Fusion of Rabies Virus Glycoprotein or gh625 to Iduronate-2-Sulfatase for the Treatment of Mucopolysaccharidosis Type II

Abstract

Mucopolysaccharidosis type II (MPS II) is a lysosomal storage disease caused by a mutation in the IDS gene, resulting in deficiency of the enzyme iduronate-2-sulfatase (IDS) causing heparan sulfate (HS) and dermatan sulfate (DS) accumulation in all cells. This leads to skeletal and cardiorespiratory disease with severe neurodegeneration in two thirds of sufferers. Enzyme replacement therapy is ineffective at treating neurological disease, as intravenously delivered IDS is unable to cross the blood–brain barrier (BBB). Hematopoietic stem cell transplant is also unsuccessful, presumably due to insufficient IDS enzyme production from transplanted cells engrafting in the brain. We used two different peptide sequences (rabies virus glycoprotein [RVG] and gh625), both previously published as BBB-crossing peptides, fused to IDS and delivered via hematopoietic stem cell gene therapy (HSCGT). HSCGT with LV.IDS.RVG and LV.IDS.gh625 was compared with LV.IDS.ApoEII and LV.IDS in MPS II mice at 6 months post-transplant. Levels of IDS enzyme activity in the brain and peripheral tissues were lower in LV.IDS.RVG- and LV.IDS.gh625-treated mice than in LV.IDS.ApoEII- and LV.IDS-treated mice, despite comparable vector copy numbers. Microgliosis, astrocytosis, and lysosomal swelling were partially normalized in MPS II mice treated with LV.IDS.RVG and LV.IDS.gh625. Skeletal thickening was normalized by both treatments to wild-type levels. Although reductions in skeletal abnormalities and neuropathology are encouraging, given the low levels of enzyme activity compared with control tissue from LV.IDS- and LV.IDS.ApoEII-transplanted mice, the RVG and gh625 peptides are unlikely to be ideal candidates for HSCGT in MPS II and are inferior to the ApoEII peptide that we have previously demonstrated to be more effective at correcting MPS II disease than IDS alone.

INTRODUCTION

Mucopolysaccharidosis type II (MPS II) is a lysosomal storage disorder caused by mutations in the iduronate-2-sulfatase (IDS) gene. This leads to a deficiency of the IDS enzyme, causing a buildup of the glycosaminoglycans (GAGs) heparan sulfate (HS) and dermatan sulfate (DS) in the lysosomes. MPS II is characterized by skeletal abnormalities, cardiorespiratory complications and, in around two thirds of patients, neurological decline.¹ Current treatment options for MPS II, such as enzyme replacement therapy, are able to alleviate non-neurological pathology yet have little effect on neurological outcomes due to the blood–brain barrier (BBB) preventing large molecules from entering the central nervous system (CNS). Unlike in the related condition MPS I, hematopoietic stem cell transplant (HSCT) is largely ineffective at treating the CNS, probably due to the levels of enzyme secreted from engrafted cells being insufficient to treat the disease.²

We have demonstrated that a hematopoietic stem cell gene therapy (HSCGT) approach, utilizing a lentiviral vector expressing codon-optimized IDS was able to reduce HS storage and improve behavioral outcomes in the mouse model of MPS II. Furthermore, the addition of a BBB-targeting peptide sequence, a tandem repeat of the receptor binding domain of ApoE (ApoEII), further improved neurological correction over the native enzyme alone.³ Typically, HSCGT results in high enzyme levels (2- to 20-fold of normal) in the periphery, but only a fraction of normal enzyme levels in the brain.⁴ As such, the use of BBB-targeting peptides may facilitate the crossing of this peripheral enzyme into the brain.

In this study, we analyzed two other peptide sequences for their ability to improve uptake of IDS enzyme into the brain. “RVG” is based on the rabies virus glycoprotein, which binds to the N-acetylcholine receptor, whereas “gh625” is derived from the herpes simplex virus. Both these viral derived peptide sequences have been shown to facilitate transport of molecules across the BBB in several studies.^5
–7 We fused these peptides to IDS via a flexible linker and packaged them into third-generation lentiviral vectors. These vectors were used to transduce lineage-depleted murine bone marrow cells that were transplanted into conditioned MPS II mice. Mice were analyzed at 6 months post-transplant for behavioral and biochemical outcomes. The data were compared with LV.IDS and LV.IDS.ApoEII samples from a previous experiment as controls.

METHODS

Expression vectors

The brain-targeting peptide sequences RVG (YTIWMPENPRPGTPCDIFTNSRGKRASNG) and gh625 (HGLASTLTRWAHYNALIRAF) were inserted downstream of the human IDS cDNA in the third-generation lentivirus (LV) (pCCL.sin.cPPT.hCD11b.IDS.WPRE [LV.IDS]), using the long invariant linker (LGGGGSGGGGSGGGGSGGGGS) and were subsequently cloned into the same third-generation lentiviral backbone to create LV.IDS.RVG and LV.IDS.gh625.

LV production and titration

Lentiviral vectors were produced by transfection of HEK293T cells with pRSV-Rev, pMDLg/pRRE, pMD2.G and LV expression plasmid, and 7.5 mM polyethylenimine (PEI) (40 kDa; Polysciences, Inc., Warrington, PA).⁸ Lentiviral titration was performed as previously described.⁹

Uptake and transcytosis assays

Human CHME3 cells were transfected with 2 μg of CD11b.IDS or CD11b.IDS.RVG or CD11b.IDS.gh625 plasmid DNA using 7.5 mM high potency linear PEI (pH 7.4, 40 kDa; Polysciences) and 150 mM NaCl. Media was collected after 48 h. Human IDS protein was detected using the Human Iduronate-2-Sulfatase DuoSet enzyme linked immunosorbant assay (ELISA) kit according to the manufacturer's instructions (R&D Systems, Abingdon, United Kingdom). For uptake studies, ∼20 ng of IDS, IDS.RVG, or IDS.gh625 was added to bEND.3 cells following seeding at 10,000 cells/cm² for 3 days for monolayer culture. For transcytosis studies, ∼70 ng of IDS, IDS.RVG, or IDS.gh625 was added to bEND.3 cells following seeding at 50,000 cells/cm² for 5 days on the luminal side of ThinCerts (0.4 μm pore size, translucent polyester). For uptake assays, cell lysate was collected after 3 h, and for transcytosis assays, media was collected after 3 h. IDS expression in cell lysate/media was detected using the Human Iduronate-2-Sulfatase DuoSet ELISA Kit according to the manufacturer's instructions.

Mice and transplant procedures

Female heterozygous for the X-linked allele on a C57BL/6 background were obtained from Prof. Joseph Muenzer (University of North Carolina at Chapel Hill, NC) and bred with wild-type (WT) C57BL/6J males (Envigo, Alconbury, United Kingdom) to obtain affected males, carrier females, and WT animals (male and females). MPS II mice were backcrossed onto a Pep3 CD45.1 congenic background (B6.SJL-PtprcaPepcb/BoyJ) to distinguish donor and recipient cells following transplant. Wild-type littermates were used as controls throughout. Group sizes were n = 4–7, determined by power calculations. This study was approved by the ethics committee of the University of Manchester. They were conducted under license from the Home Office in accordance with British legislation and the European Communities Council Directive 86/609/EEC for the care and use of experimental animals.

For transplantation studies, total bone marrow mononuclear cells from MPS II/Pep3 mice (CD45.1) were isolated and lineage depleted using the murine lineage cell depletion kit (Miltenyi Biotec, Bisley, United Kingdom) in accordance with the manufacturer's instructions. Cells were stimulated using 100 ng/ml murine stem cell factor, 100 ng/ml murine fms-like tyrosine kinase-3, and 10 ng/ml recombinant murine interleukin-3 (PeproTech, Rocky Hill, NJ) for 3 h before transduction with a lentiviral vector (expressing IDS, IDS.ApoEII, IDS.RVG, or IDS.gh625) at a multiplicity of infection of 100.

Six- to 8-week-old mice were myeloablated using 125 mg/kg busulfan (Busilvex; Pierre Fabre, Boulogne, France) in five daily doses (25 mg/kg/day) via i.p. injection. Within 24 h of myeloablation, mice received 3–4 × 10⁵ lineage-depleted transduced hematopoietic stem cells through the lateral tail vein.

Chimerism analysis using flow cytometry

Engraftment of donor cells was assessed at ∼4 weeks post-transplant in peripheral blood, by staining leukocytes with anti-mouse CD45.1-PE (donor leukocytes), anti-mouse CD45.2-fluorescein isothiocyanate (FITC) (recipient leukocytes) (both Fisher Scientific Ltd.) in 5 μM ToPro3 Iodide (Thermo Fisher). Analysis was performed on a BD FACS Canto II flow cytometer (BD Biosciences).

X-ray imaging of live mice

Control and treated mice were anesthetized using isoflurane (induction: 3 L/min in pure O₂, maintenance: 1.5 L/min in pure O₂) and radiographed (45 keV) using the Bruker InVivo Xtreme System. X-ray images were analyzed using ImageJ.

Sample processing

At 8 months of age, mice were anesthetized and perfused with room temperature phosphate-buffered saline (PBS). One brain hemisphere was fixed in 4% paraformaldehyde (PFA) for 24 h, transferred to 30% sucrose, 2 mM MgCl₂/PBS solution for 48 h before freezing at −80°C. Samples of the brain, spleen, and liver were snap-frozen on dry ice. For enzyme activity assays, samples were homogenized and sonicated in homogenization buffer (0.5 M NaCl, 0.02 M Tris, 0.1% Triton-X100, pH 7). Genomic DNA used for organ vector copy number (VCN) analysis was extracted using GenElute Mammalian Genomic DNA Miniprep Kit (Sigma).

Enzyme activity assays

IDS enzyme activity was measured in a two-step protocol using the fluorescent substrate 4-methylumbelliferone-αIdoA-2S (Carbosynth) and Aldurazyme (Genzyme), as previously described.¹⁰ Starting material was standardized to 20 μg of total protein for plasma, 40 μg for the liver and spleen, and 60 μg for the brain using a bicinchoninic acid assay (Thermo Fisher). Fluorescence was measured using the BioTek Synergy HT plate reader (excitation: 360 nm, emission: 460 nm).

Immunohistochemistry

Free-floating immunohistochemistry was performed on 30 μm PFA-fixed coronal brain sections using rabbit anti-NeuN (1:1,000, ab177487; Abcam PLC, Cambridge, United Kingdom), rabbit anti-glial fibrillary acidic protein (GFAP) (1:1,500, Z0334; Dako, Stockport, United Kingdom), and rat anti-LAMP2 (1:500, ab13524; Abcam) primary antibodies using standard protocols. Isolectin B4 (ILB4, 5 μg/mL, L5391; Sigma) was visualized on 30 μm coronal brain sections using 3,3-diaminobenzidine substrate for 40 s (Vector, Peterborough, United Kingdom) using standard protocols.

Images were acquired on a 3D-Histech Pannoramic-250 microscope slide-scanner using a 20 × /0.30 Plan Achromat objective (Zeiss) with extended focus and the 4′,6-diamidino-2-phenylindole (DAPI), FITC, and tetramethylrhodamine filter sets. Snapshots of the slide-scans were taken using Case Viewer software (3D-Histech). Nonlinear adjustments were made to all immunofluorescence images equally to eliminate background; gamma 0.72, input levels 0–190. GFAP immunofluorescence was quantified using the ImageJ software on four sections per mouse for cortex, and one section per mouse for the amygdala (n = 3 per group). Counts of ILB4-positive cells were performed on four sections per mouse at 20 × magnification and counted manually using the ImageJ software.

Vector copy number

VCN was determined from harvested tissue using quantitative polymerase chain reaction (qPCR), as previously described.⁹

Statistics

Statistical analysis was performed using GraphPad Prism 7 software (La Jolla, CA). Following normality testing, two-tailed, parametric unpaired t-tests were applied for individual group comparisons. Either one-way ANOVAs (analyses of variance) followed by Tukey's multicomparisons test (parametric) or Kruskal–Wallis test (nonparametric) was performed for multigroup analysis. Significance was set at p < 0.05.

RESULTS

Outcomes at 6 months post-transplant in LV.IDS.RVG- and LV.IDS.gh625-treated mice were compared with control tissue from previously transplanted LV.IDS- and LV.IDS.ApoEII-transplanted mice.³

Addition of gh625, but not RVG, increases uptake into bEND.3 cells in vitro

To analyze changes in uptake and transcytosis in vitro, bEND.3 cells were used as a model of the BBB. Increases in intracellular IDS expression were seen in bEND.3 cells when cultured with IDS.gh265 compared with IDS and IDS.RVG (Fig. 1A). No increases, compared with IDS, were seen in transcytosis across bEND.3 monolayers when IDS.RVG and IDS.gh625 were applied (Fig. 1B).

Figure 1.

Uptake and transcytosis in bEND.3 cells in vitro. (A) Uptake into bEND3 cells. (B) Transcytosis across bEND.3 monolayers on transwell inserts. Data are displayed as mean ± SEM. **p < 0.01, compared with IDS via one-way ANOVA. ANOVA, analysis of variance; IDS, iduronate-2-sulfatase; SEM, standard error of the mean.

Transduction of lineage-depleted murine bone marrow gives detectable levels of IDS.RVG and IDS.gh625 in cells pretransplant

Following transduction of lineage-depleted bone marrow, an IDS enzyme activity assay and qPCR were used to assess the detectable IDS activity and VCN, respectively, in cells before transplant. Both groups produced detectable levels of IDS activity, with greater activity seen in the IDS.RVG-transduced cells (Fig. 2A) with the IDS.RVG group also demonstrating the highest VCN (Fig. 2B). Chimerism was analyzed using flow cytometry, 1 month post-transplant (Fig. 2C). Chimerism in IDS.gh625-treated animals (87.73%) was higher than IDS.RVG-treated animals (82.03%), but this was not significant when analyzed with an un-paired t-test (p = 0.0774).

Figure 2.

Chimerism, IDS activity, and VCN in treated animals. (A) IDS activity in colonies pretransplant. (B) VCN in colonies pretransplant. (C) Chimerism in LV.IDS.RVG- and LV.IDS.gh625-treated animals. (D–F) Brain, liver, and spleen 8 months post-treatment. (G–I) VCN in the brain, liver, and spleen 8 months post-treatment. (J–L) IDS activity normalized to VCN in the brain, liver, and spleen. Data are displayed as mean ± SEM. **p < 0.01, ****p < 0.0001 compared with MPS II (IDS activity) or IDS (IDS activity/VCN) via one-way ANOVA. MPS, mucopolysaccharidosis; VCN, vector copy number.

HSCGT with IDS.RVG and IDS.gh625 in MPS II mice generates detectable IDS activity throughout the body

Samples of the brain, liver, and spleen were analyzed for IDS enzyme activity and detectable VCN 6 months post-transplant. Levels of IDS activity in IDS.RVG- and IDS.gh625-treated animals were about 4.08% and 1.84% of WT expression in the brain, respectively (compared with 19.54% and 21.38% for IDS and IDS.ApoEII, respectively; Fig. 2D). In the liver, expression was 173% and 68% of WT for IDS.RVG and IDS.gh625, respectively (compared with 2,073% and 2,379% for IDS and IDS.ApoEII, respectively; Fig. 2E). In the spleen, expression was 526% and 108% of WT for IDS.RVG and IDS.gh625, respectively (compared with 4,925% and 2,634% for IDS and IDS.ApoEII, respectively; Fig. 2F). Vector copies were detected in all organs, with IDS.RVG-treated animals demonstrating higher VCN in all organs (Fig. 2G–I).

When normalized to VCN, IDS activity in the brain was greatest in IDS.gh625-treated animals (however, this was not statistically significant via one-way ANOVA compared with IDS; Fig. 2J). In the liver, normalized IDS activity was significantly lower in both IDS.RVG- and IDS.gh625-treated animals compared with IDS-treated. IDS.ApoEII-treated animals had significantly greater normalized IDS activity than IDS-, IDS.RVG-, or IDS.gh625-treated animals via one-way ANOVA (Fig. 2K). Normalized IDS activity in the spleen was significantly lower in IDS.RVG- and IDS.gh625-treated animals compared with both IDS- and IDS.ApoEII-treated animals via one-way ANOVA (Fig. 2L).

LV.IDS.RVG and LV.IDS.gh625 mediate limited reduction in lysosomal accumulation in neurons throughout the brain

To determine the effects of increased IDS enzyme activity in the brains of transplanted MPS II mice 6 months post-transplantation, the lysosomal marker LAMP2 was used to stain coronal brain sections of control and treated MPS II mice for lysosomal compartment size. The neuronal marker “NeuN” used for substrate accumulation in neurons. WT animals displayed weak, punctate, and perinuclear LAMP2 staining that only partially colocalized with NeuN in the motor cortex. Untreated MPS II mice displayed strong colocalized staining of NeuN and LAMP2 in multiple areas of the brain, including the motor cortex, caudate putamen, hippocampus, and amygdala, suggesting a substantial lysosomal burden in neurons (Fig. 3A). In MPS II mice, LAMP2 quantification was increased 3.5- and 4.9-fold compared with WT. Treatment groups LV.IDS.RVG and LV.IDS.gh625 mediated significant reductions in LAMP2 staining in the amygdala (2.6- and 3.3-fold of WT levels) and a trend to reduction in the motor cortex (1.9- and 2.3-fold of WT levels) and the Amygdala (Fig. 3B and C).

Figure 3.

HSCGT with LV.IDS.RVG and LV.IDS.gh625 reduces neuronal LAMP2 in the amygdala. (A) Representative images of 30 μm brain sections of the motor cortex (M2), caudate putamen (both approximately −0.46 mm from bregma), hippocampus (CA3), and amygdala (both approximately −1.22 mm from bregma) stained with LAMP2 (lysosomal compartment; red), NeuN (neuronal nuclei; green), and DAPI (nucleus; blue) from control and treated mice. Representative images are shown following staining of four to seven mice per group, 20 × , nonlinear adjustments were made equally in all images to reduce background; gamma 0.80, input levels 0–190. Scale bar = 50 μm. LAMP2 immunofluorescence was quantified using ImageJ in the (B) cortex and (C) amygdala of MPS II mice 6 months post-transplant. Data are displayed as mean ± SEM. ***p < 0.001 compared with MPS II via one-way ANOVA. DAPI, 4',6-diamidino-2-phenylindole; HSCGT, hematopoietic stem cell gene therapy.

LV.IDS.RVG and LV.IDS.gh625 mediate improvement in neuroinflammation in MPS II mice

In MPS disorders, neuroinflammatory responses are driven by astrocytes, which can result in reactive gliosis, astrogliosis, and increased levels of inflammatory cytokines. GFAP (green), an astrocytic marker, and LAMP2 (red), a lysosomal marker, were used to stain coronal brain sections of control and treated MPS II mice. WT mice displayed minimal GFAP staining. However, in untreated MPS II mice, there was considerable staining of GFAP in the cortex, caudate putamen, hippocampus, and amygdala indicating astrogliosis (Fig. 4A). In untreated mice, strong colocalization of GFAP and LAMP2 was observed in the cortex, caudate putamen, hippocampus, and amygdala, which is indicative of lysosomal substrate accumulation in astrocytes, as well as neurons. GFAP quantification in untreated MPS II mice was 7-fold higher in the cortex and 8.1-fold higher in the amygdala relative to WT levels. LV.IDS.RVG and LV.IDS.gh625 demonstrate a significant reduction in the levels of reactive astrocytes that are present in the cortex (2.9- and 3.4-fold of WT levels) and amygdala (3.6- and 4.8-fold of WT levels) (Fig. 4B, C).

Figure 4.

LV.IDS.RVG and LV.IDS.gh625 significantly correct astrocytic lysosomal swelling and astrocytosis following HSCGT. (A) Representative images of 30 μm brain sections of the motor cortex (M2), caudate putamen (both approximately −0.46 mm from bregma), hippocampus (CA3), and amygdala (both approximately −1.22 mm from bregma) stained with GFAP (green), LAMP2 (lysosomal compartment; red), and DAPI (nucleus; blue) from control and treated mice. Representative images are shown following staining of four to seven mice per group, 20 × , nonlinear adjustments were made equally in all images to reduce background; gamma 0.80, input levels 0–190. Scale bar = 50 μm. GFAP immunofluorescence was quantified using ImageJ in the (B) cortex and (C) amygdala of MPS II mice 6 months post-transplant. Data are displayed as mean ± SEM. *p < 0.05, ****p < 0.0001 compared with MPS II via one-way ANOVA. GFAP, glial fibrillary acidic protein.

LV.IDS.RVG significantly reduces numbers of ILB4-postive cells, thereby improving neuroinflammation

In the brains of untreated MPS II mice, there is a significant increase in the number of ILB4-positive cells. A 23- and 34-fold increase in activated microglia is observed in the cortex and striatum, respectively. Therefore, this increase signifies substantial microgliosis and subsequent neuroinflammation in these animals. WT mice have minimal numbers of activated microglial cells in the cortex and striatum (Fig. 5A). Treated mice reduced ILB4 staining and activated microglia to 9- and 12-fold of WT levels in the cortex and 14- and 19-fold of WT levels in the striatum for LV.IDS.RVG and LV.IDS.gH625, respectively (Fig. 5B, C). This was a significant reduction for LV.IDS.RVG but not for LV.IDS.gh625. Levels of ILB4-positive cells were substantially reduced in treated mice compared with untreated MPS II animals.

Figure 5.

Significant reductions in number of activated brain microglial cells with LV.IDS.RVG HSCGT. (A) Representative images of 30 μm brain sections of the motor cortex and striatum stained with ILB4 (brown) to detect activated microglia from control and treated mice. Representative images are shown following staining of four to seven mice per group. (B, C) Cell counts of ILB4-positive cells from two sections per mouse in the cortex and striatum, 20 × . Scale bar = 50 μm. (Kruskal–Wallis test for nonparametric testing *p < 0.05.) Data are displayed as mean ± SEM. *p < 0.05, ***p < 0.001 compared with MPS II.

Skeletal rescue is observed with LV.IDS.RVG and LV.IDS.gh625 in MPS II mice

To collect data on the extent of skeletal symptoms, total body X-ray radiography was performed under anesthesia on control and treated mice. In MPS II mice, the widths of the zygomatic arches, humeri, and femurs are all increased over WT mice. Both LV.IDS.RVG and LV.IDS.gh625 were able to reduce zygomatic arch width to WT levels (Fig. 6A, D). Humerus widths for the treated groups were also reduced to near WT dimensions, but not completely normalized (Fig. 6B, E). Moreover, the femur widths in the LV.IDS.RVG-treated animals were significantly thinner than MPS II mice (Fig. 6C, F). LV.IDS.gh625 femur widths were not significantly reduced from MPS II.

Figure 6.

Skeletal defects are normalized by LV.IDS.RVG to wild-type levels and show reduction in bone width by LV.IDS.gh625 HSCGT. (A–C) Representative X-ray images of control and treated MPS II mouse craniums, humeri, and femurs. White arrows indicate where measurements are taken from. ImageJ software was used to analyze zygomatic arch widths (D), humerus widths (E), and femur widths (F). Data are displayed as mean ± SEM. *p < 0.05, **p < 0.01, ****p < 0.0001 compared with MPS II via one-way ANOVA.

DISCUSSION

HSCGT has shown promise in preclinical studies for MPS II as well as MPS I, MPS IIIA, and MPS IIIB.^3,8,11,12 In MPS II, both we and others have seen that the use of HSCGT with LV.IDS can alleviate the disease phenotype in MPS II mice; however, it does not achieve full correction.^3,13 In our previous study, we used IDS tagged to a BBB-crossing peptide (ApoEII) designed to improve uptake of enzyme into cells following transplant. LV.IDS.ApoEII demonstrated superior correction of behavior, GAG storage, and inflammatory markers compared with LV.IDS.³

In this study, we sought to compare two other peptide constructs for their ability to improve uptake of IDS across the BBB to treat MPS II following HSCGT. RVG and gh625 are peptides based on the RVG and herpes simplex virus, respectively, and both have demonstrated potential for increasing BBB uptake.^5
–7

HSCGT treatment with both constructs reduced lysosomal swelling in the brain cortex and amygdala, as demonstrated by LAMP2 staining (Fig. 3); however, only reductions in the amygdala were statistically significant. Reductions were also seen in astrocytosis, as demonstrated by GFAP staining, with IDS.RVG demonstrating a significant increase in reduction over IDS.gh625 in the amygdala (Fig. 4). Both IDS.RVG and IDS.gh625 facilitated reductions in microgliosis, as demonstrated by ILB4 staining (Fig. 5); however, only IDS.RVG produced a statistically significant result.

We are able to compare these data against those of Gleitz et al., as the experiments were performed identically. In comparison to the data presented in Gleitz et al., the correction seen in the brain is far less impressive with IDS.RVG or IDS.gh625 than with IDS.ApoEII. We saw partial correction with both IDS.RVG and IDS.gh625 compared with full correction with IDS.ApoEII of LAMP2 staining in the cortex and amygdala, astrocytosis in the cortex, and amygdala and microgliosis in the cortex and striatum.³ Therefore, the increased activity per VCN seen in the IDS.gh625 group is not translating to a better neuropathological outcome. IDS.RVG and IDS.gh625 also significantly reduced skeletal thickening in the zygomatic arches and humeri (Fig. 6D, E). Only IDS.RVG significantly reduced femur thickness (Fig. 6F). Skeletal outcomes in this study were similar as they were with IDS.ApoEII.³

Biochemical analyses of IDS.RVG- and IDS.gh625-treated animals were compared with samples reanalyzed from our previous study (IDS- and IDS.ApoEII-treated animals) as control samples. Analysis of IDS activity in the brain showed that transplanted cells transduced with IDS.gh625 produced less activity than those transduced with IDS.RVG (Fig. 2D). However, looking at the VCN analysis in the cells before transplant (Fig. 2B) and the VCN in Fig. 2G, this is likely to be down to a lower level of transduction at the start of the experiment. What is interesting is that, despite this lower level of activity, correction is still good in the brain. When normalized to VCN, it also seems that more IDS.gh625 is present in the brain than IDS, IDS.ApoEII, or IDS.RVG (Fig. 2J). This could potentially be due to an increase in uptake into the brain, which is consistent with uptake data seen in vitro in bEND.3 cells (Fig. 1A); however, if this were the case, an improvement in neurological outcomes might be expected.

It is also worth noting that the VCN analysis in the brain is more challenging than in other organs, due to the relatively small number of cells that engraft in the brain following transplant.⁴ A final explanation for this could be that the RVG and ApoEII tags (both of which produced less activity in the brain than native IDS when normalized to VCN) are interfering with the enzyme activity. While this may be the case for RVG, it is unlikely to be the case for ApoEII as we would expect to see reduced activity in the liver and spleen compared with native IDS and, in both tissues, IDS.ApoEII activity is equal or better than IDS. We feel that the data shown here indicate that the interaction between the tag and the enzyme is more important than the tag alone. Both gh625 and RVG have demonstrated the ability to increase transport across the BBB into the brain;^5
–7 however, when tagged to IDS, they appear to be ineffective. It is possible that to improve BBB crossing for lysosomal enzymes, appropriate tags will need to be tailored to the enzyme in question.

Overall, the neurological correction seen in these animals is limited and does not appear to be as great as in IDS.ApoEII-treated animals, as previously demonstrated. For IDS.gh625, this is despite a much greater IDS activity/VCN in the brain compared with previously published results.³ Therefore, despite the fact that both peptides were able to mediate somatic correction of skeletal phenotypes and partial correction of brain pathology such as lysosomal storage (LAMP2), and inflammation (GFAP and ILB4), RVG and gh625 are unlikely to be good candidates to improve HSCGT for MPS II patients compared with IDS.ApoEII.

In conclusion, the nature of the RVG and gh625 peptides, based on viral glycoproteins, could have contributed to the fact that although they provided encouraging data in some assays, in vitro and in vivo, neither one was as effective as the ApoEII peptide at ameliorating disease symptoms and correcting behavior in MPS II mice. It is likely that improving uptake of enzymes in the context of HSCGT can add significant benefit to the overall disease correction in the absence of improvements in blood–brain barrier crossing.

Footnotes

ACKNOWLEDGMENTS

We would like to thank the National MPS Society and the Isaac Foundation for supporting this work. We would also like to thank the Bioimaging Facility at the University of Manchester for their help with obtaining microscope images.

AUTHORs' CONTRIBUTIONS

S.R.W. performed the experiments, analyzed the data, and wrote the article. A.C., R.S., C.B., and G.T. performed the experiments and analyzed the data. S.E. performed the experiments, analyzed the data, and edited the article. G.F., C.O’.L., and J.C. performed the experiments. H.G., A.L., and R.H. performed the experiments and edited the article. B.W.B. obtained the funds, designed the study, analyzed the data, and edited the article.

AUTHOR DISCLOSURE

S.R.W., S.E., H.G., A.L., R.H., and B.W.B. are beneficiaries of a licensed program for HSCGT for MPS II from AvroBio. B.W.B. also has unrestricted grants funded by AvroBio.

FUNDING INFORMATION

This study was supported by the National MPS Society, The Isaac Foundation.

References

Wraith

. Lysosomal disorders. Semin Neonatol, 2002; 7(1):75–83.

Wood

, Bigger

. Delivering gene therapy for mucopolysaccharide diseases. Front Mol Biosci, 2022; 9:965089.

Gleitz

, Liao

, Cook

, et al. Brain-targeted stem cell gene therapy corrects mucopolysaccharidosis type II via multiple mechanisms. EMBO Mol Med, 2018; 10(7):e8730.

Wilkinson

, Sergijenko

, Langford-Smith

, et al. Busulfan conditioning enhances engraftment of hematopoietic donor-derived cells in the brain compared with irradiation. Mol Ther, 2013; 21(4):868–876.

Kumar

, Wu

, McBride

, et al. Transvascular delivery of small interfering RNA to the central nervous system. Nature, 2007; 448(7149):39–43.

Zadran

, Akopian

, Zadran

, et al. RVG-mediated calpain2 gene silencing in the brain impairs learning and memory. Neuromolecular Med, 2013; 15(1):74–81.

Valiante

, Falanga

, Cigliano

, et al. Peptide gH625 enters into neuron and astrocyte cell lines and crosses the blood-brain barrier in rats. Int J Nanomedicine, 2015; 10:1885–1898.

Sergijenko

, Langford-Smith

, Liao

, et al. Myeloid/microglial driven autologous hematopoietic stem cell gene therapy corrects a neuronopathic lysosomal disease. Mol Ther, 2013; 21(10):1938–1949.

Langford-Smith

, Wilkinson

, Langford-Smith

, et al., Hematopoietic stem cell and gene therapy corrects primary neuropathology and behavior in mucopolysaccharidosis IIIA mice. Mol Ther, 2012; 20(8):1610–1621.

10.

Voznyi

, Keulemans

, van Diggelen

. A fluorimetric enzyme assay for the diagnosis of MPS II (Hunter disease). J Inherit Metab Dis, 2001; 24(6):675–680.

11.

Holley

, Ellison

, Fil

, et al. Macrophage enzyme and reduced inflammation drive brain correction of mucopolysaccharidosis IIIB by stem cell gene therapy. Brain, 2018; 141(1):99–116.

12.

Visigalli

, Delai

, Politi

, et al. Gene therapy augments the efficacy of hematopoietic cell transplantation and fully corrects mucopolysaccharidosis type I phenotype in the mouse model. Blood, 2010; 116(24):5130–5139.

13.

Wakabayashi

, Shimada

, Akiyama

, et al. Hematopoietic stem cell gene therapy corrects neuropathic phenotype in murine model of mucopolysaccharidosis type II. Hum Gene Ther, 2015; 26(6):357–366.