A lung tropic AAV vector improves survival in a mouse model of surfactant B deficiency

Surfactant protein B (SP-B) deficiency is an autosomal recessive disorder that impairs surfactant homeostasis and manifests as lethal respiratory distress. A compelling argument exists for gene therapy to treat this disease, as de novo protein synthesis of SP-B in alveolar type 2 epithelial cells is required for proper surfactant production. Here we report a rationally designed adeno-associated virus (AAV) 6 capsid that demonstrates efficiency in lung epithelial cell transduction based on imaging and flow cytometry analysis. Intratracheal administration of this vector delivering murine or human proSFTPB cDNA into SP-B deficient mice restores surfactant homeostasis, prevents lung injury, and improves lung physiology. Untreated SP-B deficient mice develop fatal respiratory distress within two days. Gene therapy results in an improvement in median survival to greater than 200 days. This vector also transduces human lung tissue, demonstrating its potential for clinical translation against this lethal disease.


Reporting for specific materials, systems and methods
We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material, system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response.
In the majority of experiments we chose 10 animals per treatment group as the sample size. A sample size of 10 was arbitrarily chosen for our first survival study using the low dose of AAV-mSPB (10^10vg/mouse). Because there was a statistically significant improvement in median survival using n=10 with this low dose of AAV-mSPB, we chose to continue using this sample size for all our structure/function as well as survival studies in which a higher dose of AAV-SPB was used in all cases. However, in some instances the sample size per group was larger (n=11) or smaller (n=5-9) based on animal availability.
No data was excluded from this study. All experiments related to this study are included in the manuscript or in the Supplementary Figures.
In all survival as well as structure and function experiments, we used an inducible compound transgenic SP-B deficient mouse model. We performed a total of 9 survival studies with AAV-murine SPB under various conditions including using different AAV dosages, different administration methods, with or without exogenous pulmonary surfactant as a vehicle (BLES), and different times of dox removal. We independently replicated the survival study in which an intermediate dose of AAV-mSPB (10^11vg/mouse) was administered with dox removal 29 days later ( Supplementary Fig 6c and Fig 6b). In the initial study ( Supplementary Fig 6c) median survival was 9 days and in the replication study (Fig 6b) median survival was 20 days demonstrating an improvement in median survival. We also independently replicated the survival study in which an intermediate dose of AAV-mSPB (10^11vg/mouse) was administered with dox removal 7 days later (Fig 6d and Fig 6f,  Supplementary Fig 8f). In the initial study (Fig 6d) median survival was 194 days and in the replication study (Fig 6f, Supplementary Fig 8f) median survival was 202 days demonstrating successful replication. We performed 5 survival studies with AAV human SPB using different dosages, with different times of dox removal, and in both neonatal and adult mice. We performed 1 lung structure and function study using AAV murine SPB, and 1 lung structure and function study using AAV human SPB. In all instances the findings were consistent with the AAVmurine SPB and -human SPB groups significantly improving structure/function and survival compared to negative control groups.
The Animal Care and Veterinary Service Committee at the University of Ottawa allows 4 mice to be housed per cage, and they are to be grouped with their litter-mates and be of the same sex. During weaning, mice were assigned a numerical ID and uniquely ear-notched by a research technician. The ear-notch corresponded with a particular numerical ID. The survival and structure/function studies all contained both control and treatment groups within a single cage. The treatment of mice (either control or AAV-mSPB/-hSPB) was assigned based on alternating numerical IDs. For example, in a cage of mice with the IDs 1, 2, 3, and 4: 1 and 3 would be administered AAV-SPB, while 2 and 4 would receive a negative control (ie 1×PBS). The order of surgical AAV administration was based on their numerical order: 1st: 1 (AAV-SPB); 2nd: 2 (1×PBS); 3rd: 3 (AAV-SPB); 4th: 4 (1×PBS). For the structure/function analyses using AAV-mSPB or -hSPB (Figs. 4,5,7), in a cage of 4 mice 2 were untreated and 2 treated with AAV-SPB in a similar fashion to the survival studies. The mice on dox were carefully chosen for similar dates of birth and sex. A detailed summary including dates of birth and sex for these studies are provided in Supplementary Table 4. For the mouse inflammatory cytokine studies, mice were randomly matched for sex and dates of birth between mice on dox and those treated with AAV-mSPB. In the PCLS studies using human fetal lung, there was no random allocation and covariates were not controlled. The fetal lung samples were chosen if they displayed no underlying anomalies by ultrasound, had no known preexisting medical conditions, and were given consent for research use by the patient after the elective procedure. However, with each human fetal lung sample (n=3 total), negative controls (untransduced PCLS) were always performed.
The decision for identifying the respiratory signs and symptoms for euthanasia were made by the staff at ACVS or by members of our group blinded to the identity and treatment groups of the animals. Once the animals were euthanized, we identified which treatment group the mouse belonged to based on their assigned ID (ear notch). In Supplementary Table 5 we provide the information on how many mice were found dead versus euthanized in each survival study. Much of our data was collected and processed by software and so blinding was not required including: IVIS imaging quantification by LivingImage 3.2 software; flow cytometry data collected by BDFACSDiva v8.0; RT-qPCR data processed by StepOnePlus software 3.2; flexiVent data collected by FlexiWare v7.0; and cytokine measurements performed by BDFACSDiva v8.0. However, the histological, immunofluorescence, electron microscope, and confocal images were obtained by operators not blinded to the treatment groups. We did not quantitate any data from these images and provided representative images. We attempted to represent images in a forthcoming manner. For example, we obtained scanned images of the whole lungs and presented as large fields of view of the lungs in the space available. For the IF images we were forthcoming when unexpected SP-B expression was observed ( Supplementary Fig 5d), and carried out new experiments to explain these findings (Fig. 5b). For our survival studies we presented all the incidences where untreated mice survived beyond the longest expected survival point of 11 days (Supplementary Fig 6).
The c-Myc antibody was validated for Western blotting by the manufacturer (cellsignal.com). It was validated for western blotting in multiple cell lines including c-Myc knockout HEK293 cells. Reactivity for human, mouse and rat. Recommended 1:1000 dilution for western blotting. The CD31-BV421, CD45-AF647, and EpCAM-PE-Cy7 antibodies were validated for flow cytometry by the manufacturer (biolegend.com). All 3 antibodies have their lots quality control tested by immunofluorescent staining with flow cytometry analysis. All 3 antibodies are reactive for mouse. The recommended use for CD31-BV421 is 5"L per million cells in 100"L staining volume, and for both CD45-AF646 and EpCAM-PE-Cy7 is #0.25 per 10E6 cells in 100"L volume. The proSP-C and SP-B antibodies were validated for immunofluorescence using lung tissues from mice with homozygous Sftpc (Supplementary Fig. 5f) and Sftpb ( Supplementary Fig. 5g) gene deletions respectively. The proSP-C antibody was also validated for ELISA, immunohistochemistry (IHC), and western blotting by Millipore Sigma. Reactive for human, mouse and rat. Recommended use for IHC is 1:2000 to 1:4000 in human lung tissue, 1:1000 to 1:2000 in adult and 1:500 to 1:1000 in fetal mouse lung. The SP-B antibody was validated for western blotting, ELISA, and IHC by Seven Hills Bioreagents. Reactive for human, mouse, cow, and sheep. Recommended use for IHC is 1:1000. The Luciferase and mCherry antibodies were validated for immunofluorescence by the manufacturer (novusbio.com). Both antibodies were validated for western blotting, immunofluorescence (IF), and IHC. The luciferase antibody is reactive in human and mouse cell lines and its recommended use for IF is 20"g/mL. The mCherry antibody is reactive in human, mouse, rat, drosophila, and insects and its recommended use for IF is 1:500. The EpCAM antibody was validated for immunofluorescence by the manufacturer (abcam.com). It is validated for western blotting, IF, and IHC. Reactive for mouse, rat, dog, and human. Recommended use for IF is 1 to 5"g/mL.

HEK293 (ATCC; CRL-1573) and MLE12 (ATCC; CRL-2110)
The HEK293 cell line was purchased directly from ATCC and cytogenetic analysis (karyotyping) was performed by ATCC, but further authentication was not performed by our lab. The MLE12 cells were not authenticated.
The cell lines were tested and were all negative for mycoplasma contamination.
None used.
We used a compound transgenic laboratory mouse (FVB/N strain) that conditionally expresses SP-B under the control of the rtTA protein in AT2 cells. The generation of this mouse is described in detail in Melton KR et al., Am J Physiol Lung Cell Mol Physiol 285: L543-L549 (2003). All attempts were made to use equal numbers of males and females, and use mice between the ages of 6 to 8 weeks at the time of AAV administration. Occasionally we used older mice if we needed larger sample sizes, or younger mice when we performed neonatal survival studies (P8 to P15). For the structure/function and survival studies a total of 276 inducible SP-B mice were used comprising of 124 males, 138 females, and 14 of unknown sex. A detailed summary of the sex and date of birth of each animal used in each experiment in this study is listed in Supplementary Tables 4 and 5. The preceding experiments were performed at the University of Ottawa animal facility where the animals were on a 12 h on, 12 h off light cycle, room temperature was 21C, and humidity was 45%. These are all within the range of guidelines set by the Canadian Council on Animal Care. We also used five male 8week old Rosa26-Flox/LacZ (Strain: B6.129S4-Gt(ROSA)26Sortm1Sor/J; strain code 003474) mice purchased from Jackson Laboratory (Bar Harbor, ME) for targeting studies to determine the tissues and cell-types transduced by AAV6.2FF. These experiments were Three human fetal lung tissue samples were used having gestational ages of 16+0 (n=1) and 16+3 (n=2) weeks. The sex, genotype, diagnosis, and treatment categories for all human fetal lung samples were unknown.
Human fetal tissue samples were obtained after elective abortions, and after parental consent from donors. The fetuses contained no anomalies as assessed by ultrasound, and patients with known preexisting medical conditions such as HIV were excluded. The sex of the fetus was unknown. Even if there was a selection bias, there were always some PCLS from each fetus that was used as a negative control (ie untransduced).
All human fetal tissues in this study were obtained after receiving institutional review board approval from The Ottawa Hospital (Research Protocol 20170603-01H).
Eight days after IT administration of 10^11 vg of either AAV-Luc or AAV-GFP, mice were intraperitoneally injected with heparin, and euthanized. Their thorax region was surgically opened, and the lung vasculature was perfused with 1×PBS (no Ca2+ and Mg2+) and undiluted heparin to remove red blood cells (RBC). Following perfusion, an enzyme mix containing neutral protease, Collagenase I, DNase I in DPBS with Mg2+/Ca2+ was instilled into the lungs. The lungs were removed and placed in the enzyme mix for 1 hr. All traces of the trachea, bronchi, and heart tissue were removed, and then the lung lobes were dissociated into pieces and passed through a cell strainer. The filtered lung suspensions were collected and washed. Cold RBC lysis buffer was added to the cells before the total cell number was counted for each sample. A total of 5 × 10^5 cells were placed into individual wells of a 96-well plate for staining. All samples were blocked with Heavy chain (Fc) block in the dark. The samples were resuspended in FACS buffer with their respective antibodies and incubated for 30 min. The cells were washed, and then resuspended in FACS buffer plus 7-AAD for 5 to 10 min before flow cytometry. No stain, single stain, and fluorescence minus one control stains were all carried out.
The samples were analyzed on the BeckmanCoulter MoFlo XDP flow cytometry machine.
Sample compensation and data was analyzed using FlowJo version 10.
A total of 6 mice were used with n=2 controls injected with AAV6.2FF-Luciferase, and n=4 treatments injected with AAV6.2FF-GFP. In the 2 control mice, one mouse had 330,022 single viable cells isolated from their lungs with 15,305 CD45-/EpCAM++, and 5 cells CD45-/EpCAM++/GFP+; and the second mouse had 258,617 single viable cells isolated from their lungs with 13,813 CD45-/EpCAM++, and 4 cells CD45-/EpCAM++/GFP+. In the 4 treatment mice, one mouse had 271,917 single viable cells isolated from their lungs with 13,329 CD45-/EpCAM++, and 2,437 cells CD45-/EpCAM++/GFP+; the second mouse had 322,449 single viable cells isolated from their lungs with 35,502 CD45-/EpCAM++, and 11,142 cells CD45-/EpCAM++/GFP+; the third mouse had 194,580 single viable cells isolated from their lungs with 6,176 CD45-/EpCAM++, and 1,024 cells CD45-/ EpCAM++/GFP+; and the fourth mouse had 206,691 single viable cells isolated from their lungs with 7,806 CD45-/EpCAM++, and 1,565 cells CD45-/EpCAM++/GFP+. The raw data files for the flow cytometry experiment are provided in Supplementary Table 1 and the Source Data Files.
The cell populations are represented as contour plots, with outlier events greater than 2% represented as dots. The gating was set as follows: i; Forward Scatter (FSC) vs Side Scatter (SSC) to denote the total cell population from dissociated mouse lungs as represented by size and intracellular granularity respectively. ii; Doublet exclusion to ensure that only single cells are