Lipid nanoparticle-encapsulated mRNA therapy corrects serum total bilirubin level in Crigler-Najjar syndrome mouse model

Crigler-Najjar syndrome is a rare disorder of bilirubin metabolism caused by uridine diphosphate glucuronosyl transferase 1A1 (UGT1A1) mutations characterized by hyperbilirubinemia and jaundice. No cure currently exists; treatment options are limited to phototherapy, whose effectiveness diminishes over time, and liver transplantation. Here, we evaluated the therapeutic potential of systemically administered, lipid nanoparticle-encapsulated human UGT1A1 (hUGT1A1) mRNA therapy in a Crigler-Najjar mouse model. Ugt1 knockout mice were rescued from lethal post-natal hyperbilirubinemia by phototherapy. These adult Ugt1 knockout mice were then administered a single lipid nanoparticle-encapsulated hUGT1A1 mRNA dose. Within 24 h, serum total bilirubin levels decreased from 15 mg/dL (256 μmol/L) to <0.5 mg/dL (9 μmol/L), i.e., slightly above wild-type levels. This reduction was sustained for 2 weeks before bilirubin levels rose and returned to pre-treatment levels by day 42 post-administration. Sustained reductions in total bilirubin levels were achieved by repeated administration of the mRNA product in a frequency-dependent manner. We were also able to rescue the neonatal lethality phenotype seen in Ugt1 knockout mice with a single lipid nanoparticle dose, which suggests that this may be a treatment modality appropriate for metabolic crisis situations. Therefore, lipid nanoparticle-encapsulated hUGT1A1 mRNA may represent a potential treatment for Crigler-Najjar syndrome.


INTRODUCTION
Crigler-Najjar syndrome is an ultra-rare disorder of bilirubin metabolism caused by mutations in the uridine diphosphate glucuronosyl transferase 1A1 (UGT1A1) gene. [1][2][3][4][5] The most severe form of Crigler-Najjar syndrome is called type 1 (CN1) and presents as persistent jaundice at or soon after birth as a result of hyperbilirubinemia >20 mg/dL (342 mmol/L). 6 The lack of functional UGT1A1 prevents bilirubin from being conjugated in hepatocytes prior to excretion, 7 and the transit of these abnormally high levels of bilirubin across the blood-brain barrier can cause kernicterus, which can result in irreversible neurological damage.
Currently, no cure exists for CN1. Clinical management primarily consists of phototherapy, which negatively affects the quality of life, is variable in the duration required per patient, is only partially effective at lowering bilirubin levels, and becomes less effective over time. 8,9 Other illnesses can also lead to severe bilirubin excursions in which multiple clinical strategies are deployed to reverse the crisis and prevent kernicterus. Liver transplantation is a treatment option, although there is potential for morbidity and, in rare cases, mortality associated with the surgical procedure and long-term immunosuppression. These limited treatment options highlight the tremendous unmet clinical need and the potential positive impact of a novel therapeutic approach.
Previous attempts at alternative treatment strategies have utilized adeno-associated virus vectors to achieve therapeutic expression of human UGT1A1 (hUGT1A1). [10][11][12][13][14][15][16][17][18] While these gene therapy approaches have proven efficacious in other liver diseases, they are limited by interfering antibodies and concerns over the durability of expression. Others have evaluated the therapeutic potential of a lipid nanoparticle (LNP)-encapsulated mRNA therapy for the treatment of CN1. 19 mRNA delivery by LNPs results in hepatocyte distribution through opsonization, followed by receptor-mediated endocytosis. [20][21][22][23][24][25] By delivering mRNA to hepatocytes (i.e., the site of action for UGT1A1), this approach can potentially treat the fundamental cause of CN1. Previous studies utilized the naturally occurring Gunn rat model to evaluate LNP-encapsulated mRNA therapy for CN1. 19 Gunn rats lack active hepatic UGT1A1 and exhibit sustained hyperbilirubinemia due to a frameshift mutation caused by a single base deletion in UGT1A1. 26,27 Here, we evaluated the therapeutic potential of an LNP-encapsulated mRNA therapy following systemic administration in a CN1 mouse model, the Ugt1 knockout (KO) mouse. Compared with the less severe phenotype displayed in the Gunn rat, the Ugt1 KO mouse displays hyperbilirubinemia in the immediate post-natal period that is fatal within the first week of life without intervention. 14,28 Intervention with phototherapy (i.e., exposure to blue fluorescent light for 12 h/day for up to 21 days after birth) rescues this lethal neonatal phenotype and allows affected mice to survive to adulthood, albeit while displaying the hyperbilirubinemia characteristic of the disease (i.e., elevated serum total bilirubin levels of 9.1 ± 3 mg/dL following weaning from phototherapy). 11,14,29 We evaluated the pharmacokinetics and efficacy of repeated systemic dosing of LNP-encapsulated hUGT1A1 mRNA in adult Ugt1 KO mice and investigated whether the neonatal lethality phenotype seen in Ugt1 KO mice could be rescued with a single LNP dose.

RESULTS
A single dose of LNP-encapsulated mRNA therapy normalizes serum bilirubin in the Ugt1 KO mouse To evaluate the ability of LNP-encapsulated mRNA to treat CN1, we utilized the Ugt1 KO mouse model. In order to ameliorate the hyperbilirubinemia phenotype characteristic of this model, a proof-of-principle study was performed whereby phototherapy-rescued adult Ugt1 KO mice were systemically administered a single dose of LNP-encapsulated hUGT1A1 mRNA. Mice were injected with a single dose of 0.05, 0.2, or 0.5 mg/kg of LNP-encapsulated mRNA via the tail vein ( Figure 1). Mice dosed with 0.2 or 0.5 mg/kg exhibited reductions in serum total bilirubin levels to approximately wild-type levels (0.4-0.5 mg/dL) that were sustained for 2 weeks, at which time bilirubin levels began to rise before returning to baseline levels of hyperbilirubinemia by day 42 post-administration. There was no significant difference between serum total bilirubin levels in mice dosed with either 0.2 or 0.5 mg/kg, with both doses significantly improving the phenotype compared to mice treated with 0.05 mg/kg LNP-encapsulated hUGT1A1 mRNA.
Repeated administration of LNP-encapsulated mRNA therapy sustains reductions in serum bilirubin levels in the Ugt1 KO mouse Following completion of the washout period after the single dose of LNP-encapsulated mRNA (Figure 1), the same Ugt1 KO mice were repeatedly administered LNP-encapsulated hUGT1A1 mRNA. By varying the dosing interval, we determined the optimal dosing regimen to achieve sustained reductions in serum total bilirubin levels. Mice were treated every 2 weeks (Q2W; Figure 2A), every 3 weeks (Q3W; Figure 2B), or every 4 weeks (Q4W; Figure 2C) with 0.2 mg/kg LNP-encapsulated hUGT1A1 mRNA or Q2W with a negative control LNP containing green fluorescent protein (GFP) mRNA (same control group presented in Figures 2A-2C).
Q2W administration of the GFP LNP did not alter serum total bilirubin levels across the study duration, and the control Ugt1 KO mice displayed the typical degree of variability in baseline levels (8.3 ± 2.1 mg/dL; Figure 2). Mice treated with Q2W administration of the LNP hUGT1A1 mRNA had a sustained reduction in serum total bilirubin levels for the duration of the study, with some fluctuations occurring later in the study ( Figure 2). Over the course of the study, it became more challenging to perform repeated intravenous (i.v.) administrations in these mice due to limited access sites, which may have resulted in incomplete dosing. Dosing Q3W or Q4W resulted in reductions in serum bilirubin levels for 7-14 days post-injection (similar to that seen following a single dose [ Figure 1]), which returned to baseline hyperbilirubinemia levels prior to the next dose 21 or 28 days later, respectively ( Figures 2B and 2C). While liver transaminase elevations were present directly following LNP administration across the groups, we did not observe any sustained elevations throughout the study duration ( Figure S1). The livers were harvested at necropsy, and we performed histopathological analyses (Table 1). Minimal-to-mild findings of mononuclear cell infiltration were seen across all groups, including mice that received the GFP LNP. Other minimal findings of single hepatocyte necrosis, pigmented macrophages, Kupffer cell infiltrates, and oval cell hyperplasia were also observed in some mice but were considered incidental.
While all dosing intervals significantly reduced serum total bilirubin levels compared with control Ugt1 KO mice, Q2W dosing achieved average levels of 2.78 mg/dL, which was a substantial improvement on levels following Q3W and Q4W dosing. The duration of action of the LNP hUGT1A1 mRNA is dependent on both mRNA and protein stability; our data indicate that a biweekly dosing regimen can effectively and sustainably reduce bilirubin levels in the Ugt1 KO mouse.

Time course of LNP-encapsulated mRNA therapy for CN
To determine the time course of both the hUGT1A1 mRNA and expressed protein in vivo, adult C57BL/6J mice were injected with 0.5 mg/kg LNP-encapsulated hUGT1A1 mRNA and sacrificed at 6 h, 24 h, 48 h, 72 h, 7 days, 10 days, and 14 days post-administration. Necropsies were performed, and livers were harvested for immunohistological analyses ( Figure 3A) to evaluate RNA levels by in situ hybridization (ISH) ( Figure 3B) and hUGT1A1 protein levels by Western blot (Figures 3C and S2). There was detectable RNA in the liver up to 72 h post-LNP administration by ISH, and hUGT1A1 expression from the delivered mRNA was sustained through 10 days post-LNP administration ( Figure 3).
Optimization of the hUGT1A1 mRNA sequence in adult Ugt1 KO mice We further developed the hUGT1A1 mRNA sequence and evaluated the efficacy of this second mRNA with a different 5 0 untranslated region (UTR) with the same Cap, open reading frame (ORF), 3 0 UTR, and tail background compared with the original mRNA in a repeat of the single-dose study described above. Adult Ugt1 KO mice were dosed at 0.2 mg/kg LNP-encapsulated hUGT1A1 mRNA (either original or redesigned sequence) in a single i.v. injection. An additional group of mice received the vehicle as a control (phosphate-buffered saline [PBS]). Serum total bilirubin levels were decreased in a similar manner in the original and redesigned mRNA treatment groups, although a lower level was achieved on day 7 with the redesigned mRNA sequence (to 1.4 mg/dL or 15% of baseline levels compared with 2.2 mg/dL or 30% of baseline with the original mRNA) ( Figure 4A).
A single administration of LNP-encapsulated mRNA therapy to newborn Ugt1 KO mice rescues neonatal lethality We evaluated the utility of this LNP-encapsulated mRNA therapy with the redesigned sequence to rescue Ugt1 KO mice from their lethal neonatal phenotype in which they require euthanasia within the first week of life. 11,28 Litters of mice resulting from heterozygous Ugt1 +/À mating were injected with 0.1, 0.2, or 0.5 mg/kg LNP-encapsulated mRNA therapy or vehicle on the first day of life ( Figure 4B). Mice were followed for survival; vehicle-treated Ugt1 KO mice died by day 3, mimicking what we observed in untreated Ugt1 KO mice. 11 Treatment with 0.1 mg/kg or 0.2 mg/kg of the redesigned LNP-encapsulated hUGT1A1 mRNA sequence enhanced survival to a maximum of 17 and 19 days post-birth, respectively ( Figure 4B). A single treatment with the highest dose (0.5 mg/kg) allowed all Ugt1 KO mice to survive until weaning (days [21][22][23][24][25][26][27][28]; this group demonstrated survival to >100 days, which is the latest we observed the animals and is well beyond lethal neonatal exposure. While the transient nature of LNPencapsulated mRNA therapy was sufficient to rescue the neonatal Ugt1 KO mice, these animals had hyperbilirubinemia levels comparable to phototherapy-rescued Ugt1 KO animals at day 28. Liver samples collected at necropsy (>100 days post-administration) were evaluated for histopathology; minimal findings of infiltrates were seen in two out of seven Ugt1 KO mice administered with the highest dose and were considered to be background findings ( Figure S3).

DISCUSSION
We have shown the efficacy of LNP-encapsulated mRNA for treating hyperbilirubinemia resulting from UGT1A1 deficiency in the Ugt1 KO mouse model of CN1. After establishing the pharmacokinetics of both the treatment and the response following a single dose of LNP in adult mice, we were able to determine that mRNA dosing every two weeks (Q2W) was required to prevent the peaks in serum total bilirubin levels indicative of a reversal to hyperbilirubinemia that were seen in the other groups dosed less frequently. Therefore, repeated administration of LNP-encapsulated hUGT1A1 mRNA may hold potential for both short-and long-term treatment of CN1. These data are similar to those generated previously in the Gunn rat; however, we were able to demonstrate the efficacy of this approach in an animal model with an enhanced severity phenotype (the Ugt1 KO mouse) compared with that of the Gunn rat, 19 suggesting that this approach could be effective upon translation to patients with CN1 in the clinic.
In mice repeatedly administered with LNP-encapsulated hUGT1A1 mRNA, we observed some fluctuations in serum total bilirubin levels later in the study, which correlated with it becoming more challenging to perform repeated i.v. administrations in these mice. The resulting fluctuations may then have resulted from incomplete dosing. An alternative explanation could be a potential immune response to the hUGT1A1 protein in the animal model of CN1. 13,16,18,30 While we did not evaluate adaptive immune responses to hUGT1A1 protein or cells expressing this protein, mice treated with biweekly administration of the hUGT1A1 mRNA retained sustained reductions in serum total bilirubin levels for the duration of the study (154 days), suggesting an absence of deleterious antibodies or T cells. 31 In this study, we also observed transient liver transaminase elevations directly following LNP administration, similar to those previously reported following delivery of LNP-mRNA treatment for methylmalonic acidemia. 32 In addition, histopathological changes in the liver were limited to minimal-to-mild mononuclear cell infiltrations across all groups at necropsy following up to 12 administrations with LNP hUGT1A1 mRNA with no evidence of hepatocyte injury.
Current clinical management of patients with CN1 consists predominantly of daily phototherapy to stimulate the photoisomerization of bilirubin into water-soluble isomers. The number of hours of phototherapy required per day can range from >10 h to maintain young patients with CN1 below the threshold for encephalopathy 8,33 to 20 h per day. 34 Unfortunately, phototherapy is only partially effective at lowering bilirubin levels, and patients with CN1 have lifelong hyperbilirubinemia (16 ± 5 mg/dL) 8 and an ongoing risk of reaching threshold levels for kernicterus during times of illness or stress. This highly burdensome treatment approach also becomes less effective as the patient ages due to alterations in body surface area, thickening of the skin, and decreased compliance due to its obvious inconvenience. 35 While recurrent i.v. administration of an LNP-encapsulated mRNA therapy at less than monthly intervals would be associated with some inconveniences, including those based on the dosing interval, infusion volume and time, and the location where dosing could be performed, there is the expectation that this treatment approach would address the cause of CN1 without imposing the same limitations on travel, employment, or lifestyle as phototherapy.
In addition to evaluating LNP-encapsulated hUGT1A1 mRNA as a long-term treatment approach following repeated administration, we also evaluated this therapy to treat acute metabolic crises in CN1. When bilirubin levels are being managed by phototherapy, there is still a risk of other adverse events resulting in severe excursions into hyperbilirubinemia. In addition, other illnesses, stressors, or even just the interruption of phototherapy for a short period can result in brain damage. We believe that LNP-encapsulated hUGT1A1 mRNA could be rapidly deployed to treat acute metabolic crises and prevent kernicterus. We modeled this in the Ugt1 KO mice by evaluating the efficacy of this mRNA treatment to rescue the neonatal lethality seen within the first week of life in this strain. A single 0.5 mg/kg dose of LNP-encapsulated mRNA within the first 24 h of birth completely rescued the Ugt1 KO mice. Together with the data from the pharmacokinetic evaluations, this suggests that administration of LNP-encapsulated mRNA at the time of crisis could be sufficient to reverse hyperbilirubinemia within hours.
As hepatocytes are the primary site of gene expression following LNPencapsulated mRNA therapy, [21][22][23][24][25] there are potential implications for patients with liver function issues. Except for a deficiency in UGT1A1 activity, the structure and function of the liver are generally normal in CN1 patients. However, there have been recent case reports of intrahepatic cholestasis and fibrosis, which increase with age. 9,36 While the effect of liver fibrosis on efficacy and safety of LNP-encapsulated mRNA therapy has not yet been evaluated, the degree of fibrosis reportedly correlates with the lifetime average bilirubin levels in patients, 9 suggesting that early intervention with chronic LNP-encapsulated mRNA therapy could prevent future liver fibrosis. We therefore conclude that LNP mRNA therapy may be suitable for both the long-term treatment of CN1 and for managing acute hyperbilirubinemic metabolic crises associated with this disease.

mRNA/LNP production
A sequence-optimized mRNA encoding hUGT1A1 was synthesized in vitro using an optimized T7 RNA polymerase-mediated transcription reaction with complete replacement of N1-methylpseudouridine, as previously described. 37 The reaction included a DNA template con-taining the protein ORF flanked by 5 0 UTR and 3 0 UTR sequences and was terminated by an encoded polyA tail. After transcription, the Cap 1 structure was added to the 5 0 end using vaccinia capping enzyme (New England Biolabs) and vaccinia 2 0 O-methyltransferase (New England Biolabs). The mRNA was purified by oligo-dT affinity purification, buffer exchanged by tangential flow filtration into sodium acetate (pH 5), sterile filtered, and kept frozen at À20 C until further use. A second 5 0 UTR (proprietary sequence) in the same Cap, ORF, 3 0 UTR, and tail background was also evaluated.
The mRNA was encapsulated in an LNP through a modified ethanoldrop nanoprecipitation process as described previously. 38 In brief, ionizable, structural, helper, and polyethylene glycol lipids were mixed with mRNA in acetate buffer (pH 5). The mixture was neutralized with Tris-Cl (pH 7.5), sucrose was added as a cryoprotectant, and the final solution was sterile filtered. Vials were filled with formulated LNP and stored frozen at À70 C until further use. The drug product underwent analytical characterization, which included the determination of particle size and polydispersity, encapsulation, mRNA purity, double-stranded RNA content, osmolality, pH, endotoxin, and bioburden, and the material was deemed acceptable for in vivo study.

Mice
Breeding pairs of heterozygous Ugt1 +/À mice (mixed B6 and 129 background strains) were obtained from The Jackson Laboratory (Bar Harbor, ME, USA), and a colony was maintained at the University of Pennsylvania under specific-pathogen-free conditions. All mice used for this study were derived from this colony. All animal procedures and protocols were approved by the Institutional Animal Care and Use Committee of the University of Pennsylvania.  For newborn studies, Ugt1 KO mice and heterozygous and wild-type littermates received an i.v. injection with 0.1, 0.2, or 0.5 mg/kg LNPencapsulated hUGT1A1 mRNA or vehicle (20 mM Tris, 8% sucrose buffer [pH 7.4]) within 24 h of birth via the temporal vein. Mice were monitored for survival and serum total bilirubin levels throughout the in-life phase of the study.

Serum analyses
Blood was collected in serum separator tubes (Micro sample tube Serum Gel, 1.1 mL, screw cap, Sarstedt, Germany) and allowed to clot. Serum was isolated and sent to Antech Diagnostics (Irvine, CA, USA) to analyze total bilirubin levels.

Histological evaluation
Hematoxylin and eosin staining for histopathology and ISH were performed as described previously. 12 An experienced board-certified veterinary pathologist microscopically examined liver sections for histopathology. The severity of the tissue lesions was graded as follows: grade 1 (1+): minimal, an inconspicuous to barely noticeable histopathological feature (involves $10% of the tissue in an average high-power field); grade 2 (2+): mild, a noticeable but not prominent histopathological feature (involves 10% to 25% average high-power field); grade 3 (3+): moderate, a prominent but not dominant histopathological feature (involves 25% to 50% of the tissue in an average www.moleculartherapy.org high-power field); grade 4 (4+): marked, a dominant but not overwhelming histopathological feature (involves 50% to 95% of the tissue in an average high-power field); and grade 5 (5+): severe, an overwhelming histopathological feature (involves greater than $95% of the tissue in an average high-power field).
ISH images were quantified using ImageJ. 39 Four to five images per sample were acquired with a Leica DMi8 microscope using rhodamine and 4 0 ,6-diamidino-2-phenylindole (DAPI) channels to visualize ISH signals (hUGT1A1 mRNA) and nuclei, respectively. Images were converted to 8 bits, and the area of ISH staining and number of nuclei were determined after thresholding. The ratio of ISH-positive area per number of nuclei was determined for each image and averaged for each group.

Protein analysis by Western blot
Liver samples were snap frozen at the time of necropsy, and tissue homogenates were prepared in the cell lysis buffer (Thermo Fisher Scientific) containing protease inhibitor cocktails (Sigma, St. Louis, MO, USA) followed by sonication to shear DNA. These extracts (30 mg protein) were subjected to immunoblotting with human recombinant anti-UGT1A1 (Abcam, ab170858) and anti-b-actin (Abcam, ab20272) antibodies. The signal was developed with SuperSignal West Pico PLUS Chemiluminescent Substrate (Thermo Fisher Scientific), and images were acquired on the Bio-Rad Chemi-Doc Imaging system. Densitometry was performed against b-actin using Bio-Rad ImageLab 6.1 following the manufacturer's guidelines. Data are presented as fold change relative to signal strength at 6 h.

Statistical analysis
Evaluation of the statistical differences in serum total bilirubin levels over time was performed by linear mixed-effect modeling. Comparisons of differences between all serum total bilirubin values between groups were performed using t tests. All values are presented as mean ± standard deviation (SD). A p value of <0.05 was considered significant.

DATA AVAILABILITY
All data discussed in the manuscript are available in the main text or the supplemental information.

ACKNOWLEDGMENTS
We would like to thank the Program for Comparative Medicine and the Histopathology Core at the Gene Therapy Program. We would like to thank Christine Draper, Katherine Trostle, Peter Hewins, Brett Brown, and Kimberly Tulish for invaluable technical assistance. In addition, we would like to thank Hanying Yan for help with statistical analysis and Aditya Dandekar for their continued support of this project. This work was supported by internal funds from the Univer-sity of Pennsylvania's Gene Therapy Program; the LNPs were provided by Moderna Inc.