Preclinical Development of an RNAi Therapeutic Drug Candidate Targeting Hepatitis B Virus Authors: Stuart Milstein, Tuyen Nguyen, Adam Castoreno, Abigail Liebow, Vasant Jadhav, Martin Maier, Laura Sepp-Lorenzino

Affiliations: Alnylam Pharmaceuticals, Cambridge, MA 02142, USA

Abstract Current HBV therapies, including polymerase inhibitors, effectively reduce viral titers in chronic hepatitis B by inhibiting viral replication, but these therapies fail to eradicate the infection in ~90% of treated patients. Even in the absence of viral replication, high plasma levels of non-infectious, HBsAg-containing subviral particles are thought to mediate immunological tolerance. Reduction in HBsAg plasma levels is the single best predictor of immunological cure. An RNAi therapeutic targeting the HBV genome has the potential to achieve a “functional cure” by effectively decreasing expression of tolerogenic HBsAg, in addition to inhibiting all steps of the HBV life cycle. ALN-HBV, a therapeutic RNAi candidate consisting of a siRNA with Enhanced Stabilization Chemistry (ESC) design conjugated to a GalNAc-based targeting ligand and designed for SC administration has been developed. ALN-HBV effectively targets all viral transcripts as demonstrated by RNAseq. The sequence is pan-genotypic, targeting all major HBV genotypes and resides outside of an integration hotspot making it therefore unlikely to be disrupted by HBV genome integration. A single subcutaneous dose of ALN-HBV results in potent and durable silencing of tolerogenic HBsAg in preclinical models

References: Zoulim and Locarnini. Gastroenterology. 2009; 137: 1593-608 | Basnayake and Easterbrook. J. Viral Hepatitis. 2016; 23: 545-559 | Gane. Liver Int. 2017; 37 Suppl 1: 40-44 Pollicino and Raimondo. Journal of Hepatology. 2014; 61: 688–689 | Watashi et al. Int. J. Mol. Sci. 2014, 15; 2892-2905 Schweitzer et al. Lancet. 2015; 386: 1546-1555 | Gish et al. Antiviral Res. 2015; 121: 97-108

Figure 1. Unmet Need in Chronic Hepatitis B Infection

• Chronic HBV infection is significant Worldwide problem - 2B patients exposed worldwide (1 out of 3 people) - ~ 290M patients worldwide - 25M in U.S./EU/Japan with chronic infection - 10-30M New infections each year WW - ~ 1M Deaths each year WW • Clinical manifestations severeChronic inflammation leading to cirrhosis and HCC• Current therapies have significant limitations - RT inhibitors are lifelong therapy - HBsAg clearance and anti-HBs seroconversion

(immunologic cure) is ~10% after 2-3 years of therapy with PEG-IFN or RT inhibitor

Figure 2. Disrupting the HBV Life Cycle by Targeting Viral mRNAs

1. Entry- HBV attaches to surface proteoglycans after which viral envelope binds NTCP mediating cell entry via endocytosis.

2. Transcription- HBV rcDNA is converted to cccDNA. Episomal cccDNA persists as a stable minichromosome and utilizes hepatocyte transcriptional machinery to produce all viral RNAs for production and virlal replication.

3. Translation of proteins encoded by the 4 viral transcripts. 4. Reverse Transcription- pgRNA is selectively incorporated into progeny

nucleocapsids and reverse transcribed by the coassembled viral polymerase into new HBV genomes. Mature rcDNA-containing nucleocapsids can build up a pool of cccDNA or interact with the envelope proteins at ER/Golgi apparatus and be secreted as new infectious virions

5. Secretion and Immune Response- The SVPs typically outnumber the virions by a factor of 1,000- to 10,000-fold overwhelming the immune system.

Figure 3. HBV Genome structure

3.2 kb partially double stranded DNA genome containing 4 overlapping polycystronic viral transcripts that encode 7 viral proteins

1. 3.5kb pgRNA / pre-Core RNAs encoding Nucleocapsid protein HBcAg, Soluble secreted HBeAg, Pol protein

2. 2.4kb RNA encoding L envelope protein 3. 2.1kb RNA encoding M and S envelope proteins 4. 0.7kb RNA encoding X-protein

Silencing all 4 viral transcripts will likely be the most impactful strategy for targeting HBV

Figure 4. ALN-HBV Targets all 4 Viral Transcripts at a Highly Conserved Sequence in HBV X

ALN-HBV targets a sequence that is well conserved across HBV genotypes. The HBV genome is highly polymorphic creating a challenge for single agent pan-genotypic targeting. (Top left) Distribution of HBV genotypes used for analysis of sequence conservation. (Top right) An analysis of sequence conservation across a panel of 3980 HBV genomes. Percent conservation is depicted as peak height relative to position in the 4 transcripts and the target region of ALN-HBV (red box). (Bottom left) ALN-HBV is conserved in over 92% of genomes and overlaps all 4 transcripts. (Bottom right) ALN-HBV is a potent siRNA with a 20 pM IC50 in the HBV expressing cell line HepG2.2.15

Figure 5. Reduction of HBsAg in Serum Following a Single Subcutaneous Dose of in ALN-HBV

Dose dependent reduction of HBsAg is seen following as single dose of ALN-HBV. Impact of ALN-HBV on HBsAg serum levels in mice transduced with AAV-HBV viral vector after a single SC dose. 14 days after AAV-HBV transduction (day 0), mice were does with 0.3, 1, 3 or 9 mg/kg of ALN-HBV. Serum HBV was monitored periodically. N=5 mice per group. Single SC dose achieves >2 log10 HBsAg reduction lasting >30 days.

Figure 6. Reduced HBsAg in Liver After a Single Dose of ALN-HBV

Reduction of HBsAg is seen following as single dose of ALN-HBV in liver tissue. Mice that were previously transduced with AAV-HBV they were treated with 3 mg/kg ALN-HBV, an siRNA targeting mouse TTR or with PBS. Immunohistochemistry was performed on sections of mouse liver and probed with antibody against the HBV-S antigen (brown) and HBV-core antigen (purple). Reprehensive sections are shown from the PBS and TTR negative controls and from 2 different ALN-HBV treated animals. While HBsAg is almost completely absent, some residual HBV-core staining is seen.

Figure 7. ALN-HBV Silences All HBV Transcripts In Vivo

ALN-HBV treatment reduces expression of all transcripts in AAV-HBV transduced mice. 14 days after AAV-HBV transduction, mice were injected SC with 3 mg/kg ALN-HBV or an siRNA targeting mouse TTR. Livers were harvested after 28 days, RNA was isolated and RNAseq performed. In TTR control animals, a distribution or read frequencies is seen across the HBV transcriptome, whereas read frequency is uniformly reduced 21 days following a single dose of ALN-HBV. Red box highlights region targeted by ALN-HBV

Figure 8. ALN-HBV Has Similar Activity Against All HBV Genotypes

In vitro silencing of 10 HBV genotypes with ALN-HBV. ALN-HBV, which targets a conserved binding site is able to silence all HBV genotypes assayed. Representative HBV-X genes from HBV genotypes A-J were cloned into a dual-luciferase reporter construct and assayed for silencing 24 hours after transfection with 10 nM of ALN-HBV.

Figure 9. The ALN-HBV Target Site is Conserved in HCC Samples Where Viral Integration has Likely Occurred

The ALN-HBV target site is preserved in late stage HCC samples. HCC liver samples were purchased from a tissue bank and used to evaluate conservation of the ALN-HBV binding site in samples that were likely to have integrated HBV. Top- 5 nonoverlapping PCR amplicons, spanning the HBV genome from S to core were designed and used for amplification of HBV(A). Each fragment was measured by qPCR from 10 HCC samples and message quantification perfumed. (B)While there is some loss of signal as the amplicons progress from S toward X (amplicons 1 to 5), the most striking loss is between amplicons 4 and 5, showing that the ALN-HBV binding site is likely to be preserved following HBV integration in late stage disease. Bottom- HCC samples were used to measure HBV expression by RNAseq. (C)10 HCC samples (5 previously used in the qPCR experiment and 5 additional samples) were analyzed. All 10 samples show that the ALN-HBV binding site is preserved, whereas the downstream signal is lost.

Figure 11. Some Polymorphisms in the ALN-HBV Binding Site Impact Silencing and Could Lead to Resistance

In vitro activity of ALN-HBV against binding site mutations. Luciferase reporter constructs were generated for 15 binding site polymorphisms identified from the library of 3980 HBV genomes used to assess conservation. Seed based polymorphisms were favored since they would be most likely to impact activity. Luciferase constructs were cotransfected with 10 nM ALN-HBV and activity was measured after 48 hours.

Figure 12. ALN-HBV Binding Site Mutations Are Unlikely to Cause Resistance to Nucleoside Analogs

ALN-HBV targets a region that does not overlap with known nucleoside resistance mutations. ALN-HBV is unlikely to result in resistance to nuclos(t)ide analogues, which are the current standard of care for the treatment of HBV. Analysis of known resistance mutations shows no overlap with the ALN-HBV binding site.

Summary• We have developed ALN-HBV, an siRNA targeting a highly

conserved HBV target site• ALN-HBV is potent and durable, with an in vitro IC50 of

20 pM and an approximately 2 log reduction of HBsAg in vivo

• Single dose in vivo silencing for over 70 days at 9 mg/kg• ALN-HBV reduces expression of all HBV transcripts, is pan-

genotypic and is unlikely to interfere with current standard of care