing. RISC Thus, a 2.5-fold increase in the percentage of functionality Several criteria describe base preferences at specific posi was achieved by randomly combining two siRNAs. Further, 25 over 84% of pools containing two siRNAs induced more than tions of the sense Strand and are even more intriguing when 80% gene silencing. considering their potential mechanistic roles in target recog More preferably, the kit is comprised of at least three siR nition and mRNA cleavage. Base preferences for A at posi NAS, wherein one strand of each siRNA comprises a tion 19 of the sense strand but not C. are particularly interest sequence that is Substantially similar to a sequence of the ing because they reflect the same base preferences observed 30 target mRNA and the other Strand comprises a sequence that for naturally occurring miRNA precursors. That is, among the is Substantially complementary to the region of the target reported miRNA precursor sequences 75% contain a U at mRNA. As with the kit that comprises at least two siRNAs, position 1 which corresponds to an A in position 19 of the more preferably one strand will comprise a sequence that is sense strand of siRNAs, while G was under-represented in identical to a sequence that is contained in the mRNA and this same position for miRNA precursors. These observations 35 another strand that is 100% complementary to a sequence that support the hypothesis that both miRNA precursors and is contained in the mRNA. During experiments, when three siRNA duplexes are processed by very similar if not identical siRNAs were combined together, 60% of the pools induced protein machinery. The functional interpretation of the pre more than 95% gene silencing and 92% of the pools induced dominance of a U/A base pair is that it promotes flexibility at more than 80% gene silencing. the 5'antisense ends of both siRNA duplexes and miRNA 40 Further, even more preferably, the kit is comprised of at precursors and facilitates efficient unwinding and selective least four siRNAs, wherein one strand of each siRNA com strand entrance into an activated RISC. prises a sequence that is Substantially similar to a region of the Among the criteria associated with base preferences that sequence of the target mRNA, and the other strand comprises are likely to influence mRNA cleavage or possibly product a sequence that is Substantially complementary to the region release, the preference for Uat position 10 of the sense strand 45 of the target mRNA. As with the kit or pool that comprises at exhibited the greatest impact, enhancing the probability of least two siRNAs, more preferably one strand of each of the selecting an F80 sequence by 13.3%. Activated RISC prefer siRNA duplexes will comprise a sequence that is identical to entially cleaves target mRNA between nucleotides 10 and 11 a sequence that is contained in the mRNA, and another strand relative to the 5' end of the complementary targeting strand. that is 100% complementary to a sequence that is contained in Therefore, it may be that U, the preferred base for most 50 the mRNA. endoribonucleases, at this position Supports more efficient Additionally, kits and pools with at least five, at least six, cleavage. Alternatively, a U/A bp between the targeting and at least seven siRNAs may also be useful with the present siRNA strand and its cognate target mRNA may create an invention. For example, pools of five siRNA induced 95% optimal conformation for the RISC-associated "slicing gene silencing with 77% probability and 80% silencing with activity. 55 98.8% probability. Thus, pooling of siRNAs together can result in the creation of a target-specific silencing reagent Pooling with almost a 99% probability of being functional. The fact According to another embodiment, the present invention that Such high levels of Success are achievable using Such provides a pool of at least two siRNAs, preferably in the form pools of siRNA, enables one to dispense with costly and of a kit or therapeutic reagent, wherein one strand of each of 60 time-consuming target-specific validation procedures. the siRNAS, the sense Strand comprises a sequence that is For this embodiment, as well as the other aforementioned substantially similar to a sequence within a target mRNA. The embodiments, each of the siRNAs within a pool will prefer opposite strand, the antisense Strand, will preferably com ably comprise between 18 and 30 base pairs, more preferably prise a sequence that is Substantially complementary to that of between 18 and 25 base pairs, and most preferably 19 base the target mRNA. More preferably, one strand of each siRNA 65 pairs. Within each siRNA, preferably at least 18 contiguous will comprise a sequence that is identical to a sequence that is bases of the antisense strand will be 100% complementary to contained in the target mRNA. Most preferably, each siRNA the target mRNA. More preferably, at least 19 contiguous US 7,820,809 B2 31 32 bases of the antisense strand will be 100% complementary to may be able to cause silencing by acting in the nucleus. the target mRNA. Additionally, there may be overhangs on Examples of these types of modifications are described in either the sense Strand or the antisense strand, and these commonly assigned U.S. patent application Ser. Nos. 10/431, overhangs may be at either the 5' end or the 3' end of either of 027 and 10/613,077, each of which is incorporated by refer the strands, for example there may be one or more overhangs 5 ence herein. of 1-6 bases. When overhangs are present, they are not As described above, even when one selects at least two included in the calculation of the number of base pairs. The siRNAs at random, the effectiveness of the two may be two nucleotide 3' overhangs mimic natural siRNAs and are greater than one would predict based on the effectiveness of commonly used but are not essential. Preferably, the over two individual siRNAs. This additive or synergistic effect is hangs should consist of two nucleotides, most often dTdT or 10 particularly noticeable as one increases to at least three siR UU at the 3' end of the sense and antisense strand that are not NAS, and even more noticeable as one moves to at least four complementary to the target sequence. The siRNAS may be siRNAS. Surprisingly, the pooling of the non-functional and produced by any method that is now known or that comes to semi-functional siRNAs, particularly more than five siRNAs, be known for synthesizing double stranded RNA that one can lead to a silencing mixture that is as effective if not more skilled in the art would appreciate would be useful in the 15 effective than any one particular functional siRNA. present invention. Preferably, the siRNAs will be produced by Within the kit of the present invention, preferably each Dharmacon’s proprietary ACER) technology. However, other siRNA will be present in a concentration of between 0.001 methods for synthesizing siRNAs are well known to persons and 200 uM, more preferably between 0.01 and 200 nM, and skilled in the art and include, but are not limited to, any most preferably between 0.1 and 10 nM. chemical synthesis of RNA oligonucleotides, ligation of In addition to preferably comprising at least four or five shorter oligonucleotides, in vitro transcription of RNA oligo siRNAs, the kit of the present invention will also preferably nucleotides, the use of vectors for expression within cells, comprise a buffer to keep the siRNA duplex stable. Persons recombinant Dicer products and PCR products. skilled in the art are aware of buffers suitable for keeping The siRNA duplexes within the aforementioned pools of siRNA stable. For example, the buffer may be comprised of siRNAS may correspond to overlapping sequences within a 25 100 mM KC1, 30 mM HEPES-pH 7.5, and 1 mM MgC1. particular mRNA, or non-overlapping sequences of the Alternatively, kits might contain complementary strands that mRNA. However, preferably they correspond to non-overlap contain any one of a number of chemical modifications (e.g. ping sequences. Further, each siRNA may be selected ran a 2'-O-ACE) that protect the agents from degradation by domly, or one or more of the siRNA may be selected accord nucleases. In this instance, the user may (or may not) remove ing to the criteria discussed above for maximizing the 30 the modifying protective group (e.g. deprotect) before effectiveness of siRNA. annealing the two complementary strands together. Included in the definition of siRNAs are siRNAs that con By way of example, the kit may be organized such that tain substituted and/or labeled nucleotides that may, for pools of siRNA duplexes are provided on an array or microar example, be labeled by radioactivity, fluorescence or mass. ray of wells or drops for a particular gene set or for unrelated The most common substitutions are at the 2' position of the 35 genes. The array may, for example, be in 96 wells, 384 wells ribose Sugar, where moieties such as H (hydrogen) F. NH or 1284 wells arrayed in a plastic plate or on a glass slide using OCH and other O-alkyl, alkenyl, alkynyl, and orthoesters, techniques now known or that come to be known to persons may be substituted, or in the phosphorous backbone, where skilled in the art. Within an array, preferably there will be sulfur, amines or hydrocarbons may be substituted for the controls such as functional anti-lamin A/C, cyclophilin and bridging of non-bridging atoms in the phosphodiester bond. 40 two siRNA duplexes that are not specific to the gene of Examples of modified siRNAs are explained more fully in interest. commonly assigned U.S. patent application Ser. No. 10/613, In order to ensure stability of the siRNA pools prior to 077, filed Jul. 1, 2003, which is incorporated by reference usage, they may be retained in lyophilized form at minus herein. twenty degrees (-20°C.) until they are ready for use. Prior to Additionally, as noted above, the cell type into which the 45 usage, they should be resuspended; however, even once resus siRNA is introduced may affect the ability of the siRNA to pended, for example, in the aforementioned buffer, they enter the cell; however, it does not appear to affect the ability should be kept at minus twenty degrees, (-20°C.) until used. of the siRNA to function once it enters the cell. Methods for The aforementioned buffer, prior to use, may be stored at introducing double-stranded RNA into various cell types are approximately 4°C. or room temperature. Effective tempera well known to persons skilled in the art. 50 tures at which to conduct transfections are well known to As persons skilled in the art are aware, in certain species, persons skilled in the art and include for example, room the presence of proteins such as RdRP, the RNA-dependent temperature. RNA polymerase, may catalytically enhance the activity of The kit may be applied either in vivo or in vitro. Preferably, the siRNA. For example, RdRP propagates the RNAi effect in the siRNA of the pools or kits is applied to a cell through C. elegans and other non-mammalian organisms. In fact, in 55 transfection, employing standard transfection protocols. organisms that contain these proteins, the siRNA may be These methods are well known to persons skilled in the art inherited. Two other proteins that are well studied and known and include the use of lipid-based carriers, electroporation, to be a part of the machinery are members of the Argonaute cationic carriers, and microinjection. Further, one could family and Dicer, as well as their homologues. There is also apply the present invention by synthesizing equivalent DNA initial evidence that the RISC complex might be associated 60 sequences (either as two separate, complementary strands, or with the ribosome so the more efficiently translated mRNAs as hairpin molecules) instead of siRNA sequences and intro will be more susceptible to silencing than others. ducing them into cells through vectors. Once in the cells, the Another very important factor in the efficacy of siRNA is cloned DNA could be transcribed, thereby forcing the cells to mRNA localization. In general, only cytoplasmic mRNAs are generate the siRNA. Examples of vectors suitable for use with considered to be accessible to RNAi to any appreciable 65 the present application include but are not limited to the degree. However, appropriately designed siRNAs, for standard transient expression vectors, adenoviruses, retrovi example, siRNAs modified with internucleotide linkages, ruses, lentivirus-based vectors, as well as other traditional US 7,820,809 B2 33 34 expression vectors. Any vector that has an adequate siRNA one (or more) cell types in increasingly diminishing concen expression and procession module may be used. Further trations (e.g. 3.0->0.3 nM). Subsequently, the level of gene more, certain chemical modifications to siRNAS, including silencing induced by each concentration is examined and but not limited to conjugations to other molecules, may be siRNA that exhibit hyperfunctional potency (i.e. those that used to facilitate delivery. For certain applications it may be 5 induce 80% silencing or greater at e.g. picomolar concentra preferable to deliver molecules without transfection by sim tions) are identified. In a second study, the longevity profiles ply formulating in a physiological acceptable solution. of siRNA having high (>-10) SMARTscoresTM and greater This embodiment may be used in connection with any of than 80% silencing are examined. In one non-limiting the aforementioned embodiments. Accordingly, the example of how this is achieved, siRNA are introduced into a sequences within any pool may be selected by rational design. 10 test cell line and the levels of RNAi are measured over an Multigene Silencing extended period of time (e.g. 24-168 hrs). SiRNAs that In addition to developing kits that contain multiple siRNA exhibit strong RNA interference patterns (i.e. >80% interfer directed against a single gene, another embodiment includes ence) for periods of time greater than, e.g., 120 hours are thus the use of multiple siRNA targeting multiple genes. Multiple identified. Studies similar to those described above can be genes may be targeted through the use of high- or hyper 15 performed on any and all of the >10 siRNA included in this functional siRNA. High- or hyper-functional siRNA that document to further define the most functional molecule for exhibit increased potency, require lower concentrations to any given gene. Molecules possessing one or both properties induce desired phenotypic (and thus therapeutic) effects. This (extended longevity and heightened potency) are labeled circumvents RISC Saturation. It therefore reasons that if “hyperfunctional siRNA.” and earmarked as candidates for lower concentrations of a single siRNA are needed for knock future therapeutic studies. out or knockdown expression of one gene, then the remaining While the example(s) given above describe one means by (uncomplexed) RISC will be free and available to interact which hyperfunctional siRNA can be isolated, neither the with siRNA directed against two, three, four, or more, genes. assays themselves nor the selection parameters used are rigid Thus in this embodiment, the authors describe the use of and can vary with each family of siRNA. Families of siRNA highly functional or hyper-functional siRNA to knock out 25 include siRNAS directed against a single gene, or directed three separate genes. More preferably, Such reagents could be against a related family of genes. combined to knockout four distinct genes. Even more prefer The highest quality siRNA achievable for any given gene ably, highly functional or hyperfunctional siRNA could be may vary considerably. Thus, for example, in the case of one used to knockout five distinct genes. Most preferably, siRNA gene (gene X), rigorous studies such as those described above 30 may enable the identification of an siRNA that, at picomolar of this type could be used to knockout or knockdown the concentrations, induces 99--% silencing for a period of 10 expression of six or more genes. days. Yet identical studies of a second gene (gene Y) may Hyperfunctional siRNA yield an siRNA that at high nanomolar concentrations (e.g. The term hyperfunctional siRNA (hf-siRNA) describes a 100 nM) induces only 75% silencing for a period of 2 days. subset of the siRNA population that induces RNAi in cells at 35 Both molecules represent the very optimum siRNA for their low- or Sub-nanomolar concentrations for extended periods respective gene targets and therefore are designated "hyper of time. These traits, heightened potency and extended lon functional.” Yet due to a variety of factors including but not gevity of the RNAi phenotype, are highly attractive from a limited to target concentration, siRNA stability, cell type, therapeutic standpoint. Agents having higher potency require off-target interference, and others, equivalent levels of lesser amounts of the molecule to achieve the desired physi 40 potency and longevity are not achievable. Thus, for these ological response, thus reducing the probability of side reasons, the parameters described in the before mentioned effects due to “off-target interference. In addition to the assays, can vary. While the initial screen selected siRNA that potential therapeutic benefits associated with hyperfunc had SMARTscoresTM above -10 and a gene silencing capa tional siRNA, hf-siRNA are also desirable from an economic bility of greater than 80%, selections that have stronger (or position. Hyperfunctional siRNA may cost less on a per 45 weaker) parameters can be implemented. Similarly, in the treatment basis, thus reducing the overall expenditures to Subsequent studies designed to identify molecules with high both the manufacturer and the consumer. potency and longevity, the desired cutoff criteria (i.e. the Identification of hyperfunctional siRNA involves multiple lowest concentration that induces a desirable level of inter steps that are designed to examine an individual siRNA ference, or the longest period of time that interference can be agent's concentration- and/or longevity-profiles. In one non 50 observed) can vary. The experimentation Subsequent to appli limiting example, a population of siRNA directed against a cation of the rational criteria of this application is signifi single gene are first analyzed using the previously described cantly reduced where one is trying to obtain a suitable hyper algorithm (Formula VIII). Individual siRNA are then intro functional siRNA for, for example, therapeutic use. When, for duced into a test cell line and assessed for the ability to example, the additional experimentation of the type described degrade the target mRNA. It is important to note that when 55 herein is applied by one skilled in the art with this disclosure performing this step it is not necessary to testall of the siRNA. in hand, a hyperfunctional siRNA is readily identified. Instead, it is sufficient to test only those siRNA having the The siRNA may be introduced into a cell by any method highest SMARTscoresTM (i.e. SMARTscoreTM>-10). Subse that is now known or that comes to be known and that from quently, the gene silencing data is plotted against the reading this disclosure, persons skilled in the art would deter SMARTscoresTM (see FIG.9). SiRNA that (I) induce a high 60 mine would be useful in connection with the present invention degree of gene silencing (i.e. they induce greater than 80% in enabling siRNA to cross the cellular membrane. These gene knockdown) and (2) have superior SMARTscoresTM methods include, but are not limited to, any manner of trans (i.e. a SMARTscoreTM of >-10, suggesting a desirable aver fection, Such as for example transfection employing DEAE age internal stability profile) are selected for further investi Dextran, calcium phosphate, cationic lipids/liposomes, gations designed to better understand the molecule's potency 65 micelles, manipulation of pressure, microinjection, elec and longevity. In one, non-limiting study dedicated to under troporation, immunoporation, use of vectors such as viruses, standing a molecule's potency, an siRNA is introduced into plasmids, cosmids, bacteriophages, cell fusions, and cou US 7,820,809 B2 35 36 pling of the polynucleotides to specific conjugates or ligands Example I Such as antibodies, antigens, or receptors, passive introduc tion, adding moieties to the siRNA that facilitate its uptake, Sequences Used to Develop the Algorithm and the like. Having described the invention with a degree of particu Anti-Firefly and anti-Cyclophilin siRNAs panels (FIG. larity, examples will now be provided. These examples are 5A, B) sorted according to using Formula VIII predicted not intended to and should not be construed to limit the scope values. All siRNAs scoring more than 0 (formula VIII) and of the claims in any way. more then 20 (formula IX) are fully functional. All ninety 10 sequences for each gene (and DBI) appear below in Table IV. EXAMPLES TABLE IW

Cyclo 1 SEQ. D OO32 GUUCCAAAAACAGUGGAUA. General Techniques and Nomenclatures Cyclo 2 SEQ. D OO33 UCCAAAAACAGUGGAUAAU 15 Cyclo 3 SEQ. D OO34 CAAAAACAGUGGAUAAUUU siRNA nomenclature. All siRNA duplexes are referred to Cyclo 4 SEQ. D OO35 AAAACAGUGGAUAAUUUUG. by sense strand. The first nucleotide of the 5'-end of the sense Cyclo 5 SEQ. D OO36 AACAGUGGAUAAUUUUGUG strand is position 1, which corresponds to position 19 of the Cyclo 6 SEQ. D OO37 CAGUGGAUAAUUUUGUGGC Cyclo 7 SEO. D OO38 GUGGAUAAUUUUGUGGCCU antisense Strand for a 19-mer. In most cases, to compare Cyclo 8 SEQ. D OO39 GGAUAAUUUUGUGGCCUUA results from different experiments, silencing was determined Cyclo 9 SEQ. D OO4 O AUAAUUUUGUGGCCUUAGC by measuring specific transcript mRNA levels or enzymatic Cyclo 10 SEQ. D OO41 AAUUUUGUGGCCUUAGCUA. Cyclo 11 SEQ. D OO42 UUUUGUGGCCUUAGCUACA activity associated with specific transcript levels, 24 hours Cyclo 12 SEQ. D OO43 UUGUGGCCUUAGCUACAGG post-transfection, with siRNA concentrations held constant at Cyclo 13 SEQ. D OO44 GUGGCCUUAGCUACAGGAG Cyclo 14 SEQ. D OO45 GGCCUUAGCUACAGGAGAG 100 nM. For all experiments, unelss otherwise specified Cyclo 15 SEQ. D OO46 CCUUAGCUACAGGAGAGAA transfection efficiency was ensured to be over 95%, and no 25 Cyclo 16 SEQ. D OO47 UUAGCUACAGGAGAGAAAG detectable cellular toxicity was observed. The following sys Cyclo 17 SEQ. D OO48 AGCUACAGGAGAGAAAGGA Cyclo 18 SEQ. D OO49 CUACAGGAGAGAAAGGAUU tem of nomenclature was used to compare and report siRNA Cyclo 19 SEQ. D OOSO ACAGGAGAGAAAGGAUUUG. silencing functionality: “F” followed by the degree of mini Cyclo 2O SEQ. D OO51 AGGAGAGAAAGGAUUUGGC mal knockdown. For example, F50 signifies at least 50% Cyclo 21 SEQ. D OO52 GAGAGAAAGGAUUUGGCUA. 30 Cyclo 22 SEQ. D OO53 GAGAAAGGAUUUGGCUACA knockdown, F80 means at least 80%, and so forth. For this Cyclo 23 SEQ. D OOS 4 GAAAGGAUUUGGCUACAAA. study, all sub-F50 siRNAs were considered non-functional. Cyclo 24 SEQ. D OO55 AAGGAUUUGGCUACAAAAA Cyclo 25 SEQ. D OOS 6 GGAUUUGGCUACAAAAACA Cell culture and transfection. 96-well plates are coated Cyclo 26 SEQ. D OO57 AUUUGGCUACAAAAACAGC with 50 ill of 50 mg/ml poly-L-lysine (Sigma) for 1 hr., and Cyclo 27 SEQ. D OO58 UUGGCUACAAAAACAGCAA 35 Cyclo 28 SEQ. D OO59 GGCUACAAAAACAGCAAAU then washed 3x with distilled water before being dried for 20 Cyclo 29 SEQ. D OO6 O CUACAAAAACAGCAAAUUC min. HEK293 cells or HEK293Lucs or any other cell type of Cyclo 3 O SEQ. D OO61 ACAAAAACAGCAAAUUCCA interest are released from their solid support by trypsiniza Cyclo 31 SEQ. D OO62 AAAAACAGCAAAUUCCAUC Cyclo 32 SEQ. D OO63 AAACAGCAAAUUCCAUCGU tion, diluted to 3.5x10 cells/ml, followed by the addition of Cyclo 33 SEQ. D OO64 ACAGCAAAUUCCAUCGUGU Cyclo 34 SEQ. D OO65 AGCAAAUUCCAUCGUGUAA 100 uL of cells/well. Plates are then incubated overnight at 40 37° C., 5% CO. Transfection procedures can vary widely Cyclo 35 SEQ. D OO66 CAAAUUCCAUCGUGUAAUC Cyclo 36 SEQ. D OO67 AAUUCCAUCGUGUAAUCAA. depending on the cell type and transfection reagents. In one Cyclo 37 SEQ. D OO68 UUCCAUCGUGUAAUCAAGG non-limiting example, a transfection mixture consisting of 2 Cyclo 38 SEQ. D OO69 CCAUCGUGUAAUCAAGGAC mL Opti-MEM I (Gibco-BRL), 80 ul Lipofectamine 2000 Cyclo 39 SEQ. D OOf O AUCGUGUAAUCAAGGACUU Cyclo 4 O SEQ. D OO71 CGUGUAAUCAAGGACUUCA (Invitrogen), 15 uL SUPERNasin at 20 U/ul (Ambion), and 45 Cyclo 41 SEQ. D OO72 UGUAAUCAAGGACUUCAUG 1.5 ul of reporter gene plasmid at 1 g/ul is prepared in 5-ml Cyclo 42 SEQ. D OO73 UAAUCAAGGACUUCAUGAU polystyrene round bottom tubes. 100 ul of transfection Cyclo 43 SEQ. D OO74 AUCAAGGACUUCAUGAUCC Cyclo 44 SEQ. D OO75 CAAGGACUUCAUGAUCCAG reagent is then combined with 100 ul of siRNAs in polysty Cyclo 45 SEQ. D OO76 AGGACUUCAUGAUCCAGGG rene deep-well titer plates (Beckman) and incubated for 20 to Cyclo 46 SEQ. D OOff GACUUCAUGAUCCAGGGCG 30 min at room temp. 550 ul of Opti-MEM is then added to 50 Cyclo 47 SEQ. D OO78 CUUCAUGAUCCAGGGCGGA Cyclo 48 SEQ. D OO79 UCAUGAUCCAGGGCGGAGA each well to bring the final siRNA concentration to 100 nM. Cyclo 49 SEQ. D OO8 O AUGAUCCAGGGCGGAGACU Plates are then sealed with parafilm and mixed. Media is Cyclo 50 SEQ. D OO81 GAUCCAGGGCGGAGACUUC removed from HEK293 cells and replaced with 95 ul of Cyclo 51 SEQ. D OO82 UCCAGGGCGGAGACUUCAC transfection mixture. Cells are incubated overnight at 37°C., Cyclo 52 SEQ. D OO83 CAGGGCGGAGACUUCACCA. 55 Cyclo 53 SEQ. D OO84 GGGCGGAGACUUCACCAGG 5% CO. Cyclo 54 SEQ. D OO85 GCGGAGACUUCACCAGGGG Quantification of gene knockdown. A variety of quantifi Cyclo 55 SEO. D OO86 GGAGACUUCACCAGGGGAG Cyclo 56 SEQ. D OO87 AGACUUCACCAGGGGAGAU cation procedures can be used to measure the level of silenc Cyclo 57 SEO. D OO88 ACUUCACCAGGGGAGAUGG ing induced by siRNA or siRNA pools. In one non-limiting Cyclo 58 SEQ. D OO89 UUCACCAGGGGAGAUGGCA Cyclo 59 SEQ. D OO9 O CACCAGGGGAGAUGGCACA example: to measure mRNA levels 24 hrs post-transfection, 60 Cyclo 6 O SEQ. D OO91 CCAGGGGAGAUGGCACAGG QuantiGene branched-DNA (bDNA) kits (Bayer) (Wang, et Cyclo 61 SEQ. D OO92 AGGGGAGAUGGCACAGGAG al, Regulation of insulin preRNA splicing by glucose. Proc Cyclo 62 SEQ. D OO93 GGGAGAUGGCACAGGAGGA Natl Acad Sci 1997, 94:4360.) are used according to manu Cyclo 63 SEQ. D 0431 GAGAUGGCACAGGAGGAAA Cyclo 64 SEQ. D OO95 GAUGGCACAGGAGGAAAGA facturer instructions. To measure luciferase activity, media is Cyclo 65 SEQ. D OO94 UGGCACAGGAGGAAAGAGC removed from HEK293 cells 24hrs post-transfection, and 50 65 Cyclo 66 SEQ. D OO96 GCACAGGAGGAAAGAGCAU ul of Steady-GLO reagent (Promega) is added. After 5 min, Cyclo 67 SEQ. D OO97 ACAGGAGGAAAGAGCAUCU plates are analyzed on a plate reader.

US 7,820,809 B2 39 40 evaluated by B-DNA as described above and normalized to TABLE IV - continued the non-specific control. FIG. 10 shows that the siRNAs selected by the algorithm disclosed herein were significantly Lull 46 SEQ. ID O256 ACCGCGAAAAAGUUGCGCG Lull 47 SEO. ID O257 CAACCGCGAAAAAGUUGCG more potent than randomly selected siRNAs. The algorithm Lull 48 SEQ. ID O258 AACAACCGCGAAAAAGUUG. increased the chances of identifying an F50 siRNA from 48% Lull 49 SEQ. ID O259 GUAACAACCGCGAAAAAGU to 91%, and an F80 siRNA from 13% to 57%. In addition, Lull 50 SEQ. ID O26 O AAGUAACAACCGCGAAAAA pools of SMART siRNA silence the selected target better than Lull 51 SEQ. ID O261 UCAAGUAACAACCGCGAAA Lull 52 SEQ. ID O262 AGUCAAGUAACAACCGCGA randomly selected pools (see FIG. 10F). Lull 53 SEQ. ID O263 CCAGUCAAGUAACAACCGC Lull 54 SEQ. ID O264 CGCCAGUCAAGUAACAACC 10 Example III Lull 55 SEO. ID O265 GUCGCCAGUCAAGUAACAA Lull 56 SEQ. ID O266 ACGUCGCCAGUCAAGUAAC Lull 57 SEO. ID O267 UUACGUCGCCAGUCAAGUA Validation of the Algorithm Using Genes Involved in Lull 58 SEQ. ID O268 GAUUACGUCGCCAGUCAAG Clathrin-Dependent Endocytosis Lull 59 SEQ. ID O269 UGGAUUACGUCGCCAGUCA Lull 6 O SEQ. ID O270 CGUGGAUUACGUCGCCAGU 15 Components of clathrin-mediated endocytosis pathway are Lull 61 SEQ. ID O271 AUCGUGGAUUACGUCGCCA key to modulating intracellular signaling and play important Lull 62 SEQ. ID O272 AGAUCGUGGAUUACGUCGC Lull 63 SEQ. ID O273 AGAGAUCGUGGAUUACGUC roles in disease. Chromosomal rearrangements that result in Lull 64 SEQ. ID O274 AAAGAGAUCGUGGAUUACG fusion transcripts between the Mixed-Lineage Leukemia Lull 65 SEO. ID O275 AAAAAGAGAUCGUGGAUUA. gene (MLL) and CALM (Clathrin assembly lymphoid Lull 66 SEQ. ID O276 GGAAAAAGAGAUCGUGGAU Lull 67 SEO. ID O277 ACGGAAAAAGAGAUCGUGG myeloid leukemia gene) are believed to play a role in leuke Lull 68 SEQ. ID O278 UGACGGAAAAAGAGAUCGU mogenesis. Similarly, disruptions in Rab7 and Rab9, as well Lull 69 SEO. ID O279 GAUGACGGAAAAAGAGAUC as HIP1 (Huntingtin-interacting protein), genes that are Lull 7 O SEQ. ID O28O ACGAUGACGGAAAAAGAGA believed to be involved in endocytosis, are potentially respon Lull 71 SEQ. ID O281 AGACGAUGACGGAAAAAGA Lull 72 SEQ. ID O282 AAAGACGAUGACGGAAAAA sible for ailments resulting in lipid storage, and neuronal Lull 73 SEQ. ID O283 GGAAAGACGAUGACGGAAA 25 diseases, respectively. For these reasons, siRNA directed Lull 74 SEQ. ID O284 ACGGAAAGACGAUGACGGA against clathrin and other genes involved in the clathrin Lull 75 SEO. ID O285 GCACGGAAAGACGAUGACG mediated endocytotic pathway are potentially important Lull 76 SEQ. ID O286 GAGCACGGAAAGACGAUGA Lull 77 SEO. ID O287 UGGAGCACGGAAAGACGAU research and therapeutic tools. Lull 78 SEQ. ID 0288 UUUGGAGCACGGAAAGACG siRNAs directed against genes involved in the clathrin Lull 79 SEO. ID O289 GUUUUGGAGCACGGAAAGA 30 mediated endocytosis pathways were selected using Formula Lull 8 O SEQ. ID O290 UUGUUUUGGAGGACGGAAA VIII. The targeted genes were clathrin heavy chain (CHC, Lull 81 SEQ. ID O291 UGUUGUUUUGGAGCACGGA Lull 82 SEQ. ID O292 GUUGUUGUUUUGGAGCACG accession # NM 004859), clathrin light chain A (CLCa, Lull 83 SEQ. ID O293 CCGUUGUUGUUUUGGAGCA NM 001833), clathrin light chain B (CLCb, NM 001834), Lull 84 SEQ. ID O294 CGCCGUUGUUGUUUUGGAG CALM (U45976), B2 subunit of AP-2 (B2, NM 001282), Lull 85 SEO. ID O295 GCCGCCGUUGUUGUUUUGG 35 Eps15 (NM 001981), Eps15R (NM 021235), dynamin II Lull 86 SEQ. ID O296 CCGCCGCCGUUGUUGUUUU Lull 87 SEO. ID O297 UCCCGCCGCCGUUGUUGUU (DYNII, NM 004945), Rab5a (BC001267), Rab5b (NM Lull 88 SEQ. ID O298 CUUCCCGCCGCCGUUGUUG. 002868), Rab5c (AF 141304), and EEA.1 (XM 018197). Lull 89 SEQ. ID O299 AACUUCCCGCCGCCGUUGU For each gene, four siRNAs duplexes with the highest Lull 9 O SEQ. ID O3 OO UGAACUUCCCGCCGCCGUU scores were selected and a BLAST search was conducted for 40 each of them using the Human EST database. In order to minimize the potential for off-target silencing effects, only Example II those sequences with more than three mismatches against un-related sequences were used. All duplexes were synthe Validation of the Algorithm using DBI, Luciferase, sized at Dharmacon, Inc. as 21-mers with 3'-UU overhangs PLK, EGFR, and SEAP 45 using a modified method of 2'-ACE chemistry Scaringe, Advanced 5'-silyl-2'-orthoester approach to RNA oligonucle The algorithm (Formula VIII) identified siRNAs for five otide synthesis, Methods Enzymol 2000, 317:3 and the anti genes, human DBI, firefly luciferase (fLuc), renilla luciferase sense Strand was chemically phosphorylated to insure maxi (rLuc), human PLK, and human secreted alkaline phos mized activity. phatase (SEAP). Four individual siRNAs were selected on the 50 HeLa cells were grown in Dulbecco's modified Eagle's basis of their SMARTscoresTM derived by analysis of their medium (DMEM) containing 10% fetal bovine serum, anti sequence using Formula VIII (all of the siRNAs would be biotics and glutamine. siRNA duplexes were resuspended in selected with Formula IX as well) and analyzed for their 1x siRNA Universal buffer (Dharmacon, Inc.) to 20LM prior ability to silence their targets expression. In addition to the to transfection. HeLa cells in 12-well plates were transfected scoring, a BLAST search was conducted for each siRNA. To 55 twice with 4 ul of 20LM siRNA duplex in 3 ul Lipofectamine minimize the potential for off-target silencing effects, only 2000 reagent (Invitrogen, Carlsbad, Calif., USA) at 24-hour those target sequences with more than three mismatches intervals. For the transfections in which 2 or 3 siRNA against un-related sequences were selected. Semizarov, et al. duplexes were included, the amount of each duplex was Specificity of short interfering RNA determined through gene decreased, so that the total amount was the same as in trans expression signatures. Proc. Natl. Acad. Sci. U.S.A. 2003, 60 fections with single siRNAs. Cells were plated into normal 100:6347. These duplexes were analyzed individually and in culture medium 12 hours prior to experiments, and protein pools of 4 and compared with several siRNAs that were levels were measured 2 or 4 days after the first transfection. randomly selected. The functionality was measured a per Equal amounts of lysates were resolved by electrophoresis, centage of targeted gene knockdown as compared to controls. blotted, and stained with the antibody specific to targeted All siRNAs were transfected as described by the methods 65 protein, as well as antibodies specific to unrelated proteins, above at 100 nM concentration into HEK293 using Lipo PPI phosphatase and Tsg101 (not shown). The cells were fectamine 2000. The level of the targeted gene expression was lysed in Triton X 100/glycerol solubilization buffer as US 7,820,809 B2 41 42 described previously. Tebar, Bohlander, & Sorkin, Clathrin into the 3' end and cloned into pEYFP-C1 vector (CLON Assembly Lymphoid Myeloid Leukemia (CALM) Protein: TECH, Palo Alto, Calif., USA). GFP-CALM and YFP-Rab5a Localization in Endocytic-coated Pits, Interactions with were described previously Tebar, Bohlander, & Sorkin, Clathrin, and the Impact of Overexpression on Clathrin-me Clathrin Assembly Lymphoid Myeloid Leukemia (CALM) diated Traffic, Mol. Biol. Cell August 1999, 10:2687. Cell 5 Protein. Localization in Endocytic-coated Pits, Interactions lysates were electrophoresed, transferred to nitrocellulose with Clathrin, and the Impact of Overexpression on Clathrin membranes, and Western blotting was performed with several mediated Traffic, Mol. Biol. Cell August 1999, 10:2687. antibodies followed by detection using enhanced chemilumi nescence system (Pierce, Inc). Several x-ray films were ana Example IV lyzed to determine the linear range of the chemiluminescence 10 signals, and the quantifications were performed using densi Validation of the Algorithm Using Eg5, GADPH, tometry and Alphalmager v5.5 software (Alpha Innotech ATEI, MEK2. MEK1, QB, LaminA/C, c-myc, Corporation). In experiments with Eps15R-targeted siRNAs, Human Cyclophilin, and Mouse Cyclophilin cell lysates were subjected to immunoprecipitation with Ab860, and Eps15R was detected in immunoprecipitates by 15 A number of genes have been identified as playing poten Western blotting as described above. tially important roles in disease etiology. Expression profiles The antibodies to assess the levels of each protein by West of normal and diseased kidneys has implicated Edg5 in ern blot were obtained from the following sources: mono immunoglobulin A neuropathy, a common renal glomerular clonal antibody to clathrin heavy chain (TD.1) was obtained disease. Myc1, MEK1/2 and other related kinases have been from American Type Culture Collection (Rockville, Md., associated with one or more cancers, while lamins have been USA); polyclonal antibody to dynamin II was obtained from implicated in muscular dystrophy and other diseases. For Affinity Bioreagents, Inc. (Golden, Colo., USA); monoclonal these reasons, siRNA directed against the genes encoding antibodies to EEA.1 and Rab5a were purchased from BD these classes of molecules would be important research and Transduction Laboratories (Los Angeles, Calif., USA); the therapeutic tools. monoclonal antibody to Tsg101 was purchased from Santa 25 FIG.12 illustrates four siRNAs targeting 10 different genes Cruz Biotechnology, Inc. (Santa Cruz, Calif., USA); the (Table V for sequence and accession number information) monoclonal antibody to GFP was from ZYMEDLaboratories that were selected according to the Formula VIII and assayed Inc. (South San Francisco, Calif., USA); the rabbit polyclonal as individuals and pools in HEK293 cells. The level of siRNA antibodies Ab32 specific to C.-adaptins and Ab20 to CALM induced silencing was measured using the B-DNA assay. were described previously Sorkin, etal, Stoichiometric Inter 30 These studies demonstrated that thirty-six out of the forty action of the Epidermal Growth Factor Receptor with the individual SMART-selected siRNA tested are functional Clathrin-associated Protein Complex AP-2, J. Biol. Chem. (90%) and all 10 pools are fully functional. January 1995, 270:619, the polyclonal antibodies to clathrin light chains A and B were kindly provided by Dr. F. Brodsky Example V (UCSF); monoclonal antibodies to PP1 (BD Transduction 35 Laboratories) and C-Actinin (Chemicon) were kindly pro Validation of the Algorithm Using Bcl2 vided by Dr. M. Dell’Acqua (University of Colorado); Eps15 Ab577 and Eps15R Ab860 were kindly provided by Dr. P. P. Bcl-2 is a ~25 kD. 205-239 amino acid, anti-apoptotic Di Fiore (European Cancer Institute). protein that contains considerable homology with other mem FIG. 11 demonstrates the in vivo functionality of 48 indi 40 bers of the BCL family including BCLX, MCL1, BAX, BAD, vidual siRNAs, selected using Formula VIII (most of them and BIK. The protein exists in at least two forms (Bcl2a, will meet the criteria incorporated by Formula IX as well) which has a hydrophobic tail for membrane anchorage, and targeting 12 genes. Various cell lines were transfected with Bcl2b, which lacks the hydrophobic tail) and is predomi siRNA duplexes (Dup1-4) or pools of siRNA duplexes nantly localized to the mitochondrial membrane. While Bcl2 (Pool), and the cells were lysed 3 days after transfection with 45 expression is widely distributed, particular interest has the exception of CALM (2 days) and B2 (4 days). focused on the expression of this molecule in B and T cells. Note a f1-adaptin band (part of AP-1 Golgi adaptor com Bcl2 expression is down-regulated in normal germinal center plex) that runs slightly slower than P2 adaptin. CALM has B cells yet in a high percentage of follicular lymphomas, Bcl2 two splice variants, 66 and 72 kD. The full-length Eps15R (a expression has been observed to be elevated. Cytological doublet of ~130 kD) and several truncated spliced forms of 50 Studies have identified a common translocation (14:18)(q32: ~100 kD and ~70 kD were detected in Eps15R immunopre q32)) amongst a high percentage (>70%) of these lympho cipitates (shown by arrows). The cells were lysed 3 days after mas. This genetic lesion places the Bcl2 gene injuxtaposition transfection. Equal amounts of lysates were resolved by elec to immunoglobulin heavy chain gene (IgEI) encoding trophoresis and blotted with the antibody specific to a targeted sequences and is believed to enforce inappropriate levels of protein (GFP antibody forYFP fusion proteins) and the anti 55 gene expression, and resistance to programmed cell death in body specific to unrelated proteins PP1 phosphatase or O-ac the follicle center B cells. In other cases, hypomethylation of tinin, and TSG 101. The amount of protein in each specific the Bcl2 promoter leads to enhanced expression and again, band was normalized to the amount of non-specific proteins inhibition of apoptosis. In addition to cancer, dysregulated in each lane of the gel. Nearly all of them appear to be expression of Bcl-2 has been correlated with multiple sclero functional, which establishes that Formula VIII and IX can be 60 sis and various neurological diseases. used to predict siRNAs functionality in general in a genome The correlation between Bcl-2 translocation and cancer wide manner. makes this gene an attractive target for RNAi. Identification To generate the fusion of yellow fluorescent protein (YFP) of siRNA directed against the bc12 transcript (or Bcl2-Igh with Rab5b or Rab5c (YFP-Rab5b or YFP-Rab5c), a DNA fusions) would further our understanding Bcl2 gene function fragment encoding the full-length human Rab5b or Rab5c 65 and possibly provide a future therapeutic agent to battle dis was obtained by PCR using Pfu polymerase (Stratagene) with eases that result from altered expression or function of this a Sacrestriction site introduced into the 5' endanda KpnI site gene. US 7,820,809 B2 43 44 In Silico Identification of Functional siRNA To identify functional and hyperfunctional siRNA against the Bcl2 gene, the sequence for Bcl-2 was downloaded from siRNA 1. GGGAGAUAGUGAUGAAGUA SEQ. ID NO. 3 O1 the NCBI Unigene database and analyzed using the Formula VIII algorithm. As a result of these procedures, both the siRNA 2 GAAGUACAUCCAUUAUAAG SEO. ID NO. 3 O2 sequence and SMARTscoresTM of the Bcl2 siRNA were siRNA 3 GUACGACAACCGGGAGAUA SEQ. ID NO. 303 obtained and ranked according to their functionality. Subse quently, these sequences were BLASTed (database) to siRNA 4. AGAUAGUGAUGAAGUACAU SEO. ID NO. 304 insure that the selected sequences were specific and contained minimal overlap with unrealated genes. The SMART 10 siRNA 5 UGAAGACUCUGCUCAGUUU SEO. ID NO. 3 O5 scoresTM for 6 of the top 10 Bcl-2 siRNA (SEQ ID NOS siRNA 6 GCAUGCGGCCUCUGUUUGA SEO. ID NO. 3 O6 301-310) are identified in Table V under SEQ. ID NOs. 0409 0414. siRNA 7 UGCGGCCUCUGUUUGAUUU SEO. ID NO. 3 O7 In Vivo Testing of Bcl-2 SiRNA Bcl-2 siRNAs having the top ten SMARTscoresTM were 15 siRNA 8 GAGAUAGUGAUGAAGUACA SEQ. ID NO. 3O8 selected and tested in a functional assay to determine silenc siRNA 9 GGAGAUAGUGAUGAAGUAC SEQ. ID NO. 3 O9 ing efficiency. To accomplish this, each of the ten duplexes were synthesized using 2'-O-ACE chemistry and transfected siRNA 1O GAAGACUCUGCUCAGUUUG SEO. ID NO. 310 at 100 nM concentrations into cells. Twenty-four hours later assays were performed on cell extracts to assess the degree of 20 Bcl2 siRNA: Sense Strand, 5'-->3' target silencing. Controls used in these experiments included mock transfected cells, and cells that were transfected with a Example VI non-specific siRNA duplex. The results of these experiments are presented below (and Sequences Selected by the Algorithm in FIG.14) and show that all ten of the selected siRNA induce 25 80% or better silencing of the Bcl2 message at 100 nM con centrations. These data verify that the algorithm successfully Sequences of the siRNAs selected using Formulas (Algo identified functional Bcl2 siRNA and provide a set of func rithms) VIII and IX with their corresponding ranking, which tional agents that can be used in experimental and therapeutic have been evaluated for the silencing activity in vivo in the environments. present study (Formula VIII and IX, respectively).

TABLE W

Gene Accession Formula Formula Name Number SEQ. ID NO. FTllSeqTence VIII IX CLTC NM_004859 SEQ. D NO. 24 OO GAAAGAATCTGTAGAGAAA 76 94.2

CLTC NM_004859 SEQ. D NO. 24 O1 GCAATGAGCTGTTTGAAGA 65 39.9

CLTC NM_004859 SEQ. D NO. 24 O2 TGACAAAGGTGGATAAATT st 38.2

CLTC NM_004859 SEQ. D NO. 24 O3 GGAAATGGATCTCTTTGAA 54 49.4

CLTA NM_001833 SEQ. D NO. 24O4. GGAAAGTAATGGTCCAACA 22 55.5

CLTA NM_001833 SEQ. D NO. 24 OS AGACAGTTATGCAGCTATT 4. 22.9

CLTA NM_001833 SEQ. D NO. 24 O6 CCAATTCTCGGAAGCAAGA 1. 17

CLTA NM_001833 SEQ. D NO. 24. Of GAAAGTAATGGTCCAACAG -1 -13

CLTB NM_001834 SEQ. D NO. 24 O8 GCGCCAGAGTGAACAAGTA 17 sfs

CLTB NM_001834 SEQ. D NO. 24 O9 GAAGGTGGCCCAGCTATGT 15 -8. 6

CLTB NM_001834 SEQ. D NO O311 GGAACCAGCGCCAGAGTGA 13 4 O. 5

CLTB NM_001834 SEQ. D NO O312 GAGCGAGATTGCAGGCATA 2O 61.7

CALM U45976 SEQ. D NO O313 GTTAGTATCTGATGACTTG 36 -34 .. 6

CALM U45976 SEQ. D NO O314 GAAATGGAACCACTAAGAA 33 46.1

CALM U45976 SEQ. D NO O315 GGAAATGGAACCACTAAGA 3O 61.2

CALM U45976 SEQ. D NO O316 CAACTACACTTTCCAATGC 28 6.8

EPS15 NM_001981 SEQ. D NO O317 CCACCAAGATTTCATGATA 48 25.2

EPS15 NM_001981 SEQ. D NO O318 GATCGGAACTCCAACAAGA 43 49.3

EPS15 NM_001981 SEQ. D NO O319 AAACGGAGCTACAGATTAT 39 11.5

US 7,820,809 B2 47 48

TABLE V- continued

Gene Accession Formula Formula Name Number SEO. ID NO. FTllSeqTence VIII IX

GAPDH NM_002046 SEQ. NO. O358 GAAATCCCATCACCATCTT 24 GAPDH NM_002046 SEQ. NO. O359 GACCTCAACTACATGGTT 22 -22.9 GAPDH NM_002046 SEQ. NO. O360 TGGTTTACATGTTCCAATA c-Myc SEQ. NO. O.361 GAAGAAATCGATGTTGTT 31 -117 c-Myc SEQ. NO. O362 ACACAAACTTGAACAGCTA 22 51.3 c-Myc SEQ. NO. O363 GGAAGAAATCGATGTTGTT 18 26 c-Myc SEQ. NO. O364 GAAACGACGAGAACAGTTG 18

MAP2K1 NM_002 755 SEQ. NO. O365 GCACATGGATGGAGGTTCT 26 16 MAP2K1 NM_002 755 SEQ. NO. O366 GCAGAGAGAGCAGATTTGA 16 MAP2K1 NM_002 755 SEQ. NO. O367 GAGGTTCTCTGGATCAAG 14 15.5 MAP2K1 NM_002 755 SEQ. NO. O368 GAGCAGATTTGAAGCAACT 14 18.5 MAP2K2 NM 030662 SEQ. NO. O369 CAAAGACGATGACTTCGAA 37 26. 4 MAP2K2 NM 030662 SEQ. NO. O370 GATCAGCATTTGCATGGAA 24 MAP2K2 NM 030662 SEQ. NO. O371 TCCAGGAGTTTGTCAATAA 17 MAP2K2 NM 030662 SEQ. NO. O372 GGAAGCTGATCCACCTTGA 16 59.2 KNSL1 (EGS) NM_004523 SEQ. NO. O373 GCAGAAATCTAAGGATATA 53 35.8 KNSL1 (EGS) NM_004523 SEQ. NO. O374 CAACAAGGATGAAGTCTAT SO 18.3 KNSL1 (EGS) NM_004523 SEQ. NO. O375 CAGCAGAAATCTAAGGATA 41 32.7 KNSL1 (EGS) NM_004523 SEQ. NO. O376 CTAGATGGCTTTCTCAGTA 39 CyclophilinA NM 021130 SEQ. NO. O377 AGACAAGGTCCCAAAGACA -16 58.1 CyclophilinA NM 021130 SEQ. NO. O378 GGAATGGCAAGACCAGCAA 36 CyclophilinA NM 021130 SEQ. NO. O379 AGAATTATTCCAGGGTTTA 16.1 CyclophilinA NM 021130 SEQ. NO. O38O GCAGACAAGGTCCCAAAGA LAMIN A/C NM 170707 SEQ. NO. O381 AGAAGCAGCTTGAGGATGA 31 38.8 LAMIN A/C NM 170707 SEQ. NO. O.382 GAGCTTGACTTCCAGAAGA 33 22.4 LAMIN A/C NM 170707 SEQ. NO. O383 CCACCGAAGTTCACCCTAA 21 27.5 LAMIN A/C NM 170707 SEQ. NO. O384 GAGAAGAGCTCCTCCATCA 55 30.1

CyclophilinB M60857 SEQ. NO. O385 GAAAGAGCATCTACGGTGA 41 83.9

CyclophilinB M60857 SEQ. NO. O386 GAAAGGATTTGGCTACAAA. 53 59.1

CyclophilinB M60857 SEQ. NO. O387 ACAGCAAATTCCATCGTGT 28.8

CyclophilinB M60857 SEQ. NO. O388 GGAAAGACTGTTCCAAAAA 27

DB1. NM_020548 SEQ. NO. O389 CAACACGCCTCATCCTCTA 27 DB12 NM_020548 SEQ. NO. O390 CATGAAAGCTTACATCAAC 25 -30.8 DB13 NM_020548 SEQ. NO. O391 AAGATGCCATGAAAGCTTA. 17 22 DB14 NM_020548 SEQ. NO. O392 GCACATACCGCCTGAGTCT 15 rLUC1 SEQ. NO. O393 GATCAAATCTGAAGAAGGA 49.2

LUC2 SEQ. NO. O394 GCCAAGAAGTTTCCTAATA SO 13.7

LUC3 SEQ. NO. O395 CAGCATATCTTGAACCATT 41 US 7,820,809 B2 49 50

TABLE V- continued

Gene Accession Formula Formula Name Number SEO. ID NO. FTllSeqTence VIII IX

rLUC4 SEQ. NO. O396 GAACAAAGGAAACGGATGA 39 29.

SeAP1 NM 031313 SEQ. NO. O397 CGGAAACGGTCCAGGCTAT 6 26.

SeAP2 NM 031313 SEQ. NO. O398 GCTTCGAGCAGACATGATA 4. -11.

SeAP3 NM 031313 SEQ. NO. O399 CCTACAGGGTCCTCCTATA 4. 4.

SeAP4 NM 031313 SEQ. NO. O4 OO GCCAAGAACCTCATCATCT 1. -9. fLUC1 SEQ. NO. O401 GATATGGGCTGAATACAAA. 54 4 O. fLUC2 SEQ. NO. O4 O2 GCACTCTGATTGACAAATA 47 54. fLUC3 SEQ. NO. O4 O3. TGAAGTCTCTGATTAAGTA 46 34. fLUC4 SEQ. NO. O404 TCAGAGAGATCCTCATAAA 4 O 11. mCyclo1 NM_008907 SEQ. NO. O4 O5 GCAAGAAGAT CACCATTTC 52 46. mCyclo2 NM_008907 SEQ. NO. O4 O6 GAGAGAAATTTGAGGATGA 36 7 O. mCyclo3 NM_008907 SEQ. NO. O4. Of GAAAGGATTTGGCTATAAG 35 -1. mCyclo 4 NM_008907 SEQ. NO. O408 GAAAGAAGGCATGAACATT 27 1.O.

BCL2 1 NM 000633 SEQ. NO. O409 GGGAGATAGTGATGAAGTA 21 72

BCL2 2 NM 000633 SEQ. NO. O410 GAAGTACATCCATTATAAG 1. 3.

BCL2 3 NM 000633 SEQ. NO. O411 GTACGACAACCGGGAGATA 1. 35.

BCL2 4 NM 000633 SEQ. NO. O412 AGATAGTGATGAAGTACAT -12 22.

BCL2 5 NM 000633 SEQ. NO. O413. TGAAGACTCTGCTCAGTTT 36 19.

BCL2 6 NM 000633 SEQ. NO. O414 GCATGCGGCCTCTGTTTGA 5 -9.

QB1 NM_003365.1 SEQ. NO. O415 GCACACAGCUUACUACAUC 52 - 4.

QB2 NM_003365.1 SEQ. NO. O416 GAAAUGCCCUGGUAUCUCA 49 22.

QB3 NM_003365.1 SEQ. NO. O417 GAAGGAACGUGAUGUGAUC 34 22.

QB4 NM_003365.1 SEQ. NO. O418 GCACUACUCCUGUGUGUGA 28 2O.

ATE1-1 NM_007041 SEQ. NO. O419 GAACCCAGCUGGAGAACUU 45 15.

ATE1-2 NM_007041 SEQ. NO. O42O GAUAUACAGUGUGAUCUUA. 4 O 12.

ATE1-3 NM_007041 SEQ. NO. O421. GUACUACGAUCCUGAUUAU 37 32.

ATE1 - 4 NM_007041 SEQ. NO. O422 GUGCCGACCUUUACAAUUU 35 18.

EGFR-1 NM_005228 SEQ. NO. O423 GAAGGAAACTGAATTCAAA. 68 79.

EGFR-1 NM_005228 SEQ. NO. O424 GGAAATATGTACTACGAAA 49 49.

EGFR-1 NM_005228 SEQ. NO. O425 CCACAAAGCAGTGAATTTA 41 7.

EGFR-1 NM_005228 SEQ. NO. O426 GTAACAAGCTCACGCAGTT 4 O 25.

Example VII be used to develop functional, highly functional, or hyper functional siRNA to each gene. To accomplish this for the Genome-Wide Application of the Algorithm 60 human genome, the entire online ncbi refseq database was accessed through Entrez (efetch). The database was pro The examples described above demonstrate that the algo rithm(s) can successfully identify functional siRNA and that cessed through Formula VIII. For each gene the top 80-100 these duplexes can be used to induce the desirable phenotype scores for siRNAs were obtained and BLASTed to insure of transcriptional knockdown or knockout. Each gene or fam 65 that the selected sequences are specific in targeting the gene ily of genes in each organism plays an important role in of choice. These sequences are provided on the enclosed CDs maintaining physiological homeostasis and the algorithm can in electronic form. US 7,820,809 B2 51 52 With respect to the disks, there are four tables on each disk siRNA=SEQ. ID NO. 0433 UGAAACGGCUGCCUGA copy in text format: Tables XII-XV. Table XII, which is UUU; SEQ. ID NO. 0434 GAAGWUACCUCUUACAUUU; located in a file entitled Table 12.txt, provides a list of the SEQ. ID NO. 0435 GUACGAAACUGUAUGAWUA: SEQ. 80-100 sequences for each target, identified by Formula VIII ID NO. 0436 GGACGAGGCUAUCAUUAUG) may contrib as having the highest relative SMARTscoresTM for the target ute to treatment of this disorder. analyzed. Table XIII, which is located in a file entitled Table In another example, the von Hippel-Lindau (VHL) tumor 13.txt, provides the SMARTscorest M, and for each gene, a Suppressor has been observed to be inactivated at a high pool pick of up to four sequences is denoted. (The denotation frequency in sporadic clear cell renal cell carcinoma (RCC) of “1” in Table XIII means that it is a pool pick.) These pool and RCCs associated with VHL disease. The VHL tumor pick sequences represent the most functional siRNAs for the 10 corresponding target. Any 1, 2, 3, or 4 of the pool pick Suppressor targets hypoxia-inducible factor-1 alpha (HIF-1 sequences could be used for gene silencing. Further, alpha), a transcription factor that can induce vascular endot sequences that are not denoted as pool pick sequences, but helial growth factor (VEGF) expression, for ubiquitination and degradation. Inactivation of VHL can lead to increased that are included on the compact disks may also be used for levels of HIF-1alpha, and subsequent VEGF overexpression. gene silencing either alone or in combination with other 15 sequences. However, their individual relative functionality Such over expression of VEGF has been used to explain the would be less than that of a pool pick sequence. Table XIV. increased (and possibly necessary) Vascularity observed in which is located in a file entitled Table 14.txt, provides an RCC. Thus, functional, highly functional, or hyperfunctional identification of genes by accession number, and Table XV, siRNAs directed against either HIF-1alpha (SEQ. ID NO. which is located in a file entitled Table 15.txt, provides a 0437 GAAGGAACCUGAUGCUUUA: SEQ ID NO. 0438 short name for the genes identified on the disk. The informa GCAUAUAUCUAGAAGGUAU; SEQ. ID NO. 0439 GAA tion contained on the disks is part of this patent application CAAAUACAUGGGAUUA: SEQ. ID NO. 0440 GGACA and are incorporated into the specification by reference. One CAGAUUUAGACUUG) or VEGF (SEQ. ID NO. 0441 may use these tables in order to identify functional siRNAs GAACGUACUUGCAGAUGUG: SEQ. ID NO. 0442 for the gene provided therein, by simply looking for the gene 25 GAGAAAGCAUUUGUUUGUA: SEQ. ID NO. 0443 of interest and an siRNA that is listed as functional. Prefer GGAGAAAGCAUUUGUUUGU; SEQ. ID NO. 0444 ably, one would select one or more of the siRNA that most CGAGGCAGCUUGAGUUAAA) may be useful in the treat optimized for the target of interest and is denoted as a pool ment of renal cell carcinoma. pick. In another example, gene expression of platelet derived Table XII: siRNA selected by Formula VIII 30 growth factor A and B (PDGF-A and PDGF-B) has been See data submitted herewith on a CD-ROM in accordance observed to be increased 22- and 6-fold, respectively, in renal with PCT Administrative Instructions Section 801(a) tissues taken from patients with diabetic nephropathy as com Table XIII: SMARTscoresTM pared with controls. These findings suggest that over expres See data submitted herewith on a CD-ROM in accordance sion of PDGFA and B may play a role in the development of with PCT Administrative Instructions Section 801(a) 35 the progressive fibrosis that characterizes human diabetic Table XIV: Identification of Targets kidney disease. Thus, functional, highly functional, or hyper See data submitted herewith on a CD-ROM in accordance functional siRNAs directed against either PDGFA (SEQ. ID with PCT Administrative Instructions Section 801(a) NO. 0445: GGUAAGAUAWUGUGCUUUA: SEQ. ID NO. Table XV: Description of Targts 0446: CCGCAAAUAUGCAGAAUUA: SEQ. ID NO. 0447: See data submitted herewith on a CD-ROM in accordance 40 GGAUGUACAUGGCGUGUUA: SEQ. ID NO. 0448: with PCT Administrative Instructions Section 801(a) GGUGAAGUUUGUAUGWUUUA) or PDGF B (SEQ. ID Many of the genes to which the described siRNA are NO. 0449: CCGAGGAGCWUUAUGAGAU; SEQ. ID NO. directed play critical roles in disease etiology. For this reason, 0450: GCUCCGCGCUUUCCGAUUU; SEQ. ID NO. 0451 the siRNA listed in the accompanying compact disk may GAGCAGGAAUGGUGAGAUG, SEQ. ID NO. 0452: potentially act as therapeutic agents. A number of prophetic 45 GAACUUGGGAUAAGAGUGU; SEQ. ID NO. 0453 examples follow and should be understood in view of the CCGAGGAGCUUUAUGAGAU; SEQ. ID NO. 0454 siRNA that are identified on the accompanying CD. To isolate UUUAUGAGAUGCUGAGUGA) may be useful in the treat these siRNA, the appropriate message sequence for each gene ment of this form of kidney disorder. is analyzed using one of the before mentioned formulas (pref In another example, a strong correlation exists between the erably formula VIII) to identify potential siRNA targets. Sub 50 over-expression of glucose transporters (e.g. GLUT12) and sequently these targets are BLASTed to eliminate homology cancer cells. It is predicted that cells undergoing uncontrolled with potentially off-targets. cell growth up-regulate GLUT molecules so that they can The list of potential disease targets is extensive. For cope with the heightened energy needs associated with instance, over-expression of Bcl10 has been implicated in the increased rates of proliferation and metastasis. Thus, siRNA development of MALT lymphoma (mucosa associated lym 55 based therapies that target the molecules such as GLUT1 phoid tissue lymphoma) and thus, functional, highly func (also known as SLC2A1: siRNA=SEQ. ID NO.: 0455 tional, or hyperfunctional siRNA directed against that gene GCAAUGAUGUCCAGAAGAA; SEQ. ID NO.: 0456 (e.g. SEQ. ID NO. 0427: GGAAACCUCUCAUUGCUAA; GAAGAAUAUUCAGGACUUA: SEQ.IDNO.: 0457 GAA SEQ. ID NO. 0428: GAAAGAACCUUGCCGAUCA: SEQ. GAGAGUCGGCAGAUGA: SEQ. ID NO.: 0458 CCAA ID NO. 0429: GGAAAUACAUCAGAGCUUA, or SEQ. ID 60 GAGUGUGCUAAAGAA) GLUT12 (also known as NO. 0430: GAAAGUAUGUGUCUUAAGU) may contrib SLCA12: siRNA SEQ. ID NO. 0459: GAGACACU ute to treatment of this disorder. CUGAAAUGAUA: SEQ. ID NO. O460: In another example, studies have shown that molecules that GAAAUGAUGUGGAUAAGAG: SEQ. ID NO. 0461: inhibit glutamine:fructose-6-phosphate aminotransferase GAUCAAAUCCUCCCUGAAA; SEQ. ID NO. 0462: (GFA) may act to limit the symptoms suffered by Type II 65 UGAAUGAGCUGAUGAUUGU) and other related trans diabetics. Thus, functional, highly functional, or hyperfunc porters, may be of value in treating a multitude of malignan tional siRNA directed against GFA (also known as GFPT1: C1GS. US 7,820,809 B2 53 54 The siRNA sequences listed above are presented in a 5'3" siRNAs designed to silence the above-specified genes were sense strand direction. In addition, siRNA directed against the functional, while 100% of the pools that contain the same targets listed above as well as those directed against other siRNA duplexes were functional. targets and listed in the accompanying compact disk may be useful as therapeutic agents. Example X Example VIII Highly Functional siRNA Evidence for the Benefits of Pooling Pools of five siRNAs in which each two siRNAs overlap to 10 10-90% resulted in 98% functional entities (>80% silencing). Evidence for the benefits of pooling have been demon Pools of siRNAs distributed throughout the mRNA that were strated using the Reporter gene, luciferase. Ninety siRNA evenly spaced, covering an approximate 20-2000 base pair duplexes were synthesized using Dharmacon proprietary range, were also functional. When the pools of siRNA were ACER chemistry against one of the standard reporter genes: positioned continuously head to tail relative to mRNA firefly luciferase. The duplexes were designed to start two 15 sequences and mimicked the natural products of Dicer base pairs apart and to cover approximately 180 base pairs of cleaved long double stranded RNA, 98% of the pools evi the luciferase gene (see sequences in Table IV). Subse denced highly functional activity (>95% silencing). quently, the siRNA duplexes were co-transfected with a luciferase expression reporter plasmid into HEK293 cells Example XI using standard transfection protocols and luciferase activity Human Cyclophyline was assayed at 24 and 48 hours. Transfection of individual siRNAs showed standard distri Table IV above lists the siRNA sequences for the human bution of inhibitory effect. Some duplexes were active, while cyclophyline protein. A particularly functional siRNA may others were not. FIG. 15 represents a typical screen of ninety be selected by applying these sequences to any of Formula I siRNA duplexes (SEQ. ID NO. 0032-0120) positioned two 25 to VII above. base pairs apart. As the figure Suggests, the functionality of Alternatively, one could pool 2, 3, 4, 5 or more of these the siRNA duplex is determined more by a particular sequences to create a kit for silencing a gene. Preferably, sequence of the oligonucleotide than by the relative oligo within the kit there would be at least one sequence that has a nucleotideposition within a gene or excessively sensitive part relatively high predicted functionality when any of Formulas of the mRNA, which is important for traditional anti-sense 30 technology. 1-VII is applied. When two continuous oligonucleotides were pooled Example XII together, a significant increase in gene silencing activity was observed (see FIGS. 16A and 16B). A gradual increase in Sample Pools of siRNAs and their Application to efficacy and the frequency of pools functionality was 35 Human Disease observed when the number of siRNAs increased to 3 and 4 (see FIGS. 16A, 16B, 17A, and 17B). Further, the relative The genetic basis behind human disease is well docu positioning of the oligonucleotides within a pool did not mented and siRNA may be used as both research or diagnos determine whether a particular pool was functional (see tic tools and therapeutic agents, either individually or in FIGS. 18A and 18B, in which 100% of pools of oligonucle 40 pools. Genes involved in signal transduction, the immune otides distanced by 2, 10 and 20 base pairs were functional). response, apoptosis, DNA repair, cell cycle control, and a However, relative positioning may nonetheless have an variety of other physiological functions have clinical rel impact. An increased functionality may exist when the siRNA evance and therapeutic agents that can modulate expression are positioned continuously head to toe (5' end of one directly of these genes may alleviate Some or all of the associated adjacent to the 3' end of the others). 45 symptoms. In some instances, these genes can be described as Additionally, siRNA pools that were tested performed at a member of a family or class of genes and siRNA (randomly, least as well as the best oligonucleotide in the pool, under the conventionally, or rationally designed) can be directed experimental conditions whose results are depicted in FIG. against one or multiple members of the family to induce a 19. Moreover, when previously identified non-functional and desired result. marginally (semi) functional siRNA duplexes were pooled 50 To identify rationally designed siRNA to each gene, the together in groups of five at a time, a significant functional sequence was analyzed using Formula VIII to identify a cooperative action was observed. (See FIG. 20) In fact, pools SMARTpool containing the functional sequences. To confirm of semi-active oligonucleotides were 5 to 25 times more the activity of these sequences, the siRNA are introduced into functional than the most potent oligonucleotide in the pool. a cell type of choice (e.g. HeLa cells, HEK293 cells) and the Therefore, pooling several siRNA duplexes together does not 55 levels of the appropriate message are analyzed using one of interfere with the functionality of the most potent siRNAs several art proven techniques. SiRNA having heightened lev within a pool, and pooling provides an unexpected significant els of potency can be identified by testing each of the before increase in overall functionality mentioned duplexes at increasingly limiting concentrations. Similarly, siRNA having increased levels of longevity can be Example IX 60 identified by introducing each duplex into cells and testing functionality at 24, 48, 72, 96, 120, 144, 168, and 192 hours Pooling Across Species after transfection. Agents that induce >95% silencing at sub nanomolar concentrations and/or induce functional levels of Experiments were performed on the following genes: silencing for >96 hours are considered hyperfunctional. B-galactosidase, Renilla luciferase, and Secreted alkaline 65 The following are non-limiting examples of families of phosphatase, which demonstrates the benefits of pooling. (see proteins to which siRNA described in this document are FIGS. 21A, 21B and 21C) Approximately 50% of individual targeted against: US 7,820,809 B2 55 56 Transporters, Pumps, and Channels Superior mesenteric ganglia. Similarly, the kinetic properties Transporters, pumps, and channels represent one class of of the three erg potassium channels are not homogeneous. genes that are attractive targets for siRNAs. One major class Ergil and erg2 channels are relatively slow activating delayed of transporter molecules are the ATP-binding cassette (ABC) rectifiers whereas the erg3 current activates rapidly and then transporters. To date, nearly 50 human ABC-transporter 5 exhibits a predominantly transient component that decays to genes have been characterized and have been shown to be a Sustained plateau. The current properties of all three chan involved in a variety of physiological functions including nels are sensitive to methanesulfonanilides, suggesting a high transport of bile salts, nucleosides, chloride ions, cholesterol, degree of conservation in the pore structure of all three pro toxins, and more. Predominant among this group are MDR1 teins. (which encodes the P-glycoprotein, NP 000918), the MDR 10 Recently, the erg family of K channels has been impli related proteins (MRP1-7), and the breast cancer resistance cated in human disease. Consistent with the observation that protein (B.CRP). In general, these transporters share a com ergl is expressed in the heart, single strand conformation mon structure, with each protein containing a pair of ATP polymorphism and DNA sequence analyses have identified binding domains (also known as nucleotide binding folds, HERG (human erg1) mutations in six long-QT-syndrome NBF) and two sets of transmembrane (TM) domains, each of 15 (LQT) families, an inherited disorder that results in sudden which typically contains six membrane-spanning C-helices. death from a ventricular tachyarrythmia. Thus siRNA The genes encoding this class of transporter are organized as directed against this group of molecules (e.g. KCNH1-8) will either full transporters (i.e. containing two TM and two NBF be of extreme therapeutic value. domains) or as half transporters that assemble as either Another group of channels that are potential targets of homodimers or heterodimers to create functional transport siRNAs are the CLCA family that mediate a Ca"-activated ers. As a whole, members of the family are widely dispersed Cl conductance in a variety of tissues. To date, two bovine throughout the genome and show a high degree of amino acid (bCLC1; b0LCA2 (Lu-ECAM-1)), three mouse (mCLCA1; sequence identify among eukaryotes. mCLCA2; mCLCA3) and four human (hCLCA1; hocLCA2; ABC-transporters have been implicated in several human hCLCA3; hCLCA4) CLCA family members have been iso diseases. For instance, molecular efflux pumps of this type 25 lated and patch-clamp studies with transfected human embry play a major role in the development of drug resistance exhib onic kidney (HEK-293) cells have shown that bOLCA1, ited by a variety of cancers and pathogenic microorganisms. mCLCA1, and hCLCA1 mediate a Ca"-activated CI con In the case of human cancers, increased expression of the ductance that can be inhibited by the anion channel blocker MDR1 gene and related pumps have been observed to gen DIDS and the reducing agent dithiothreitol (DTT). erate drug resistance to a broad collection of commonly used 30 The protein size, structure, and processing seem to be chemotherapeutics including doxorubicin, daunorubicin, similar among different CLCA family members and has been vinblastine, Vincristine, colchicines. In addition to the contri studied in greatest detail for Lu-ECAM-1. The Lu-ECAM-1 bution these transporters make to the development of multi open reading frame encodes a precursor glycoprotein of 130 drug resistance, there are currently 13 human genetic diseases kDa that is processed to a 90-kDa amino-terminal cleavage associated with defects in 14 different transporters. The most 35 product and a group of 30- to 40-kDa glycoproteins that are common of these conditions include cystic fibrosis, Stargardt glycosylation variants of a single polypeptide derived from its disease, age-related macular degeneration, adrenoleukodys carboxy terminus. Both subunits are associated with the outer trophy, Tangier disease, Dubin-Johnson syndrome and pro cell surface, but only the 90-kDa subunit is thought to be gressive familial intrahepatic cholestasis. For this reason, anchored to the cell membrane via four transmembrane siRNAs directed against members of this, and related, fami 40 domains. lies are potentially valuable research and therapeutic tools. Although the protein processing and function appear to be With respect to channels, analysis of Drosophila mutants conserved among CLCA homologs, significant differences has enabled the initial molecular isolation and characteriza exist in their tissue expression patterns. For example, bovine tion of several distinct channels including (but not limited to) Lu-ECAM-1 is expressed primarily in vascular endothelia, potassium (K-) channels. This list includes shaker (Sh), 45 bCLCA1 is exclusively detected in the trachea, and hCLCA1 which encodes a voltage activated K' channel, slowpoke is selectively expressed in a Subset of human intestinal epi (Slo), a Ca" activated K" channel, and ether-a-go-go (Eag). thelial cells. Thus the emerging picture is that of a multigene The Eag family is further divided into three subfamilies: Eag, family with members that are highly tissue specific, similar to Elk (eag-like K channels), and Erg (Eag related genes). the CIC family of voltage-gated Cl channels. The human The Erg subfamily contains three separate family members 50 channel, hCLCA2, is particular interesting from a medical (Erg 1-3) that are distantly related to the sh family of voltage and pharmacological standpoint. CLCA2 is expressed on the activated K' channels. Like sh, erg polypetides contain the luminal Surface of lung vascular endothelia and serves as an classic six membrane spanning architecture of K channels adhesion molecule for lung metastatic cancer cells, thus (S1-S6) but differ in that each includes a segment associated mediating vascular arrest and lung colonization. Expression with the C-terminal cytoplasmic region that is homologous to 55 of this molecule in normal mammary epithelium is consis cyclic nucleotide binding domains (cNBD). Like many iso tently lost in human breast cancer and in nearly all tumori lated ion channel mutants, erg mutants are temperature-sen genic breast cancer cell lines. Moreover, re-expression of sitive paralytics, a phenotype caused by spontaneous repeti hCLCA2 in human breast cancer cells abrogates tumorige tive firing (hyperactivity) in neurons and enhanced nicity in nude mice, implying that hCLCA2 acts as a tumour transmitter release at the neuromuscular junction. 60 Suppressor in breast cancer. For these reasons, siRNA Initial studies on the tissue distribution of all three mem directed against CLCA family members and related channels bers of the erg subfamily show two general patterns of expres may prove to be valuable in research and therapeutic venues. Sion. Erg1 and erg3 are broadly expressed throughout the nervous system and are observed in the heart, the Superior Transporters Involved in Synaptic Transmission. mesenteric ganglia, the celiac ganglia, the retina, and the 65 Synaptic transmission involves the release of a neurotrans brain. In contrast, erg2 shows a much more restricted pattern mitter into the synaptic cleft, interaction of that transmitter of expression and is only observed in celiac ganglia and with a postsynaptic receptor, and Subsequent removal of the US 7,820,809 B2 57 58 transmitter from the cleft. In most synapses the signal is transporters with overlapping Substrate specificities. The terminated by a rapid reaccumulation of the neurotransmitter class of organic anion transporters plays a critical role in the into presynaptic terminals. This process is catalyzed by spe elimination of a large number of drugs (e.g., antibiotics, che cific neurotransmitter transporters that are often energized by motherapeutics, diuretics, nonsteroidal anti-inflammatory the electrochemical gradient of sodium across the plasma drugs, radiocontrast agents, cytostatics); drug metabolites membrane of the presynaptic cells. (especially conjugation products with glutathione, glucu Aminobutyric acid (GABA) is the major inhibitory neu ronide, glycine, Sulfate, acetate); and toxicants and their rotransmitter in the central nervous system. The inhibitory metabolites (e.g., mycotoxins, herbicides, plasticizers, glu action of GABA, mediated through GABA/GABA recep tathione S-conjugates of polyhaloalkanes, polyhaloalkenes, tors, and is regulated by GABA transporters (GATs), integral 10 hydroquinones, aminophenols), many of which are specifi membrane proteins located perisynaptically on neurons and cally harmful to the kidney. glia. So far four different carriers (GAT1-(GAT4) have been Over the past couple of years the number of identified cloned and their cellular distribution has been partly worked anion transporting molecules has grown tremendously. out. Comparative sequence analysis has revealed that GABA Uptake of organic anions (OA) across the basolateral mem transporters are related to several other proteins involved in 15 brane is mediated by the classic sodium-dependent organic neurotransmitter uptake including gamma-aminobutyric acid anion transport system, which includes C.-ketoglutarate transporters, monoamine transporters, amino acid transport (C-KG)/OA exchange via the organic anion transporter ers, certain “orphan transporters, and the recently discov (OAT1) and sodium-ketoglutarate cotransport via the Na/ ered bacterial transporters. Each of these proteins has a simi dicarboxylate cotransporter (SDCT2). The organic anion lar 12 transmembrane helices topology and relies upon the transporting polypetide, Oatp1, and the kidney-specific OAT Na+/Cl- gradient for transport function. Transport rates are K1 and OAT-K2 are seen as potential molecules that mediate dependent on Substrate concentrations, with half-maximal facilitated OA efflux but could also be involved in reabsorp effective concentrations for transport frequently occurring in tion via an exchange mechanism. Lastly the PEPT1 and the Submicromolar to low micromolar range. In addition, PEPT2 mediate luminal uptake of peptide drugs, whereas transporter function is bidirectional, and non-vesicular efflux 25 CNT1 and CNT2 are involved in reabsorption of nucleosides of transmitter may contribute to ambient extracellular trans The organic anion-transporting polypeptide 1 (Oatp1) is a mitter levels. Na"- and ATP-independent transporter originally cloned Recent evidence Suggests that GABA transporters, and from rat liver. The tissue distribution and transport properties neurotransmitter transporters in general, are not passive play of the Oatp1 gene product are complex. Oatp1 is localized to ers in regulating neuronal signaling; rather, transporter func 30 the basolateral membrane of hepatocytes, and is found on the tion can be altered by a variety of initiating factors and signal apical membrane of S3 proximal tubules. Studies with tran transduction cascades. In general, this functional regulation siently transfected cells (e.g. HeLa cells) have indicated that occurs in two ways, either by changing the rate of transmitter Oatp1 mediates transport of a variety of molecules including flux through the transporter or by changing the number of taurocholate, estrone-3-sulfate, aldosterone, cortisol, and functional transporters on the plasma membrane. A recurring 35 others. The observed uptake of taurocholate by Oatp1 theme in transporter regulation is the rapid redistribution of expressed in X. laevis oocytes is accompanied by efflux of the transporter protein between intracellular locations and the GSH, Suggesting that transport by this molecule may be glu cell Surface. In general, this functional modulation occurs in tathione dependent. part through activation of second messengers such as kinases, Computer modeling Suggests that members of the Oatp phosphatases, arachidonic acid, and pH. However, the 40 family are highly conserved, hydrophobic, and have 12 trans mechanisms underlying transporter phosphorylation and membrane domains. Decreases in expression of Oatp family transporter redistribution have yet to be fully elucidated. members have been associated with cholestatic liver diseases GABA transporters play a pathophysiological role in a and human hepatoblastomas, making this family of proteins number of human diseases including temporal lobe epilepsy of key interest to researchers and the medical community. For and are the targets of pharmacological interventions. Studies 45 these reasons, siRNAs directed against OAT family members in seizure sensitive animals show some (but not all) of the (including but not limited to SLC21A2, 3, 6, 8, 9, 11, 12, 14, GAT transporters have altered levels of expression at times 15, and related transporters) are potentially useful as research prior to and post seizure, Suggesting this class of transporter and therapeutic tools. may affect epileptogenesis, and that alterations following seizure may be compensatory responses to modulate seizure 50 Nucleoside Transporters. activity. For these reasons, siRNAs directed against members Nucleoside transporters play key roles in physiology and of this family of genes (including but not limited to pharmacology. Uptake of exogenous nucleosides is a critical SLCG6A1-12) may prove to be valuable research and thera first step of nucleotide synthesis in tissues such as bone mar peutic tools. row and intestinal epithelium and certain parasitic organisms 55 that lack de novo pathways for purine biosynthesis. Nucleo Organic Ion Transporters. side transporters also control the extracellular concentration The human body is continuously exposed to a great variety of adenosine in the vicinity of its cell surface receptors and of xenobiotics, via food, drugs, occupation, and environment. regulate processes such as neurotransmission and cardiovas Excretory organs such as kidney, liver, and intestine defend cular activity. Adenosine itself is used clinically to treat car the body against the potentially harmful effects of these com 60 diac arrhythmias, and nucleoside transport inhibitors such as pounds by transforming them into less active metabolites that dipyridamole, dilaZep, and draflazine function as coronary are Subsequently secreted from the system. vasodilators. Carrier-mediated transport of xenobiotics and their In mammals, plasma membrane transport of nucleosides is metabolites exist for the active secretion of organic anions brought about by members of the concentrative, Na'-depen and cations. Both systems are characterized by a high clear 65 dent (CNT) and equilibrative, Na'-independent (ENT) ance capacity and tremendous diversity of Substances nucleoside transporter families. CNTs are expressed in a tis accepted, properties that result from the existance of multiple Sue-specific fashion; ENTs are present in most, possibly all, US 7,820,809 B2 59 60 cell types and are responsible for the movement of hydro Ion Exchangers philic nucleosides and nucleoside analogs down their concen Intracellular pH regulatory mechanisms are critical for the tration gradients. In addition, structure/function studies of maintenance of countless cellular processes. For instance, in ENT family members have predicted these molecules to con muscle cells, contractile processes and metabolic reactions tain eleven transmembrane helical segments with an amino are influenced by pH. During periods of increased energy terminus that is intracellular and a carboxyl terminus that is demands and ischemia, muscle cells produce large amounts extracellular. The proteins have a large glycosylated loop of lactic acid that, without quick and efficient disposal, would between TMs 1 and 2 and a large cytoplasmic loop between lead to acidification of the sarcoplasm. TMs 6 and 7. Recent investigations have implicated the TM Several different transport mechanisms have evolved to 3-6 region as playing a central role in Solute recognition. The 10 maintain a relatively constant intracellular pH. The relative medical importance of the ENT family of proteins is broad. In contribution of each of these processes varies with cell type, humans adenosine exerts a range of cardioprotective effects the metabolic requirements of the cell, and the local environ and inhibitors of ENTs are seen as being valuable in allevi mental conditions. Intracellular pH regulatory processes that ating a variety of cardio/cardiovascular ailments. In addition, have been characterized functionally include but are not lim responses to nucleoside analog drugs has been observed to 15 ited to the Na/H" exchange, the Na(HCO), cotransport, and vary considerably amongst e.g. cancer patients. While some the Na-dependent and -independent Cl/base exchangers. forms of drug resistance have been shown to be tied to the As bicarbonate and CO comprise the major pH buffer of up-regulation of ABC-transporters (e.g. MDR1), resistance biological fluids, sodium biocarbonate cotransporters may also be the result of reduced drug uptake (i.e. reduced (NBCs) are critical. Studies have shown that these molecules ENT expression). Thus, a clearer understanding of ENT exist in numerous tissues including the kidney, brain, liver, transporters may aid in optimizing drug treatments for cornea, heart, and lung, Suggesting that NBCs play an impor patients suffering a wide range of malignancies. For these tant role in mediating HCO transport in both epithelial as reasons, siRNAS directed against this class of molecules (in well as nonepithelial cells. cluding SLC28A1-3, SLC29A1-4, and related molecules) Recent molecular cloning experiments have identified the may be useful as therapeutic and research tools. Sulfate 25 existence of four NBC isoforms (NBC1, 2, 3 and 4) and two Transporters. NBC-related proteins, AE4 and NCBE (Anion Exchanger 4 All cells require inorganic sulfate for normal function. and Na-dependent Chloride-Bicarbonate Exchanger). The Sulfate is the fourth most abundant anion in human plasma secondary structure analyses and hydropathy profile of this and is the major source of sulfur in many organisms. Sulfation family predict them to be intrinsic membrane proteins with 12 of extracellular matrix proteins is critical for maintaining 30 putative transmembrane domains and several family mem normal cartilage metabolism and Sulfate is an important con bers exhibit N-linked glycosylation sites, protein kinases. A stituent of myelin membranes found in the brain and C, casein kinase II, and ATP/GTP-binding consensus Because Sulfate is a hydrophilicanion that cannot passively phosphorylation sites, as well as potential sites for myristy cross the lipid bilayer of cell membranes, all cells require a lation and amidation. AE4 is a relatively recent addition to mechanism for Sulfate influx and efflux to ensure an optimal 35 this family of proteins and shows between 30–48% homology Supply. To date, a variety of Sulfate transporters have been with the other family members. When expressed in COS-7 identified in tissues from many origins. These include the cells and Xenopus oocytes AE4 exhibits sodium-independent renal sulfate transporters (NaSi-1 and Sat-1), the ubiquitously and DIDS-insensitive anion exchanger activity. Exchangers expressed diastrophic dysplasia sulfate transporter (DTDST), have been shown to be responsible for a variety of human the intestinal sulfate transporter (DRA), and the erythrocyte 40 diseases For instance, mutations in three genes of the anion anion exchanger (AE 1). Most, if not all, of these molecules transporterfamily (SLC) are believed to cause known heredi contain the classic 12 transmembrane spanning domain archi tary diseases, including chondrodysplasia (SLC26A2, DTD), tecture commonly found amongst members of the anion diarrhea (A3, down-regulated in adenoma/chloride-losing transporter Superfamily. diarrhea protein: DRA/CLD), and goiter/deafness syndrome Recently three different sulfate transporters have been 45 (A4, pendrin). Moreover, mutations in Na+/HCO3 co-trans associated with specific human genetic diseases. Family porters have also been associated with various human mala members SLC26A2, SLC26A3, and SLC26A4 have been dies. For these reasons, siRNAs directed against these sorts of recognized as the disease genes mutated in diastrophic dys genes (e.g. SLC4A4-10, and related genes) may be useful for plasia, congenital chloride diarrhea (CLD), and Pendred syn therapeutic and research purposes. drome (PDS), respectively. DTDST is a particularly complex 50 disorder. The gene encoding this molecule maps to chromo Receptors Involved in Synaptic Transmission Some 5q and encodes two distinct transcripts due to alterna In all vertebrates, fast inhibitory synaptic transmission is tive exon usage. In contrast to other sulfate transporters (e.g. the result of the interaction between the neurotransmitters Sat-1) anion movement by the DTDST protein is markedly glycine (Gly) and Y-aminobutyric acid (GABA) and their inhibited by either extracellular chloride or bicarbonate. 55 respective receptors. The Strychnine-sensitive glycine recep Impaired function of the DTDST gene product leads to under tor is especially important in that it acts in the mammalian Sulfation of proteoglycans and a complex family of reces spinal cord and brain stem and has a well-established role in sively inherited osteochondrodysplasias (achondrogenesis the regulation of locomotor behavior. type I B, atelosteogenesis type II, and diastrophic dysplasia) Glycine receptors display significant sequence homology with clinical features including but not limited to, dwarfism, 60 to several other receptors including the nicotinic acetylcho spinal deformation, and specific joint abnormalities. Interest line receptor, the aminobutyric acid receptor type A ingly, while epidemiological studies have shown that the dis (GABAR), and the serotonin receptor type 3 (5-HTR) sub ease occurs in most populations, it is particularly prevalent in units. As members of the Superfamily of ligand-gated ion Finland owing to an apparent founder effect. For these rea channels, these polypeptides share common topological fea Sons, siRNAS directed against this class of genes (including 65 tures. The glycine receptor is composed of two types of gly but not limited to SLC26A1-9, and related molecules) may be cosylated integral membrane proteins (C.1-C4 and B) potentially helpful in both therapeutic and research venues. arranged in a pentamenc Suprastructure. The alpha Subunit US 7,820,809 B2 61 62 encodes a large extracellular, N-terminal domain that carries reasons, siRNAS directed against this class of genes (e.g. the structural determinants essential for agonist and antago KLK1-15) may prove valuable in both research and therapeu nist binding, followed by four transmembrane spanning tic settings. regions (TM1-TM4), with TM2 playing the critical role of ADAM Proteins forming the inner wall of the chloride channel. 5 The process of fertilization takes place in a series of dis The density, location, and subunit composition of glycine crete steps whereby the sperm interacts with, i) the cumulus neurotransmitter receptors changes over the course of devel cells and the hyaluronic acid extracellular matrix (ECM) in opment. It has been observed that the amount of GlyR gene which they are embedded, ii) the egg's own ECM, called the translation (assessed by the injection of developing rat cere zona pellucida (ZP), and iii) the egg plasma membrane. Dur bral cortex mRNA into Xenopus oocytes) decreases with age, 10 ing the course of these interactions, the "acrosome reaction.” whereas that of GABARs increases. In addition, the type and the exocytosis of the acrosome vesicle on the head of the location of mRNAs coding for GlyR changes over the course sperm, is induced, allowing the sperm to penetrate the ZP and of development. For instance in a study of the expression of gain access to the perivitelline space. This process exposes alpha 1 and alpha 2 subunits in the rat, it was observed that (in new portions of the sperm membrane, including the inner embryonic periods E11-18) the mantle Zone was scarce in the 15 acrosomal membrane and the equatorial segment, regions of alpha 1 mRNA, but the germinal Zone (matrix layer) at E11 the sperm head that can participate in initial gamete mem 14 expressed higher levels of the message. At postnatal day 0 brane binding. (PO), the alpha 1 signals became manifested throughout the The interactions of the gamete plasma membranes appear gray matter of the spinal cord. By contrast, the spinal tissues to involve multiple ligands and receptors and are frequently at P0 exhibited the highest levels of alpha 2 mRNA, which compared to leukocyte-endothelial interactions. These inter decreased with the postnatal development. actions lead to a series of signal transduction events in the In both, man and mouse mutant lines, mutations of GlyR egg, known as collectively as egg activation and include the Subunit genes result in hereditary motor disorders character initiation of oscillations in intracellular calcium concentra ized by exaggerated Startle responses and increased muscle tion, the exit from meiosis, the entry into the first embryonic tone. Pathological alleles of the Glral gene are associated 25 mitosis, and the formation of a block to polyspermy via the with the murine phenotypes oscillator (spd”) and spasmodic release of ZP-modifying enzymes from the eggs cortical (spd). Similarly, a mutant allele of Glrb has been found to granules. Ultimately, sperm and egg not only adhere to each underly the molecular pathology of the spastic mouse (spa). other but also go on to undergo membrane fusion, making one Resembling the situation in the mouse, a variety of GLRA1 cell (the Zygote) from two. mutant alleles have been shown to be associated with the 30 Studies on the process of sperm-egg interactions have iden human neurological disorder hyperekplexia or startle disease. tified a number of proteins that are crucial for fertilization. For these reasons, siRNA directed against glycine receptors One class of proteins, called the ADAM family (A Disintegrin (GLRA1-3, GLRB, and related molecules), glutamate recep And Metalloprotease), has been found to be important in tors, GABA receptors, ATP receptors, and related neurotrans spermatogenesis and fertilization, as well as various develop mitter receptor molecules may be valuable therapeutic and 35 mental systems including myogenesis and neurogenesis. research reagents. Members of the family contain a disintegrin and metallopro tease domain (and therefore have (potentially) both cell adhe Proteases sion and protease activities), as well as cysteine-rich regions, Kallikreins 40 epidermal growth factor (EGF)-like domains, a transmem One important class of proteases are the kallikreins, serine brane region, and a cytoplasmic tail. Currently, the ADAM endopeptidases that split peptide Substrates preferentially on gene family has 29 members and constituents are widely the C-terminal side of internal arginyl and lysyl residues distributed in many tissues including the brain, testis, epid Kallikreins are generally divided into two distinct groups, idymis, ovary, breast, placenta, liver, heart, lung, bone, and muscle. plasma kallikreins and tissue kallikreins. Tissue kallikreins 45 represent a large group of enzymes that have Substantial simi One of the best-studied members of the ADAM family is larities at both the gene and protein level. The genes encoding fertilin, a heterodimeric protein comprised of at least two this group are frequently found on a single chromosome, are subunits, fertilin alpha and fertilin beta. The fertilin beta gene organized in clusters, and are expressed in a broad range of (ADAM2) has been disrupted with a targeting gene construct tissues (e.g. pancreas, ovaries, breast). In contrast, the plasma 50 corresponding to the exon encoding the fertilin beta disinte form of the enzyme is encoded by a single gene (e.g. KLK3) grin domain. Sperm from males homozygous for disruptions that has been localized to chromosome 4q24-35 in humans. in this region exhibit defects in multiple facets of sperm The gene encoding plasma kallikrein is expressed solely in function including reduced levels of sperm transit from the the liver, contains 15 exons, and encodes a glycoprotein that uterus to the oviduct, reduced sperm-ZPbinding, and reduced is translated as a preprotein called prekallikrein Kallikreins 55 sperm-egg binding, all of which contribute to male infertility. are believed to play an important role in a host of physiologi Recently, four new ADAM family members (ADAM cal events For instance, the immediate consequence of plasma 24-27) have been isolated. The deduced amino acid prekallikrein activation is the cleavage of high molecular sequences show that all four contain the complete domain weight kininogen (HK) and the Subsequent liberation of organization common to ADAM family members and North bradykinin, a nine amino acid vasoactive peptide that is an 60 ern Blot analysis has shown all four to be specific to the testes. important mediator of inflammatory responses. Similarly, SiRNAs directed against this class of genes (e.g. ADAM2 and plasma kallikrein promotes single-chain urokinase activation related proteins) may be useful as research tools and thera and Subsequent plasminogen activation, events that are criti peutics directed toward fertility and birth control. cal to blood coaggulation and wound healing. Aminopeptidases Disruptions in the function of kallikreins have been impli 65 Aminopeptidases are proteases that play critical roles in cated in a variety of pathological processes including imbal processes such as protein maturation, protein digestion in its ances in renal function and inflammatory processes. For these terminal stage, regulation of hormone levels, selective or US 7,820,809 B2 63 64 homeostatic protein turnover, and plasmid stabilization. vesicles are introduced into the target cell. Studies over the These enzymes generally have broad Substrate specificity, last two decades have identified the toxic components as occur in several forms and play a major role in physiological Granzymes, a family of serine proteases that are expressed homeostasis. For instance, the effects of bradykinin, angio exclusively by cytotoxic T lymphocytes and natural killer tensin converting enzyme (ACE), and other vasoactive mol (NK) cells. These agents are stored in specialized lytic gran ecules are muted by one of several peptidases that cleave the ules and enter the target cell via endocytosis. Like caspases, molecule at an internal position and eliminate its ability to cysteine proteases that play an important role in apoptosis, bind its cognate receptor (e.g. for bradykinin, the B2-recep granzymes can cleave proteins after acidic residues, espe tor). cially aspartic acid, and induce apoptosis in the recipient cell. Among the enzymes that can cleave bradykinin is the 10 Granzymes have been grouped into three subfamilies membrane bound aminopeptidase P, also referred to as ami according to Substrate specificity. Members of the granzyme noacylproline aminopeptidase, proline aminopeptidase; family that have enzymatic activity similar to the serine pro X-Pro aminopeptidase (eukaryote) and XPNPEP2. Ami tease chymotrypsin are encoded by a gene cluster termed the nopeptidase P is an aminoacylproline aminopeptidase spe chymase locus. Similarly, granzymes with trypsin-like cific for NH-terminal Xaa-proline bonds. The enzyme i) is a 15 specificities are encoded by the tryptase locus, and a third mono-zinc-containing molecule that lacks any of the typical subfamily cleaves after unbranched hydrophobic residues, metal binding motifs found in other Zinc metalloproteases, ii) especially methionine, and are encoded by the Met-ase has an active-site configuration similar to that of other mem locus. All granzymes are synthesized as Zymogens and, after bers of the MG peptidase family, and iii) is present in a variety clipping of the leader peptide, obtain maximal enzymatic of tissues including but not limited to the lung, kidney, brain, activity Subsequent to the removal of an amino-terminal and intestine. dipeptide. Aminopeptidases play an important role in a diverse set of Granzymes have been found to be important in a number of human diseases. Low plasma concentrations of aminopepti important biological functions including defense against dase Pare a potential predisposing factor for development of intracellular pathogens, graft versus host reactions, the Sus angio-oedema in patients treated with ACE inhibitors, and 25 ceptibility to transplantable and spontaneous malignancies, inhibitors of aminopeptidase P may act as cardioprotectors lymphoid homeostasis, and the tendency toward auto-im against other forms of illness including, but not limited to mune diseases. For these reasons, siRNAS directed against myocardial infarction. For these reasons, siRNAs directed granszymes (e.g. GZMA, GZMB, GZMH, GZHK, GZMM) against this family of proteins (including but not limited to and related serine proteases may be useful research and thera XPNPEP1 and related proteins) may be useful as research 30 peutic reagents. and therapeutic tools. Kinases Serine Proteases Protein Kinases (PKs) have been implicated in a number of biological processes. Kinase molecules play a central role in One important class of proteases are the serine proteases. modulating cellular physiology and developmental deci Serine proteases share a common catalytic triad of three 35 amino acids in their active site (serine (nucleophile), aspartate sions, and have been implicated in a large list of human (electrophile), and histidine (base)) and can hydrolyze either maladies including cancer, diabetes, and others. esters or peptide bonds utilizing mechanisms of covalent During the course of the last three decades, over a hundred catalysis and preferential binding of the transition state. distinct protein kinases have been identified, all with pre Based on the position of their introns serine proteases have 40 Sumed specific cellular functions. A few of these enzymes been classified into a minimum of four groups including those have been isolated to sufficient purity to perform in vitro in which 1) the gene has no introns interrupting the exon studies, but most remain intractable due to the low abundance coding for the catalytic triad (e.g. the haptoglobin gene.); 2) of these molecules in the cell. To counter this technical diffi culty, a number of protein kinases have been isolated by each gene contains an intron just downstream from the codon molecular cloning strategies that utilize the conserved for the histidine residue at the active site, a second intron 45 sequences of the catalytic domain to isolate closely related downstream from the exon containing the aspartic acid resi homologs. Alternatively, some kinases have been purified due of the active site and a third intron just upstream from the (and Subsequently studied) based on their interactions with exon containing the serine of the active site (e.g. trypsinogen, other molecules. chymotrypsinogen, kallikrein and proelastase); 3) the genes p58 is a member of the p34.cdc2-related supergene family contain seven introns interrupting the exons coding the cata 50 lytic region (e.g. complement factor B gene); and 4) the genes and contains a large domain that is highly homologous to the cell division control kinase, cdc2. This new cell division contain two introns resulting in a large exon that contains both control-related protein kinase was originally identified as a the active site aspartatic acid and serine residues (e.g. factor component of semipurified galactosyltransferase; thus, it has X, factor 1x and protein C genes). been denoted galactosyltransferase-associated protein kinase Cytotoxic lymphocytes (e.g. CD8(+) cytotoxic T cells and 55 natural killer cells) form the major defense of higher organ (GTA-kinase). GTA-kinase has been found to be expressed in isms against virus-infected and transformed cells. A key func both adult and embryonic tissues and is known to phospho tion of these cells is to detect and eliminate potentially harm rylate a number of Substrates, including histone H1, and ful cells by inducing them to undergo apoptosis. This is casein. Interestingly enough, overexpression of this molecule achieved through two principal pathways, both of which 60 in CHO cells has shown that elevated levels of p58 result in a require direct but transient contact between the killer cell and prolonged late telophase and an early G1 phase, thus hinting its target. The first pathway involves ligation of TNF receptor of an important role for GTA-kinase in cell cycle regulation. like molecules such as Fas/CD95 to their cognate ligands, and Cyclin Dependent Kinases results in mobilization of conventional, programmed cell The cyclin-dependent kinases (Cdks) area family of highly death pathways centered on activation of pro-apoptotic 65 conserved serine/threonine kinases that mediate many of the caspases. The second mechanism consists of a pathway cell cycle transitions that occur during duplication. Each of whereby the toxic contents of a specialized class of secretory these Cak catalytic subunits associates with a specific Subset US 7,820,809 B2 65 66 of regulatory Subunits, termed cyclins, to produce a distinct and neurotransmitters. GPCRs play a central role in cell pro Cdk-cyclin kinase complex that, in general, functions to liferation, differentiation, and have been implicated in the execute a unique cell cycle event. etiology of disease. Activation of the Cdk-cyclin kinases during cellular tran The mechanism by which 6 protein-coupled receptors sitions is controlled by a variety of regulatory mechanisms. translate extracellular signals into cellular changes was ini tially envisioned as a simple linear model: activation of the For the Cdc2cyclin B complex, inhibition of kinase activity receptor by agonist binding leads to dissociation of the het during S phase and G is accomplished by phosphorylation of erotrimeric GTP-binding G protein (Gs, Gi, or Gq) into its two Cdc2 residues, Thr' and Tyr', which are positioned alpha and beta/gamma Subunits, both of which can activate or within the ATP-binding cleft. Phosphorylation of Thr'' and/ 10 inhibit various downstream effector molecules. More specifi or Tyr' suppresses the catalytic activity of the molecule by cally, activation of the GPCR induces a conformational disrupting the orientation of the ATP present within this cleft. change in the Goa Subunit, causing GDP to be released and In contrast, the abrupt dephosphorylation of these residues by GTP to be bound in its place. The Ga and Gfy subunits then the Cdc25 phosphatase results in the rapid activation of dissociate from the receptor and interact with a variety of Cdc2cyclin B kinase activity and Subsequent downstream 15 effector molecules. For instance in the case of the Gs family, mitotic events. While the exact details of this pathway have the primary function is to stimulate the intracellular messen yet to be elucidated, it has been proposed that Thr/Tyr' geradenylate cyclase (AC), which catalyzes the conversion of phosphorylation functions to permit a cell to attain a critical cytoplasmic ATP into the secondary messenger cyclic AMP concentration of inactive Cakcyclin complexes, which, upon (cAMP). In contrast, the Gi family inhibits this pathway and activation, induces a rapid and complete cell cycle transition the Gq family activates phospholipases C (PLC), which Furthermore, there is evidence in mammaliancells that Thr'? cleaves phosphatidylinositol 4.5, bisphosphate (PIP2) to gen Tyr' phosphorylation also functions to delay Cdk activation erate inositol-1,4,5-phosphate (IP3) and diacylglycerol after DNA damage. (DAG). The Schizosaccharomyces pombe weel gene product was More recently, studies have shown that the functions of 25 GPCRs are not limited to their actions on G-proteins and that the first kinase identified that is capable of phosphorylating considerable cross-talk exists between this diverse group of Tyr' in Cdc2. Homologs of the Weel kinase have been sub receptor molecules and a second class of membrane bound sequently identified and biochemically characterized from a proteins, the receptor tyrosine kinases (RTKs). A number of wide range of species including human, mouse, frog, Saccha GPCRs such as endothelin-1, thrombin, bombesin, and romyces cerevisiae, and Drosophila. In Vertebrate systems, 30 dopamine receptors can activate MAPKs, a downstream where Thr'' in Cdc2 is also phosphorylated, the Weel kinase effector of the RTK/Ras pathway. Interestingly, the interac was capable of phosphorylating Cdc2 on Tyr', but not Thr'', tion between these two families is not unidirectional and indicating that another kinase was responsible for Thr' 'phos RTKs can also modulate the activity of signaling pathways phorylation. This gene, Myt1 kinase, was recently isolated traditionally thought to be controlled exclusively by ligands from the membrane fractions of Xenopus egg extracts and has 35 that couple to GPCRs. For instance, EGF, which normally been shown to be capable of phosphorylating Thr'' and, to a activates the MAPK cascade via the EGF receptor can stimu lessor extent, Tyr' in Cdc2. Ahuman Myt1 homolog display late adenylate cyclase activity by activating GO.S. ing similar properties has been isolated, as well as a non There are dozens of members of the G Protein-Coupled membrane-associated molecule with Thr' kinase activity. Receptor family that have emerged as prominent drug targets In the past decade it has been shown that cancer can origi 40 in the last decade. One non-limiting list of potential GPCR nate from overexpression of positive regulators, such as siRNA targets is as follows: cyclins, or from underexpression of negative regulators (e.g. CMKLR1 p16 (INK4a), p.15 (INK4b), p21 (Cip1)). Inhibitors such as CML1/CMKLR1 (Accession No. Q99788) is a member of Myt1 are the focus of much cancer research because they are the chemokine receptor family of GPCRs that may play a role capable of controlling cell cycle proliferation, now consid 45 in a number of diseases including those involved in inflam ered the Holy Grail for cancer treatment. For these reasons, mation and immunological responses (e.g. asthma, arthritis). siRNA directed against kinases and kinase inhibitors includ For this reason, siRNA directed against this protein may ing but not limited to ABL1, ABL2, ACK1, ALK, AXL, BLK, prove to be important therapeutic reagents. BMX, BTK, C20orf64, CSF1R, SCK, DDR1, DDR2, Studies of juvenile-onset neuronal ceroid lipofuscinosis DKFZp761P1010, EGFR, EPHA1, EPHA2, EPHA3, 50 (JNCL, Batten disease), the most common form of childhood EPHA4, EPHA7, EPHA8, EPHB1, EPHB2, EPHB3, encephalopathy that is characterized by progressive neural EPHB4. EPHB6, ERBB2, ERBB3, ERBB4, FER, FES, degeneration, show that it is brought on by mutations in a FGFR1, FGFR2, FGFR3, FGFR4, FGR, FLT1, FLT3, FLT4, novel lysosomal membrane protein (CLN3). In addition to FRK, FYN, HCK, IGF1R, INSR, ITK, JAK1, JAK2, JAK3, being implicated in JNCL, CLN3 (GPCR-like protein, Acces KDR, KIAA1079, KIT, LCK, LTK, LYN, MATK, MERTK, 55 sion No. A57219) expression studies have shown that the MET, MSTIR, MUSK, NTRK1, NTRK2, NTRK3, PDG CLN3 mRNA and protein are highly over-expressed in a FRA, PDGFRB, PTK2, PTK2B, PTK6, PTK7, PTK9, number of cancers (e.g. glioblastomas, neuroblastomas, as PTK9L, RET, ROR1, ROR2, ROS1, RYK, SRC, SYK, TEC, well as cancers of the prostate, ovaries, breast, and colon) TEK, TIE, TNK1, TXK, TYK2, TYRO3, YES1, and related Suggesting a possible contribution of this gene to tumor proteins, may be useful for researchand therapeutic purposes. 60 growth. For this reason, siRNA directed against this protein G Protein Coupled Receptors may prove to be important therapeutic reagents. One important class of genes to which siRNAs can be CLACR directed are G-protein coupled receptors (GPCRs). GPCRs The calcitonin receptor (CTR/CALCR, Accession No. constitute a Superfamily of seven transmembrane spanning 65 NM 001742) belongs to “family B' of GPCRs which typi proteins that respond to a diverse array of sensory and chemi cally recognized regulatory peptides such as parathyroid hor cal stimuli. Such as light, odor, taste, pheromones, hormones mone, Secretin, glucagons and vasoactive intestinal polypep US 7,820,809 B2 67 68 tide. Although the CT receptor typically binds to calcitonin to known diseases. Pathway activation is initiated with the (CT), a 32 amino acid peptide hormone produced primarily secretion of Sonic hedgehog. There are three closely related by the thyroid, association of the receptor with RAMP (Re members of the Shh family (Sonic hedgehog, Desert, and ceptor Activity Modulating Protein) enables it to readily bind Indian) with Shh being the most widely expressed form of the other members of the calcitonin peptide family including 5 group. The Shh gene product is secreted as a small pro-signal amylin (AMY) and other CT gene-related peptides (e.g. molecule. To successfully initiate its developmental role, Shh. C.(CGRP and BCGRP). While the primary function of the is first cleaved, whereupon the N-terminal truncated fragment calcitonin receptor pertains to regulating osteoclast mediated is covalently modified with cholesterol. The addition of the sterol moiety promotes the interaction between Shh and its bone resorption and enhanced Ca" excretion by the kidney, cognate membrane bound receptor, Patched (Ptch). There are recent studies have shown that CT and CTRs may play an 10 at least two isoforms of the Patched gene, Ptch1 and Ptch2. important role in a variety of processes as wide ranging as Both isoforms contain a sterol-sensing domain (SSD); a embryonic/foetal development and sperm function/physiol roughly 180 amino acid cluster that is found in at least seven ogy. In addition, studies have shown that patients with par different classes of molecules including those involved in ticular CTR genotypes may beat higher risk to lose bone mass cholesterol biosynthesis, Vesicular traffic, signal transduc and that this GPCR may contribute to the formation of cal 15 tion, cholesterol transport, and sterol homeostasis. In the cium oxalate urinary stones. For this reason, siRNA directed absence of Shh, the Patched protein is a negative regulator of against CTR may be useful as therapeutic reagents. the pathway. In contrast, binding of Shih-cholesterol to the OXTR Patched receptor releases the negative inhibition which that The human oxytocin receptor (OTR, OXTR) is a 389 molecule enforces on a G-protein coupled receptor known as amino acid polypeptide that exhibits the seven transmem Smoothened. Subsequent activation of Smoothened (directly brane domainstructure and belongs to the Class-I (rhodopsin or indirectly) leads to the triggering of a trio of transcription type) family of G-protein coupled receptors. OTR is factors that belong to the Gli family. All three factors are expressed in a wide variety of tissues throughout develop relatively large, contain a characteristic C2-H2 zinc-finger ment and mediates physiological changes through G(q) pro pentamer, and recognize one of two consensus sequences teins and phospholipase C-beta. Studies on the functions of 25 (SEQ. ID NO 0463 GACCACCCA or SEQ. ID NO. 0464 oxytocin and the oxytocin receptor have revealed abroad list GAACCACCCA). In the absence of Shh, Gli proteins are of duties. OT and OTR play a role in a host of sexual, maternal cleaved by the proteosome and the C-terminally truncated and Social behaviors that include egg-laying, birth, milk fragment translocates to the nucleus and acts as a dominant letdown, feeding, grooming, memory and learning. In addi transcription repressor. In the presence of Shih-cholesterol, tion, it has been hypothesized that abnormalities in the func 30 Gli repressor formation is inhibited and full-length Gli func tionality of oxytocin-OTR receptor-ligand system can lead to tions as a transcriptional activator. a host of irregularities including compulsive behavior, eating Shh and other members of the Shh-PTCH-Gli pathway are disorders (such as anorexia), depression, and various forms of expressed in a broad range of tissues (e.g. the notochord, the neurodegenerative diseases. For these reasons, siRNA floorplate of the neural tube, the brain, and the gut) at early directed against this gene (NM 000916) may play an impor stages in development. Not Surprisingly, mutations that lead tant role in combating OTR-associated illnesses. 35 to altered protein expression or function have been shown to induce developmental abnormalities. Defects in the human EDG GPCRS Shh gene have been shown to cause holoprosencephaly, a Lysophosphatidic acid and other lipid-based hormones/ midline defect that manifests itself as cleft lip or palate, CNS growth factors induce their effects by activating signaling septation, and a wide range of other phenotypes. Interest pathways through the G-protein coupled receptors (GPCRs) 40 and have been observed to play important roles in a number of ingly, defects in cholesterol biosynthesis generate similar human diseases including cancer, asthma, and Vascular Shh-like disorders (e.g. Smith-Lemli-Opitz syndrome) sug pathologies. For instance, during studies of immunoglobulin gesting that cholesterol modification of the Shh gene product A nephropathy (IgAN), researchers have observed an is crucial for pathway function. Both the Patched and enhanced expression of EDG5 (NP 004221) suggesting a as Smoothened genes have also been shown to be clinically contribution of this gene product in the development of IgAN. relevant with Smoothened now being recognized as an onco For that reasons, siRNA directed against Edg5 (NM gene that, like PTCH-1 and PTCH-2, is believed to be the 0.04230), Edg4 (NM 004720), Edg7 (Nm 012152) and causative agent of several forms of adult tumors. For these related genes may play an important role in combating human reasons, siRNA directed against Smoothened (SMO, disease. 50 NM 005631), Patched (PTCH, nm 000264), and addi Genes Involved in Cholesterol Signaling and Biosynthesis tional genes that participate in cholesterol signaling, biosyn Studies on model genetic organisms such as Drosophila thesis, and degradation, have potentially useful research and and C. elegans have led to the identification of a plethora of therapeutic applications. genes that are essential for early development. Mutational Targeted Pathways. analysis and ectopic expression studies have allowed many of these genes to be grouped into discreet signal transduction 5 In addition to targeting siRNA against one or more mem pathways and have shown that these elements play critical bers of a family of proteins, siRNA can be directed against roles in pattern formation and cell differentiation. Disruption members of a pathway. Thus, for instance, siRNA can be of one or more of these genes during early stages of develop directed against members of a signal transduction pathway ment frequently leads to birth defects whereas as alteration of (e.g. the insulin pathway, including AKT1-3, CBL, CBLB. gene function at later stages in life can result in tumorigen 60 EIF4EBP1, FOXO1A, FOXO3A, FRAP1, GSK3A, GSK3B, CS1S. IGF1, IGF1R, INPP5D, INSR, IRS1, MLLT7, PDPK1, One critical set of interactions known to exist in both PIK3CA, PIK3CB, PIK3R1, PIK3R2, PPP2R2B, PTEN, invertebrates and vertebrates is the Sonic Hedgehog-Patched RPS6, RPS6KA1, RPX6KA3, SGK, TSC1, TSC2, AND Gli pathway. Originally documented as a Drosophila seg XPO1), an apoptotic pathway (CASP3,6,7,8,9, DSH1/2, mentation mutant, several labs have recently identified 65 P110, P85, PDK1/2, CATENIN, HSP90, CDC37, P23, BAD, human and mouse orthologs of many of the pathways mem BCLXL, BCL2, SMAC, and others), pathways, involved in bers and have successfully related disruptions in these genes DNA damage, cell cycle, and other physiological (p53, US 7,820,809 B2 69 70 MDM2, CHK1/2, BRCA1/2, ATM, ATR, P15INK4, P27, BRCA1/2, ATM, ATR, p15, P27, P21, Skp2, Cdc25C/A, P21, SKP2, CDC25C/A, 14-3-3, PLK, RB, CDK4, GLUT4, 14-3-3 Ofe, PLK, Rb, Cdk4, Gluta, iNOS, mTOR, FKBP. Inos, Mtor, FKBP. PPAR, RXR, ER). Similarly, genes PPARy, RXRcC. ERC. and related genes may play a critical involved in immune system function including TNFR1, IL role in combating diseases associated with disruptions in IR, IRAK1/2, TRAF2, TRAF6, TRADD, FADD, IKKe. 5 DNA repair, and cell cycle abnormalities. IKKy., IKKB, IKKC., IkBC, IkB,B, p50, p85, Rac, RhoA, Tables VI-Table X below provide examples of useful pools Cdc42, ROCK, Pak1/2/3/4/5/6, cIAP, HDAC1/2, CBP, for inhibiting different genes in the human insulin pathway B-TrCP. Rip2/4, and others are also important targets for the and tyrosine kinase pathways, proteins involved in the cell siRNAs described in this document and may be useful in cycle, the production of nuclear receptors, and other genes. treating immune system disorders. Genes involved in apop 10 These particular pools are particularly useful in humans, but tosis, such as Dsh1/2, PTEN, P110 (pan), P85, PDK1/2, Akt1, would be useful in any species that generates an appropriately Akt2, Akt (pan), p70', GSK3 B, PP2A (cat), B-catenin, homologous mRNA. Further, within each of the listed pools HSP90, Cdc27/p50, P23, Bad, BclxL, Bcl2, Smac/Diablo, any one sequence maybe used independently but preferably at and Ask1 are potentially useful in the treatment of diseases least two of the listed sequences, more preferably at least that involve defects in programmed cell death (e.g. cancer), 15 three, and most preferably all of the listed sequences for a while siRNA agents directed against p53, MDM2, Chk1/2, given gene is present.

TABLE WI

Gene SEQ. Name Acci GI L. L. Duplex H: Sequence ID NO.

AKT1. NM_005163 4.885. O6O 2 Of D-OO3 OOO-Os GACAAGGACGGGCACATTA 465

AKT1. NM_005163 4.885. O6O 2 Of D-OO3 OOO-O6 GGACAAGGACGGGCACATT 466

AKT1. NM_005163 4.885. O6O 2 Of D- OO3 OOO- Of GCTACTTCCTCCTCAAGAA 467

AKT1. NM_005163 4.885. O6O 2 Of D- OO3 OOO-08 GACCGCCTCTGCTTTGTCA, 468

AKT2

AKT2 NM_001626 6715585. 208 D- OO3 OO1 - OS GTACTTCGATGATGAATTT 469

AKT2 NM_001626 6715585. 208 D-OO3 OO1 - O6 GCAAAGAGGGCATCAGTGA 470

AKT2 NM_001626 6715585. 208 D-OO3 OO1 - Of GGGCTAAAGTGACCATGAA 471

AKT2 NM_001626 6715585. 208 D-OO3 OO1 - 08 GCAGAATGCCAGCTGATGA 472

AKT3

AKT3 NM_005465 323 Of1641 OOOO D- OO3OO2- 05 GGAGTAAACTGGCAAGATG 473

AKT3 NM_005465 3.23. Of 1641 OOOO D- OO3 OO2- O6 GACATTAAATTTCCTCGAA 474

AKT3 NM_005465 323 Of1641 OOOO D- OO3OO2- Of GACCAAAGCCAAACACATT 475

AKT3 NM_005465 323 Of1641 OOOO D- OO3OO2- 08 GAGGAGAGAATGAATTGTA 476

CBL

CBL NM_005188 4.885.116 867 D- OO3 OO3- OS GGAGACACATTTCGGATTA 477

CBL NM_005188 4.885.116 867 D- OO3 OO3- 06 GATCTGACCTGCAATGATT 478

CBL NM_005188 4.885.116 867 D- OO3 OO3- Of GACAATCCCTCACAATAAA 479

CBL NM_005188 4.885.116 867 D- OO3 OO3- 08 CCAGAAAGCTTTGGTCATT 48O

CBLB

CBLB NM 170662 293668Of 86.8 D-OO3 OO4 - 05 GACCATACCTCATAACAAG 481

CBLB NM 170662 293668Of 86.8 D-OO3 OO4 - 06 TGAAAGACCTCCACCAATC 482

CBLB NM 170662 293668Of 86.8 D- OO3 OO4- Of GATGAAGGCTCCAGGTGTT 483

CBLB NM 170662 293668Of 86.8 D- OO3 OO4 - 08 TATCAGCATTTACGACTTA 484

EIF4EBP1

EIF4EBP1 NM_004095 2007 O179 1978 D- OO3OOs-Os GCAATAGCCCAGAAGATAA 485 EIF4EBP1 NM_004095 2007 O179 1978 D- OO3OO5 - 06 CGCAATAGCCCAGAAGATA 486