Vol. 51 No. 2/2004

431–444

QUARTERLY

Review

Targeting BACE with small inhibitory nucleic acids — a future for Alzheimer’s disease therapy?*.

Barbara Nawrot½

Centre of Molecular and Macromolecular Studies of Polish Academy of Sciences, Łódź, Poland Received: 03 May, 2004; accepted: 14 May, 2004

Key words: b-secretase, b-amyloid peptide, Alzheimer’s disease, hammerhead , , small interfering

b-Secretase, a b-site amyloid precursor protein (APP) cleaving (BACE), par- ticipates in the secretion of b-amyloid peptides (Ab), the major components of the toxic amyloid plaques found in the of patients with Alzheimer’s disease (AD). According to the amyloid hypothesis, accumulation of Ab is the primary influence driving AD pathogenesis. Lowering of Ab secretion can be achieved by decreasing BACE activity rather than by down-regulation of the APP substrate protein. There- fore, b-secretase is a primary target for anti-amyloid therapeutic drug design. Se- veral approaches have been undertaken to find an effective inhibitor of human b-secretase activity, mostly in the field of peptidomimetic, non-cleavable substrate analogues. This review describes strategies targeting BACE mRNA recognition and its down-regulation based on the antisense action of small inhibitory nucleic acids (siNAs). These include antisense , catalytic nucleic acids — ribo- zymes and deoxyribozymes — as well as small interfering RNAs (siRNAs). While antisense oligonucleotides were first used to identify an aspartyl protease with b-secretase activity, all the strategies now demonstrate that siNAs are able to inhibit

*Presented as invited lecture at the 29th Congress of the Federation of European Biochemical Societies, Warsaw, Poland, 26 June–1 July 2004. .This work was supported by the State Committee for Scientific Research (project PBZ-KBN- 059/T09/09 to B.N.). ½ Correspondence to: Barbara Nawrot, Centre of Molecular and Macromolecular Studies of Polish Acad- emy of Sciences, Sienkiewicza 112, 90-363 Łódź, Poland; phone: (48 42) 681 6970; fax: (48 42) 681 5483; e-mail: [email protected] Abbreviations:Ab, b-amyloid peptide; AD, Alzheimer’s disease; APP, amyloid precursor protein; CAD cells, catecholaminergic cell line; CNS, ; CTE, constitutive transport element; miRNA, micro-RNA; RISC, RNA induced silencing complex; RNAi, RNA interference; shRNA, short hairpin RNA; siNAs, small inhibitory nucleic acids; siRNAs, small interfering RNAs. 432 B. Nawrot 2004

BACE in a sequence-specific manner, measured both at the level of its mRNA and at the level of protein. Moreover, knock-down of BACE reduces the intra- and extracellular population of Ab40 and Ab42 peptides. An anti-amyloid effect of siNAs is observed in a wide spectrum of cell lines as well as in primary cortical neurons. Thus targeting BACE with small inhibitory nucleic acids may be beneficial for the treatment of Alzheimer’s disease and for future drug design.

BACKGROUND vaccine AN-1792, specific antibodies against AD, and the antibiotic ciloquinol) there is still Alzheimer’s disease (AD) is the most com- a need for development of new and more ef- mon cause of dementia in the increasing pop- fective treatments of this dementia (Helmut, ulation of elderly people. According to the 2000; Evans, 2001). World Health Organization about 5.5% of peo- Alzheimer’s disease is a neurodegenerative ple who live beyond 60 suffer from some form disorder characterized by progressive deposi- of neurodegeneration. Up to now there are no tion of senile plaques in the of patients drugs available that specifically and on a mo- with ageing dementia. The major component lecular level cure the cause of Alzheimer’s dis- of these aggregates is a 4 kDa b-amyloid pep- ease. Despite many novel therapies investi- tide (Ab), a product of proteolytic cleavage of gated in the last decade (donepezil, tacrine, the amyloid precursor protein (APP) (Selkoe,

Figure 1. Processing of amyloid precursor protein (APP). APP proteolysis occurs in vivo by two different ways. Non-amyloidogenic pathway includes hydrolysis catalysed by a-secretase, resulting in release of soluble a-APPs, and the intra-membrane proteolysis catalysed by g-secretase se- creting peptides P3 and P7. The second, amyloidogenic pathway involves b-secretase activity which releases b-APPs, and, after cleveage of the C-terminal fragment by g-secretase, peptide P7 and Ab peptides, ranging in length from 39 to 43 amino acids, are secreted. Ab42 peptide is prone to aggregation and formation of amyloid plaques, accu- mulation of which is the primary factor driving AD pathogenesis. Vol. 51 Small inhibitory nucleic acids for BACE 433

1991). The processing of this transmembrane (Asp2), also called b-site APP cleaving en- glycoprotein occurs in vivo by two different zyme (BACE) or memapsin 2. This 501-amino pathways (Fig. 1). acid transmembrane protein contains two ac- The conventional non-amyloidogenic path- tive site motifs at amino acids 93-96 and way proceeds via proteolytic cleavage of APP 289-292 (Fig. 2). by a- and g-secretases and results in release of These are highly conserved signature se- non-toxic, soluble a-APPs protein and two quences of aspartyl proteases, DTGS and other shorter products P3 and P7 (Haass & DSGT, respectively. The active sites of the Selkoe, 1993; Hendriks & Van Broeckhoven, protein are situated within the lumenal do- 1996). In normal, healthy individuals these main at the right orientation to attack the products protect neuronal cells against oxida- b-site of APP. The transmembrane domain is tive stress and participate in wound repair located between amino acids 461 and 477. (Mattson et al., 1993; Barger et al., 1995; The remaining 24 amino acids of the C-termi- Mattson & Furukawa, 1997; Kummer et al., nal part of the protein are located in a cytosol. 2002). Another APP processing pathway is BACE contains four N-glycosylation sites and amyloidogenic. APP is cleaved at the N-termi- six cysteine residues forming intramolecular nus of the Ab region by b-secretase and at the disulfide bonds. The optimum activity of C-terminus by g-secretase. The cleavage prod- BACE is at pH 5.5. b-Secretase catalyses hy- ucts are b-APPs protein, C-terminal peptide drolysis of the peptide bond of APP according P7 and 39-43 amino acid long beta-amyloid to the general mechanism known for aspartyl peptide, of which the major products are proteases (Schmidt, 2003). Two aspartic acid Ab40 and Ab42. The predominant form of Ab residues are involved in catalysis; one of them found in the cerebrospinal fluid is the shorter must be negatively charged. These residues Ab40. However, the longer Ab42 peptide, work as acceptors and donors of protons and which is released at a much slower rate and activate water molecule participating in the thus appears in a much lower concentration, peptide bond hydrolysis. is more prone to aggregation (Jarrett et al., Shortly after BACE was identified, another 1993). According to the amyloid hypothesis, aspartyl protease, cleaving the APP protein accumulation of Ab is the primary factor driv- within the b-amyloid sequence, was sequenced ing AD pathogenesis (Hardy & Allsop, 1991; (Farzan, 2000). This protein is called BACE2 Hardy & Selkoe, 2002). Lowering of Ab secre- (memapsin 1, Asp1) and thus BACE was re- tion can be achieved by decreasing b-secre- named BACE1. BACE1 has 64% sequence tase activity rather than by down-regulation identity with BACE2. Both homologues co-ex- of APP substrate protein or g-secretase inhibi- press in a variety of cells. BACE1 is present in tion. Therefore, b-secretase is a primary tar- all tissues with the highest level of expression get for anti-amyloid therapeutic drug design in the brain (Sinha et al., 1999; Vassar, 1999; (Citron, 2002). Lin et al., 2000). BACE2 is mostly expressed in astrocytes and is not detected in neurons and microglia, where BACE1 is exclusively present CHARACTERISTICS OF (Basi et al., 2003). BETA-SECRETASE

The gene encoding b-secretase was se- BACE1 AS A THERAPEUTIC TARGET quenced by five independent research groups (Vassar et al., 1999; Hussain et al., 1999; Yan It is a widely accepted opinion that inhibi- et al., 1999; Sinha et al., 1999; Lin et al., tion of Ab generation by lowering the activity 2000). b-Secretase is an aspartyl protease 2 of b-secretase may be beneficial for AD treat- 434 B. Nawrot 2004 ment (Citron, 2002a; 2002b). This idea is b-secretase and b-APP cleavage (Simons & strongly supported by Roberds et al. (2001) Ehehalt, 2002). Therefore, b-secretase is an who have shown that BACE knockout mice excellent target for anti-amyloid therapeutic are healthy and show no phenotypic differ- drug design (Citron, 2002a; 2002b). Several ences from their wild-type littermates. Corti- approaches were immediately undertaken in order to find an effective inhibitor of human b-secretase activity (Ghosh et al., 2000; 2001; Turner et al., 2001; Brady et al., 2004). The main strategy for designing BACE inhibitors utilized peptide-based analogues of the sub- strate. However, due to the 30% sequence identity of BACE1 with other aspartyl pro- N N teases the selectivity of BACE peptidomi- metic inhibitors is not satisfactory (Ghosh et N al., 2000). Several non-peptidomimetic inhibi- tors have been found which effectively block N BACE activity (Miyamoto et al., 2001; Qiao & Etcheberrigaray, 2002; Bhisetti et al., 2002). Two crystal structures of BACE1 complexed with its inhibitors OM99-2 and OM00-3 were solved by Hong et al. (2000; 2002). The struc- ture of an active site and APP binding domain is helpful in designing low molecular mass in- hibitors for BACE, especially because it re- vealed that the active site of BACE1 is more open and less hydrophobic than typical sites of other aspartyl proteases. The progress in Figure 2. Schematic diagram of aspartyl protease the search of selective b-secretase inhibitors 2 (Asp2, BACE1). has been recently summarized in several re- BACE1 is a 501 amino-acid transmembrane protein. view papers (Ghosh et al., 2002; Roggo, 2002; Signal peptide (1–21 aa) and pro-peptide (22–45 aa) Schmidt, 2003; John et al., 2003). are cleaved during protein maturation. Within the lumenal domain there are two active site motifs (signa- ture sequences of aspartyl proteases) situated at amino acids 93–96 and 289–292, four N-glycosylation sites STRATEGIES BASED ON DOWN- and six cysteine residues involved in intramolecular REGULATION OF GENE EXPRESSION disulfide bonds. The transmembrane domain is located BY SMALL INHIBITORY NUCLEIC between amino acids 461 and 477, and the C-terminal ACIDS part of the protein (24 aa) is located in a cytosol. Strategies targeting mRNA recognition and cal neurons from such mice showed no detect- its down-regulation are based on the anti- able b-secretase activity and remarkably low sense action of small inhibitory nucleic acids levels of Ab peptide. Moreover, it was shown (siNAs) including antisense oligonucleotides, recently that elevated Ab production in spo- catalytic nucleic acids — ribozymes and radic Alzheimer’s disease patients is corre- deoxyribozymes as well as small interfering lated with increased b-secretase activity (Li et RNAs (siRNAs) (Fig. 3). All of these nucleic al., 2004). The level of cholesterol and acids recognize the target molecule via se- rafts is probably crucial for activation of quence-specific Watson–Crick base- pairing Vol. 51 Small inhibitory nucleic acids for BACE 435 and lead to the formation of a complementary winding the intramolecular helices of the siNA/mRNA duplex. The mechanism of ac- messenger RNA, facilitates the high effi- tion of particular siNAs is different. Anti- ciency of ribozymes in in vivo experiments, in- sense oligonucleotides, when bound to the dependently of the secondary structure of the target molecule form DNA/mRNA duplexes target mRNA (Warashina et al., 2001; block the by “hybridization ar- Kawasaki et al., 2004).

Figure 3. Strategies for down-regulation of gene expression by small inhibitory nucleic acids (antisense oligonucleotides, catalytic nucleic acids and siRNA). All presented strategies concern mRNA degradation on post-transcriptional level, based on a sequence-specific tar- get molecule recognition. Antisense strategy involves RNase H activation by DNA/RNA hybrid and cleavage of RNA within the sequence of the duplex. Ribozymes and deoxyribozymes exhibit catalytic activity for RNA hydroly- sis. Short interfering RNAs (siRNAs) induce via activation of RISC — a protein complex with helicase/ activity. rest” or activate RNase H and lead to RNA Antisense, and degradation (Stein & Cheng, 1993). Ribozy- strategies are widely used to design nucleic mes and deoxyribozymes facilitate the cleav- acids for therapeutic applications (Christof- age of RNA phosphodiester bonds via a cata- fersen & Marr, 1995; Lewin & Hauswirth, lytic mechanism (Emilsson et al., 2003). Up to 2001; Opalinska & Gewirtz, 2003). In recent now, it has been demonstrated that years various antisense oligonucleotides and -coded ribozymes, coupled on their ribozymes have been the subjects of many 5¢-ends with a tRNAVal sequence, express pre-clinical and clinical trials (Kurreck, 2003), intracellularly with high efficiency (Koseki et including ex vivo treatment of HIV-1 infected al., 1999). Such ribozymes, when mimicking patients (Sullenger & Gilboa, 2002). More- the 3¢-immature tRNA molecule, are recog- over, the first antisense Vomi- nized by the nuclear protein exportin-t (Xpo-t) virsen was successfully introduced for the and effectively exported from the nucleus to treatment of AIDS patients infected with the cytoplasm (Kuwabara et al., 2001). Conju- cytomegalovirus (Highleyman, 1998). Nucleic gation of the 3¢-end of the ribozymes with an acid technology has been also considered as a aptameric sequence for cellular helicases, un- possible therapeutic approach for treatment 436 B. Nawrot 2004 of disorders of the central nervous system sion when siRNAs were used at high concen- (CNS). The latest achievements in identifying trations (>100 nM) (Bridge et al., 2003; Jack- target genes and the use of inhibition ap- son et al., 2003; Persengiev et al., 2004). How- proaches have been summarised recently by ever, these off-target effects can be excluded Trülzsch & Wood (2004). by selection of the best target sites and the Tuschl’s discovery in 2001 that chemically lowering the amount of siRNAs applied. synthesized short RNA duplexes (small inter- Accessibility of the target mRNA sequences fering RNAs, siRNAs) exhibit a potency for to siNA molecules is one of the limiting fac- sequence-specific cognate gene silencing in tors for efficient down-regulation of gene ex- mammalian cells via an RNAi pathway pression. Even though RNA helicases partici- (Elbashir et al., 2001a) opened a Pandora’s pate in the siRNA-induced gene silencing, box. The search for siRNA sequences able to none of small inhibitory nucleic acids have in- specifically and efficiently down-regulate ternal potency for unwinding mRNA intra- gene expression is of interest from several molecular helices, which would facilitate ac- points of view. Besides their relevance for ge- cessibility to complementary target se- notype/phenotype relationship investigations quences. Some effort has been done to design siRNAs have been immediately considered as helicase-attached hybrid ribozymes which are molecular tools for target identification and active in cellular systems independently of implemented for therapeutic application the secondary structure of the target mRNA (Dorsett & Tuschl, 2004). The mechanism of (Warashina et al., 2001; Kawasaki et al., RNA interference involves activation of the 2004). Besides, screening techniques are used nuclease/helicase protein complex RISC to find active antisense, ribozyme and siRNA (RNA induced silencing complex) and guiding sequences that are able to suppress, to a it to the complementary mRNA via an reasonable extent, the expression of the antisense strand of siRNA. The nucleolytic ac- cognate gene. tivity of the RISC complex leads to cleavage of the target mRNA and, in consequence to the inhibition of protein biosynthesis. The ef- DOWN-REGULATION OF BACE GENE fect of gene silencing can last for several days EXPRESSION if siRNA molecules are generated endoge- nously from coding siRNAs or short Approaches used to inhibit BACE gene ex- hairpin RNAs (Miyagishi & Taira, 2002; pression include applications of antisense Paddison et al., 2002; Paul et al., 2002; oligonucleotides, catalytic nucleic acids and Brummelkamp et al., 2002; Paddison et al., small interfering RNAs. Although each of 2002), the substrates of the these strategies works via antisense recogni- (Bernstein et al., 2001). Interestingly, tion of the target RNA and uses different in mammalian primary neurons the effect of mechanisms of action, all of them are de- gene silencing induced by small interfering signed to down-regulate gene expression at RNAs can last for at least three weeks (Omi et the post-transcriptional stage. The secondary al., 2004). It has been postulated that siRNA structure of the target BACE mRNA molecule duplexes that are not fully complementary to (accession number AF190725, start codon the cognate mRNA can induce repression of 5¢-AUG456 –3¢) generated with the help of the protein translation using the same RISC com- MFOLD program (Zuker, 2003) is shown at plex and micro-RNA mechanism (Schwarz et Fig. 4. al., 2003). An antisense approach was used by Yan et Recently, severe non-specific effects have al. (1999) to identify an aspartyl protease been observed in the profile of gene expres- with the b-secretase activity. Experiments Vol. 51 Small inhibitory nucleic acids for BACE 437 were done in HEK293 cells overexpressing BACE is involved in Ab secretion in both so- APP695 protein with the Swedish mutation matic and neuronal cells. KM ÷ NL, (which makes the substrate pro- Similarly, Vassar et al. (1999) used a set of tein more prone to b-cleavage), and with two antisense oligonucleotides to establish the ne-

Figure 4. Secondary structure of b-site APP cleaving enzyme (BACE1) mRNA (accession number AF190725, start codon 5¢-AUG456–3’) generated with the MFOLD program (Zuker, 2003). lysines at the C terminus (HEK293sw-KK cessity of BACE for the APPsw cleavage at cells). Two of the antisense oligomers di- the b-site. The 25 nt long antisense oligo- rected toward aspartyl protease mRNA signi- (entry 7–9, Table 1) reduced the ficantly reduced the BACE message level (Ta- BACE mRNA level by up to about 25–30%; ble 1). The amounts of Ab peptides 1–40 and while extracellular Ab40 and Ab42 peptides 1–42 released into the culture medium were were secreted in about 60–70% yield in com- also reduced by 50–80%. Similar BACE gene parison to the control oligomer-treated cells. down-regulation and lowering of Ab produc- An approach based on RNA-cleaving tion were observed in human neuroblastoma ribozymes was developed to control expres- IMR-32 and murine Neuro-2A cells. These ex- sion of b-secretase (Nawrot et al., 2002; 2003). periments proved that aspartyl protease Two sites of BACE mRNA were chosen as tar- 438 B. Nawrot 2004 get sequences for endogenously generated (Table 1, entries 10 and 11). A similar ribozymes. The ribozyme cassette was de- anti-amyloid effect was observed in experi- signed to constitute a catalytic hammerhead ments of BACE down-regulation performed core and substrate recognition arms, flanked on a SH-SY5Y human neuroblastoma cell line at the 5¢-terminus with tRNAVal and at the (Nawrot et al., unpublished). Therefore, such 3¢-terminus with CTE (constitutive transport ribozymes, especially when coded in viral vec- element) sequences. Ribozyme cassettes were tors, may be considered as inhibitors of cloned into a pUC19 plasmid and used for b-secretase activity, and further, as diagnos-

Table 1. Approaches for inhibition of BACE gene expression in cultured cell lines (BACE mRNA – accession number AF190725).

a b c BACE accession number NM_012104; as the amount of recombinant BACE protein; primary cortical neurons of wild d type mice; primary cortical neurons of the APPsw transgenic mice. transient of HEK293 cells. Such tic and therapeutic agents for AD treatment. ribozymes efficiently inhibit BACE gene ex- Several DNAzymes (deoxyribozymes) pression at the level of mRNA (up to 95%) and 10–23, complementary to BACE mRNA, were at the level of protein (up to 90%). Inhibition screened for their cleavage efficiency. A few of BACE activity directly influences the intra- of the molecules were active in the cellular and extracellular population of Ab40 peptide test system (HEK293 cells), reducing BACE Vol. 51 Small inhibitory nucleic acids for BACE 439 gene expression by 50–70% (Nawrot et al., un- duced the amount of overexpressed BACE1 published). Further experiments are neces- protein (>90%). siRNA targeting BACE1 led sary to demonstrate the influence of the ac- to a decrease in secreted sAPPb as well as Ab tive deoxyribozymes at the level of the BACE peptide, consistent with earlier cellular and in protein and the released b-amyloid peptides. vivo studies (Roberds et al., 2001; Cai et al., Small molecules, such as short 2001; Luo et al., 2001). Moreover, siRNA to interfering nucleic acids (called also siNA), BACE1 did not affect expression of BACE2 short interfering RNA (siRNA), dou- and vice versa. Silenced BACE1 and non-al- ble-stranded RNA (dsRNA), micro-RNA tered BACE2 gene expression was beneficial (miRNA) and short hairpin RNA (shRNA), ca- for the reduction of Ab, proving the antago- pable of mediating RNA interference against nistic activities of both proteases (Basi et al., BACE gene expression, were patented by 2003). McSwinggen & Beigelman (2003) as an ap- Although there is 64% homology between proach to the treatment of Alzheimer’s dis- the aspartyl proteases BACE1 and BACE2, ease. The patent also covers chemically modi- Kao et al. (2004) designed four siRNA se- fied siRNAs, comprising e.g. 2¢-deoxy- quences targeted to BACE1 which specifically 2¢-fluoro or 2¢-amino pyrimidine and purine and selectively suppressed BACE1 and not nucleotides, in which the was ad- BACE2 expression in various cell systems ditionally modified with 5¢- and 3¢-terminal (entry 22–25, Table 1). First, the activity of inverted deoxyabasic caps and the antisense the siRNAs was tested in a catecholaminergic strand comprised a 3¢-phosphorothioate cell line (CAD cells) overexpressing APP and internucleotide linkage. Four siRNAs di- BACE1 proteins. Two of the selected siRNAs rected toward BACE mRNA sequences are il- demonstrated >90% efficacy on suppressing lustrated (entry 14–17, Table 1). The siRNAs, recombinant BACE1, and the others (entry containing 19 nt non-modified RNA duplexes 23 and 25) down-regulated the gene of BACE 3¢-terminated with double thymidine over- by 70 and 50%, respectively. Similar effects of hangs, were used for lipid-assisted trans- BACE1 silencing were observed in this sys- fection of human lung cancer A549 cells. The tem when siRNAs were generated from short siRNAs, at a final concentration of 25 nM, hairpin RNAs coded in a pSilencer vector. proved to be very efficient down-regulation Secondly, it was also demonstrated that agents, reducing the level of BACE messen- siRNAs directed toward sequences 22 and 24 ger RNA by up to 90% of that in control cells (Table 1) significantly reduced the level of en- transfected only with . dogenously expressed BACE1 in HEK293 Small interfering RNAs directed toward en- cells (Kao et al., 2004). Furthermore, selected dogenous and plasmid expressed BACE1 and siRNA molecules reduced Ab production in BACE2 genes were used by Basi et al. (2003) primary cortical neurons derived from both to investigate the role of endogenous BACE2 wild type and APPsw transgenic mice. While protein in the production of Ab peptide. To in- siRNAs used in all test systems significantly duce BACE1 gene silencing by RNA interfer- reduced the secretion of Ab40 and Ab42, ence, three synthetic siRNAs, designed as re- there were no changes in APP and presenilin commended (Elbashir et al., 2001b), were 1 (PS1) protein levels and their subcellular tested in HEK293 cells stably transfected distribution. Importantly, pre-treating neu- with wild type APP or APPsw genes. The rons with BACE1 siRNAs reduced H2O2 in- siRNA, directed toward sequence 709–728 nt duced cell death, and thus the neurotoxicity of BACE1 mRNA, suppressed endogenous of Ab caused by oxidative stress could be sig- message level up to 36%, and specifically re- nificantly decreased. 440 B. Nawrot 2004

BACE KNOCKOUT: HOPES AND rected toward the CNS in in vivo experiments HYPES (Pardridge, 2002; Davidson & Breakefield, 2003; Trülzsch & Wood, 2004) keep that ther- As has been shown in cellular systems and apeutic strategy still promising. in in vivo experiments, BACE1 mRNA knock- In summary, the sequence-specific down- out is directly linked with protein expression, regulation of BACE1 results in lowering of as well as with APP metabolite production. b-secretase activity and decreasing of Ab for- The down-regulation of BACE1 by small in- mation. Thus, small inhibitory nucleic acids, hibitory nucleic acids specifically impacts on including antisense oligonucleotides, cata- the b-cleavage of APP and Ab production in lytic nucleic acids and small interfering RNAs both, somatic and neuronal cells. Although can be considered as future anti-amyloid the use of siNA may constitute a potential agents useful in designing novel Alzheimer’s anti-amyloidal approach to the treatment of disease therapies. Alzheimer’s disease there are several disad- vantages for BACE1 targeted therapies. b-Secretase, like the other transmembrane The author expresses the words of gratitude proteases, may have multiple substrates. Two to Professor Wojciech J. Stec for scientific in- of them, an a-2,6-sialyltransferase (ST6Gal I) spiration, continuous support, and critical and P-selectin glycoprotein ligand-1 (PSGL-1) reading of this manuscript. have been identified by Kitazume et al. (2001) and Lichtenthaler et al. (2003), respectively. These findings indicate that only partial REFERENCES BACE1 down-regulation would be beneficial to other BACE1 substrates. BACE1 exhibits Banks WA, Farr SA, Butt W, Kumar VB, >30% sequence homology with other aspartyl Franko MW, Morley JE. (2001) Delivery proteases of the pepsin family, and about 60% across the blood-brain barrier of antisense di- b sequence identity with its BACE2 homologue. rected against amyloid : reversal of learn- ing and memory deficits in mice BACE1 inhibitors cross-interacting with overexpressing amyloid precursor protein. J other aspartyl proteases have to be excluded Pharmacol Exp Ther.; 297: 1113–21. as AD therapeutics. Besides specificity, the delivery of small inhibitory nucleic acids to Barger SW, Fiscus RR, Ruth P, Hofmann F, tissues and primary neurons of the central Mattson MP. (1995) Role of cyclic GMP in nervous system (CNS) hampers progress in the regulation of neuronal calcium and sur- vival by secreted forms of b-amyloid precur- the design of novel oligonucleotide therapeu- sor. J Neurochem.; 64: 2087–96. tics. However, it has been demonstrated that in- Basi G, Frigon N, Barbour R, Doan T, Gordon tact antisense oligonucleotides crossed the G, McConlogue L, Sinha S, Zeller M. (2003) blood-brain barrier (BBB) in mice over- Antagonistic effects of BACE1 and BACE2 on Ab production in cells. J Biol Chem.; expressing APP (Banks et al., 2001). The re- 278: 31512–20. cent report that oligonucleotides encapsu- lated with nanogels (polyethylene glycol and Bernstein E, Caudy AA, Hammond SM, polyethyleneimine) can effectively cross the Hannon GJ. (2001) Role for a bidentate BBB both in cellular and in mouse models ribonuclease in the initiation step of RNA in- terference. Nature.; 409: 363–6. (Vinogradov et al., 2004) kindled some hopes that this obstacle can be overcome. Also the Bhisetti GR, Saunders JO, Murcko MA, Lepre recent reports that virus-based vectors are ef- CA, Britt SD, Come JH, Deninger DD, Wang fective carriers of gene/siNA therapy di- T. (2002) Inhibitors of BACE. Patent WO 02/88101. Vol. 51 Small inhibitory nucleic acids for BACE 441

Brady SF, Singh S, Crouthamel MC, Holloway Emilsson GM, Nakamura S, Roth A, Breaker MK, Coburn CA, Garsky VM, Bogusky M, RR. (2003) Ribozyme speed limits. RNA.; 9: Pennington MW, Vacca JP, Hazuda D, Lai 907–18. MT. (2004) Rational design and synthesis of Evans J. (2001) New paths to Alzheimer’s selective BACE-1 inhibitors. Bioorg Med drugs. Chemistry in Britain.; April, 47–51. Chem Lett.; 14: 601–4. Farzan M, Schnitzler CE, Vasilieva N, Leung D, Bridge AJ, Pebernard S, Ducraux A, Nicoulaz Choe H. (2000) BACE2, a b-secretase AL, Iggo R. (2003) Induction of an inter- homolog, cleaves at the b site and within the feron response by RNAi vectors in mamma- amyloid-beta region of the amyloid-b precur- lian cells. Nat Genet.; 34: 263–4. sor protein. Proc Natl Acad Sci USA.; 97: Brummelkamp TR, Bernards R, Agami R. 9712–7. (2002) A system for stable expression of Ghosh AK, Shin D, Downs D, Koelsch G, Lin X, short interfering RNAs in mammalian cells. Ermolieff J, Tang J. (2000) Design of potent Science.; 296: 550–3. inhibitors for human brain memapsin 2 Cai H, Wang Y, McCarthy D, Wen H, Borchelt (b-secretase). J Am Chem Soc.; 122: 3522–3. DR, Price DL, Wong PC. (2001) BACE1 is Ghosh AK, Bilcer G, Harwood C, Kawahama R, the major b-secretase for generation of Ab Shin D, Hussain KA, Hong L, Loy JA, peptides by neurons. Nat Neurosci.; 4: Nguyen C, Koelsch G, Ermolieff J, Tang J. 233–4. (2001) Structure-based design: potent inhibi- Christoffersen RE, Marr JJ. (1995) Ribozymes tors of human brain memapsin 2 (b-se- as human therapeutic agents. J Med Chem.; cretase). J Med Chem.; 44: 2865–8. 38: 2023–37. Ghosh AK, Hong L, Tang J. (2002) Beta-secre- Citron M. (2002a) Beta-secretase as a target for tase as a therapeutic target for inhibitor the treatment of Alzheimer’s disease. J drugs. Curr Med Chem.; 9: 1135–44. Neurosci Res.; 70: 373–9. Haass C, Selkoe DJ. (1993) Cellular processing Citron M. (2002b) Emerging Alzheimer’s disease of b-amyloid precursor protein and the gene- therapies: inhibition of b-secretase. Neurobiol sis of amyloid b-peptide. Cell.; 75: 1039–42. Aging.; 23: 1017–22. Hardy J, Allsop D. (1991) Amyloid deposition as Davidson BL, Breakefield XO. (2003) Viral vec- the central event in the aetiology of Alzhei- tors for gene delivery to the nervous system. mer’s disease. Trends Pharmacol Sci.; 12: Nat Rev Neurosci.; 4: 353–64. 383–8. Dorsett Y, Tuschl T. (2004) siRNAs: applica- Hardy J, Selkoe DJ. (2002) The amyloid hypoth- tions in functional genomics and potential as esis of Alzheimer’s disease: progress and therapeutics. Nat Rev Drug Discov.; 3: problems on the road to therapeutics. Sci- 318–29. ence.; 297: 2209. Elbashir SM, Harborth J, Lendeckel W, Yalcin Helmut L. (2000) Antibiotic to treat Alzhei- A, Weber K, Tuschl T. (2001a) Duplexes of mer’s? Science.; 290: 1273–4. 21- RNAs mediate RNA interfer- Hendriks L, Van Broeckhoven Ch. (1996) A ence in cultured mammalian cells. Nature.; beta A4 amyloid precursor protein gene and 411: 494–8. Alzheimer’s disease. Eur J Biochem.; 237: Elbashir SM, Martinez J, Patkaniowska A, 6–15. Lendeckel W, Tuschl T. (2001b) Functional Highleyman L. (1998) FDA approves anatomy of siRNAs for mediating efficient fomivirsen, famciclovir, and Thalidomide. RNAi in Drosophila melanogaster embryo Food and Drug Administration. BETA. Oct.; lysate. EMBO J.; 20: 6877–88. 5. 442 B. Nawrot 2004

Hong L, Koelsch G, Lin X, Wu S, Terzyan S, b-secretase, b-site amyloid precursor pro- Ghosh AK, Zhang XC, Tang J. (2000) Struc- tein-cleaving enzyme, is responsible for cleav- ture of the protease domain of memapsin 2 age of a Golgi-resident sialyltransferase. (b-secretase) complexed with inhibitor. Sci- Proc Natl Acad Sci USA.; 98: 13554–9. ence.; 290: 150–3. Koseki S, Tanabe T, Tani K, Asano S, Shioda T, Hong L, Turner RT 3rd, Koelsch G, Shin D, Nagai Y, Shimada T, Ohkawa J, Taira K. Ghosh AK, Tang J. (2002) Crystal structure (1999) Factors governing the activity in vivo of memapsin 2 (b-secretase) in complex with of ribozymes transcribed by RNA polymer- an inhibitor OM00-3. Biochemistry.; 41: ase III. J Virology.; 73: 1868–77. 10963–7. Kummer C, Wehner S, Quast T, Werner S, Hussain I, Powell D, Howlett DR, Tew DG, Herzog V. (2002) Expression and potential Meek TD, Chapman C, Gloger IS, Murphy function of beta-amyloid precursor proteins KE, Southan CD, Ryan DM, Smith TS, during cutaneous wound repair. Exp Cell Simmons DL, Walsh FS, Dingwall C, Christie Res.; 280: 222–32. G. (1999) Identification of a novel aspartyl Kurreck J. (2003) Antisense technologies, im- protease (Asp 2) as b-secretase. Mol Cell provement through novel chemical modifica- Neurosci.; 14: 419–27. tions. Eur J Biochem.; 270: 1628–44. Jackson AL, Bartz SR, Schelter J, Kobayashi Kuwabara T, Warashina M, Sano M, Tang H, SV, Burchard J, Mao M, Li B, Cavet G, Wong-Staal F, Munekata E, Taira K. (2001) Linsley PS. (2003) Expression profiling re- Recognition of engineered tRNAs with an ex- veals off-target gene regulation by RNAi. Nat tended 3¢ end by exportin-t (Xpo-t) and Biotechnol.; 21: 635–7. transport of tRNA-attached ribozymes to the Jarrett JT, Berger EP, Lansbury PT Jr. (1993) cytoplasm in somatic cells. Biomacro- The carboxy terminus of the b amyloid pro- molecules.; 2: 1229–42. tein is critical for the seeding of amyloid for- Lewin AS, Hauswirth WW. (2001) Ribozyme mation: implications for the pathogenesis of : applications for molecular Alzheimer’s disease. Biochemistry.; 32: medicine. Trends Mol Med.; 7: 221–8. 4693–7. Li Q, Sudhof TC. (2004) Cleavage of amy- John V, Beck JP, Bienkowski MJ, Sinha S, loid-beta precursor protein and amyloid-b Heinrikson RL. (2003) Human b-secretase precursor-like protein by BACE 1. J Biol (BACE) and BACE inhibitors. J Med Chem.; Chem.; 279: 10542–50. 46: 4625–30. Li R, Lindholm K, Yang LB, Yue X, Citron M, Kao SC, Krichevsky AM, Kosik KS, Tsai LH. Yan R, Beach T, Sue L, Sabbagh M, Cai H, (2004) BACE1 suppression by RNA interfer- Wong P, Price D, Shen Y. (2004) Amyloid ence in primary cortical neurons. J Biol beta peptide load is correlated with in- Chem.; 279: 1942–9. creased b-secretase activity in sporadic Alz- Kawasaki H, Warashina M, Kuwabara T, Taira heimer’s disease patients. Proc Natl Acad K. (2004) Helicase-attached novel hybrid Sci USA.; 101: 3632–7. ribozymes. Methods Mol Biol.; 252: 237–44. Lichtenthaler SF, Dominguez DI, Westmeyer Kitazume S, Tachida Y, Oka R, Kotani N, GG, Reiss K, Haass C, Saftig P, De Strooper Ogawa K, Suzuki M, Dohmae N, Takio K, B, Seed B. (2003) The cell adhesion protein Saido TC, Hashimoto Y. (2003) Characteriza- P-selectin glycoprotein ligand-1 is a substrate tion of alpha 2,6-sialyltransferase cleavage by for the aspartyl protease BACE1. J Biol Alzheimer’s beta-secretase (BACE1). J Biol Chem.; 278: 48713–9. Chem.; 278: 14865–71. Lin X, Koelsch G, Wu S, Downs D, Dashti A, Kitazume S, Tachida Y, Oka R, Shirotani K, Tang J. (2000) Human aspartic protease Saido TC, Hashimoto Y. (2001) Alzheimer’s memapsin 2 cleaves the b-secretase site of Vol. 51 Small inhibitory nucleic acids for BACE 443

beta-amyloid precursor protein. Proc Natl Opalinska JB, Gewirtz AM. (2002) Nucleic acid Acad Sci USA.; 97: 1456–60. therapeutics: a work in progress. Curr Opin Investig Drugs.; 3: 928–33. Luo Y, Bolon B, Kahn S, Bennett BD, Babu-Khan S, Denis P, Fan W, Kha H, Paddison PJ, Caudy AA, Bernstein E, Hannon Zhang J, Gong Y, Martin L, Louis JC, Yan GJ, Conklin DS. (2002) Short hairpin RNAs Q, Richards WG, Citron M, Vassar R. (2001) (shRNAs) induce sequence-specific silencing Mice deficient in BACE1, the Alzheimer’s in mammalian cells. Genes Dev.; 16: 948–58. beta-secretase, have normal phenotype and Paddison PJ, Caudy AA, Hannon GJ. (2002) abolished b-amyloid generation. Nat Stable suppression of gene expression by Neurosci.; 4: 231–2. RNAi in mammalian cells. Proc Natl Acad Mattson MP, Cheng B, Culwell AR, Esch FS, Sci USA.; 99: 1443–8. Lieberburg I, Rydel RE. (1993) Evidence for Pardridge WM. (2002) Drug and gene targeting excitoprotective and intraneuronal cal- to the brain with molecular trojan horses. cium-regulating roles for secreted forms of Nat Rev Drug Discov.; 1: 131–9. the b-amyloid precursor protein. Neuron.; 10: 243–54. Paul CP, Good PD, Winer I, Engelke DR. (2002) Effective expression of small interfering RNA Mattson MP, Furukawa K. (1997) Alzheimer’s in human cells. Nature Biotechnol.; 20: disease. Short precursor shortens memory. 505–8. Nature.; 387: 457–8. Persengiev SP, Zhu X, Green MR. (2004) Non- McSwinggen J, Beigelman L. (2003) RNA inter- specific, concentration-dependent stimulation ference mediated treatment of Alzheimer’s and repression of mammalian gene expres- disease using short interfering nucleic acids sion by small interfering RNAs (siRNAs). (siNA). Patent WO 03/070895. RNA.; 10: 12–8. Miyagishi M, Taira K. (2002) U6 pro- Qiao L, Etcheberrigaray R. (2002) moter-driven siRNAs with four uridine 3¢ Phosphinylmethyl and phosphorylmethyl overhangs efficiently suppress targeted gene succinic and glutaric acid analogs as expression in mammalian cells. Nat b-secretase inhibitors. Patent WO 02/96897. Biotechnol.; 20: 497–500. Roberds SL, Anderson J, Basi G, Bienkowski Miyamoto M, Matsui J, Fukumoto H, Tarui N. MJ, Branstetter DG, Chen KS, Freedman (2001) b-Secretase inhibitors. Patent WO SB, Frigon NL, Games D, Hu K, John- 01/187293. son-Wood K, Kappenman KE, Kawabe TT, Nawrot B, Antoszczyk S, Maszewska M, Kola I, Kuehn R, Lee M, Liu W, Motter R, Kuwabara T, Warashina M, Taira K, Stec Nichols NF, Power M, Robertson DW, WJ. (2003) Efficient inhibition of b-secretase Schenk D, Schoor M, Shopp GM, Shuck ME, (BACE) gene expression in HEK293T cells by Sinha S, Svensson KA, Tatsuno G, Tintrup tRNAVal/CTE-driven hammerhead ribozymes. H, Wijsman J, Wright S, McConlogue L. Eur J Biochem.; 270, 3962–70. (2001) BACE knockout mice are healthy de- spite lacking the primary beta-secretase ac- Nawrot B, Antoszczyk S, Maszewska M, tivity in brain: implications for Alzheimer’s Rebowski G, Kuwabara T, Warashina M, disease therapeutics. Hum Mol Genet.; 10: Taira K, Stec WJ. (2002) Modulation of 1317–24. b-secretase gene expression by action of cata- lytic nucleic acids. Nucleic Acids Res.; Suppl. Roggo S. (2002) Inhibition of BACE, a promis- 2: 105–6. ing approach to Alzheimer’s disease therapy. Curr Top Med Chem.; 2: 359–70. Omi K, Tokunaga K, Hohjoh H. (2004) Long-lasting RNAi activity in mammalian Schmidt B. (2003) Aspartic proteases involved neurons. FEBS Lett.; 558: 89–95. in Alzheimer’s disease. ChemBioChem.; 4: 366–78. 444 B. Nawrot 2004

Schwarz D, Hutvagner G, Du T, Xu Z, Aronin J. (2001) Subsite specificity of memapsin 2 N, Zamore PD. (2003) Assymetry in the as- (b-secretase): implications for inhibitor de- sembly of the RNAi enzyme complex. Cell.; sign. Biochemistry.; 40: 10001–6. 115: 199–208. Vassar R, Bennett BD, Babu-Khan S, Kahn S, Selkoe DJ. (1991) The molecular pathology of Mendiaz EA, Denis P, Teplow DB, Ross S, Alzheimer’s disease. Neuron.; 6: 487–98. Amarante P, Loeloff R, Luo Y, Fisher S, Fuller J, Edenson S, Lile J, Jarosinski MA, Simons K, Ehehalt R. (2002) Cholesterol, lipid Bier AL, Curran E, Burgess T, Louis JC, Col- rafts, and disease. J Clin Invest.; 110: lins F, Treanor J, Rogers G, Citron M. 597–603. (1999) b-secretase cleavage of Alzheimer’s Sinha S, Anderson JP, Barbour R, Basi GS, amyloid precursor protein by the trans- Caccavello R, Davis D, Doan M, Dovey HF, membrane aspartic protease BACE. Science.; Frigon N, Hong J, Jacobson-Croak K, Jewett 286: 735–41. N, Keim P, Knops J, Lieberburg I, Power M, Vinogradov SV, Batrakova EV, Kabanov AV. Tan H, Tatsuno G, Tung J, Schenk D, (2004) Nanogels for oligonucleotide delivery Seubert P, Suomensaari SM, Wang S, to the brain. Bioconjug Chem.; 15: 50–60. Walker D, John V., et al. (1999) Purification and cloning of amyloid precursor protein Warashina M, Kuwabara T, Kato Y, Sano M, b-secretase from human brain. Nature.; 402: Taira K. (2001) RNA-protein hybrid 537–40. ribozymes that efficiently cleave any mRNA independently of the structure of the target Stein CA, Cheng YC. (1993) Antisense RNA. Proc Natl Acad Sci USA.; 98: 5572–7. oligonucleotides as therapeutic agents — is the bullet really magical? Science.; 261: Yan R, Bienkowski MJ, Shuck ME, Miao H, 1004–12. Tory MC, Pauley AM, Brashier JR, Stratman NC, Mathews WR, Buhl AE, Carter DB, Sullenger BA, Gilboa E. (2002) Emerging clini- Tomasselli AG, Parodi LA, Heinrikson RL, cal applications of RNA. Nature.; 418: Gurney ME. (1999) Membrane-anchored 252–8. aspartyl protease with Alzheimer’s disease Trülzsch B, Wood M. (2004) Application of nu- b-secretase activity. Nature.; 402: 533–7. cleic acid technology in the CNS. J Zuker M. (2003) Mfold web server for nucleic Neurochem.; 88: 257–65. acid folding and hybridization prediction. Nu- Turner RT 3ed, Koelsch G, Hong L, cleic Acids Res.; 31: 3406–15. Castanheira P, Ermolieff J, Ghosh AK, Tang