Preclinical use of a clinically-relevant scAAV9/SUMF1 vector for the treatment of multiple sulfatase deficiency

Abstract

Background: Multiple Sulfatase Deficiency (MSD) is a rare inherited lysosomal storage disorder characterized by loss of function mutations in the SUMF1 gene that manifests as a severe pediatric neurological disease. There are no available targeted therapies for MSD.

Methods: We engineered a viral vector (AAV9/SUMF1) to deliver working copies of the SUMF1 gene and tested the vector in Sumf1 knock out mice that generally display a median lifespan of 10 days. Mice were injected as pre-symptomatic neonates via intracerebroventricular administration, or as post-symptomatic juveniles via intrathecal alone or combination intrathecal and intravenous delivery. Cohorts were assessed for survival, behavioral outcomes, and post-mortem for sulfatase activity.

Results: We show that treatment of neonates extends survival up to 1-year post-injection. Importantly, delivery of SUMF1 through cerebral spinal fluid at 7 days of age alleviates MSD symptoms. The treated mice show wide distribution of the SUMF1 gene, no signs of toxicity or neuropathy, improved vision and cardiac function, and no behavioral deficits. One-year post treatment, tissues show increased sulfatase activity, indicating functional SUMF1. Further, a GLP toxicology study conducted in rats demonstrates favorable overall safety of this approach.

Conclusions: These preclinical studies highlight the potential of our AAV9/SUMF1 vector, the design of which is directly translatable for clinical use, as a gene replacement therapy for MSD patients.

Fig. 1. ICV delivery of scAAV9/SUMF1 rescues neonatal lethality in Sumf1^(−/−) mice.

a Schematic of the self-complementary vector, scAAV9/SUMF1, carrying a full-length codon optimized human SUMF1 cDNA (hSUMF1opt) under the control of a CBh promoter and with a synthetic SV40 poly(A) tail. b Schematic of experimental design depicting AAV9 delivery and subsequent evaluation. scAAV9/hSUMF1 (2.8 × 10¹¹ vg/mouse) was delivered at P1 via ICV injection. Sumf1^(−/−) and Sumf1^(+/+) pups were randomized across experimental groups, and N = 8–10 by sex enrolled into each study arm. Animals were monitored for survival and growth and tested for behavioral outcomes at 1.5, 6, 12 months. Study end point was 14 months. c Kaplan-Meier survival curves represent the survival of mice. Log-rank (Mantel-Cox) test was applied for survival curve comparisons to Sumf1^(−/−) untreated group. d Enzymatic activity for ARSA, ARSB, ARSC, ARSL, IDS, and SGSH in brain and liver tissues of mice at study end point. Values represent the enzymatic activity as a percentage of average Sumf1^(+/+) sulfatase activity. e Behavior was assessed in Sumf1^(−/−) -AAV9/SUMF1 and Sumf1^(+/+)-untreated groups, starting at 1.5 month, and retested at 6 and 12 months. Rotarod data represents the average latency to fall in a three trials test. The average total distance traveled and vertical activity time in Open field, is reported at each time point. Spontaneous alternation is indicated as the average percentage of spontaneous alternations in a Y-maze. All data points represent individual mouse, bar graphs indicate mean ± SEM. Groups were compared by two-way ANOVA with Šídák’s correction for multiple comparisons. Exact p value is indicated in the figure and Supplementary Data 1. All schemes were created with BioRender.com.

Fig. 2. IT delivery of scAAV9/SUMF1 extends life span of Sumf1^(−/−) mice in a dose-dependent manner.

a Sumf1^(−/−) mice were IT injected at P7 with scAAV9/SUMF1 at high dose (HD, 7 × 10¹¹ vg/mouse), 1:4 dilution (1.8 × 10¹¹ vg/mouse), 1:16 dilution (4.4 × 10¹⁰ vg/mouse), 1:64 dilution (1.1 × 10¹⁰ vg/mouse), vehicle, and untreated. Sumf1^(+/+) mice were injected IT at P7 with either HD scAAV9/SUMF1 or vehicle. Balanced numbers of females and males by genotype were randomized across all the experimental groups. Animals were monitored daily until 3 weeks age, for survival and growth, and then weekly until end point at 18 months. HD groups were tested for behavior at 3 time points. Created in BioRender.com (b), Survival data for the dose-response groups is represented with Kaplan-Meier curves. c To test the efficacy of a combined gene therapy treatment (IT/IV), Sumf1^(−/−) mice at P7, received dual IT and IV injections of scAAV9/SUMF1 (1.2 × 10¹² vg/mouse). Survival and growth were monitored daily until 3 weeks of age, then weekly until end point at 18 months of age. Created in BioRender.com (d), Kaplan-Meier survival curves are reported. Log-rank (Mantel-Cox) test was applied to comparisons between survival of IT single dose vs combined IT/IV (p = 0.4155) and combined IT/IV vs Sumf1^(−/−) -vehicle (p = 0.001). e Behavior testing of IT and IT/IV treated mice at 1.5, 6, and 12 months of age. Rotarod data is reported as the average latency to fall from a three-trial test. For open field test, average total distance traveled, and vertical activity time is reported. The spontaneous alternation test is indicated as the average percentage of spontaneous alternations in a Y-maze. Sulfatase activity was measured in f brains and g liver at the study end point. Data is represented as the percentage of sulfatase activity relative to the average levels in Sumf1^(+/+) mice. All data points represent individual mouse, bar graphs indicates mean ± SEM. Statistical significance was determined by two-way ANOVA with Šídák’s correction for multiple comparisons. p values are indicated and can be accessed in Supplementary Data 1.

Fig. 3. scAAV9/SUMF1 rescues cardiac and vision deficits in Sumf1^(−/−) mice.

a Heart function was assessed in vivo by electrocardiography of Sumf1^(−/−) mice treated with gene therapy as single intrathecal (IT) high (7 × 10¹¹ vg) and low dose (1.1 × 10¹⁰ vg) or combined IT/IV dose (1.2 × 10¹² vg). Sumf1^(+/+) control and Sumf1^(−/−) -vehicle mice were included. Bar graphs represent the average heart rate, measured at 9 and 12 months of age. b Visual acuity was tested by opto-kinetics. The bar graphs represent the frequency thresholds (cycles/degree), measured at 9 and 12 months. All data points represent individual mouse, and bar graphs indicates mean ± SEM. Statistical significance was determined by two-way ANOVA with Šídák’s correction for multiple comparisons, Sumf1^(+/+) -vehicle was used as reference for multiple comparison. P values are indicated and accessible in Supplementary Data 1.

Fig. 4. Vector biodistribution and expression following IT and combined IT/IV gene therapy.

a scAAV9/SUMF1 vector copy number was assessed in all the dosing groups by qPCR analysis of hSUMF1opt in genomic DNA from brain, spinal cord, and liver at study end point (14 months for ICV delivery, and 18 months for IT or IT/IV delivery). Data points represents individual mice, and bar graphs indicates mean ± SEM of vg copies/mouse diploid genome. b hSUMF1opt expression was determined at the mRNA level by RT-qPCR analysis of brain, spinal cord, liver, heart, lung, kidney, spleen, and muscle. Vector cDNA is reported as the number of vector copies per copy of host Gapdh. Data points represents individual mice, and bar graphs indicates the mean ± SEM, N = 5–9 per experimental group.

Fig. 5. SUMF1 positive cells in tissues following IT and combined IT/IV gene therapy.

The transduction efficiency and tissue distribution of scAAV9/SUMF1 was assessed by RNA ISH (RNAscope™) at end point (18 months old) on tissues from mice injected with scAAV9/SUMF1 via single-IT high dose (IT-HD, 7 × 10¹¹ vg), single diluted 1:4 dose (IT, 1.8 × 10¹¹ vg), or combined high dose IT/IV (IT/IV, 1.2 × 10¹² vg). Sumf1^(−/−) untreated control samples, were collected at an average age of 248 days. hSUMF1opt-specific probes were utilized to detect the frequency for transduced cells. Representative images and quantification of hSUMF1opt in a brain (cortex and hippocampus), b liver, c heart, and d eyes (GCL ganglion cell layer, INL inner nuclear layer, ONL outer nuclear layer). Scale bar represents 50 µm. Image quantification, box plots represents median, min and max of the average number of AAV9/SUMF1 positive cells as a percentage of the total number of cells per section.

Fig. 6. Dose-dependent reduction of GAGs from IT delivery of scAAV9/SUMF1.

GAG tissue accumulation was detected by Alcian blue staining in liver sections of mice IT injected with scAAV9/SUMF1; high dose (HD, 7 × 10¹¹ vg/mouse), 1:4 dilution (1.8 × 10¹¹ vg/mouse), 1:16 dilution (4.4 × 10¹⁰ vg/mouse), and 1:64 dilution (1.1 × 10¹⁰ vg/mouse). Sumf1^(−/−) and Sumf1^(+/+) vehicle-injected were controls. a Representative images of liver sections. Scale bar represents 50 µm. b Alcian blue quantification represented as percentage of the stained area. Box plots represent the median, min, and max.

Fig. 7. scAAV9/SUMF1 gene therapy ameliorates brain pathology in Sumf1^(−/−) mice.

Brain sections from Sumf1^(−/−) mice treated with scAAV9/SUMF1 as high dose IT (IT, 7 × 10¹¹ vg), diluted 1:4 (IT, 1.8 × 10¹¹ vg), combined high dose IT/IV (IT/IV, 1.2 × 10¹² vg), vehicle, and Sumf1^(+/+) vehicle-injected were immune-stained for a LAMP1 and b GFAP. Representative images of each stain are shown for the motor cortex and hippocampus. All images are presented at 40× magnification. Scale bar represents 50 µm. Image quantification represents the percentage of stained area for each marker. Data points represent individual mouse, and bar graphs indicates mean ± SEM, N = 2–9 per group. Statistical significance was determined by two-way ANOVA with Šídák’s correction for multiple comparisons relative to the Sumf1^(+/+) vehicle group. The adjusted p value is indicated.