Non-invasive detection of adeno-associated viral gene transfer using a genetically encoded CEST-MRI reporter gene in the murine heart

Figures

• 

  Figure 1. Transfection of HEK293T cells using the AAV9 vector encoding for eGFP, V5 and LRP. (Left) Phase contrast image of HEK293T cells and (middle) corresponding fluorescence microscopy image showing expression of eGFP. (Right) Western blot of cell homogenates for cells infected with AAV9-LRP (+) or control uninfected cells (−) demonstrates positive staining for the V5 epitope. The position of the LRP-V5 band with respect to the ladder is influenced by the strongly positive charge of LRP compared to other proteins of similar size.

• 

  Figure 2. Direct injection of AAV9 encoding for the Lysine Rich Protein results in regionally specific CEST contrast at 45 days after injection. (A) Cardiac CEST image acquired following radiofrequency saturation at −3.76ppm (control excitation). The site of injection is indicated by the red arrow. (B) Corresponding cardiac CEST image acquired following saturation at +3.76ppm. The excitation of LRP and subsequent exchange of saturated magnetization results in a reduced steady state longitudinal magnetization, and thus signal intensity, in voxels containing LRP expressing cells. (C) CEST contrast calculated as MTRasym from images A and B is elevated in the septum and inferior wall about the inferior right ventricular insertion point (red arrow). Slight elevations in MTRasym are observed in the interventricular septum. (D) Immuno-fluorescent staining of the corresponding tissue section confirms expression of the LRP-V5 construct in the same tissue regions as observed using CEST MRI. (E) Injection of an empty AAV9 vector did not result in observable CEST contrast or (F) positive staining of isolated tissue sections. Higher magnification images of D and F are shown in Supplemental Figure 1.

• 

  Figure 3. Time-course of regional changes in CEST contrast. Box and whisker plots summarize measurement of regional MTRasym values at 15 and 45 days after intra-myocardial injection of either AAV9 encoding for LRP or an empty AAV9 vector (grey bar represents the mean value). At both time points, MTRasym values within tissues that positively stained for V5 expression were significantly higher than corresponding regions of interest from mice receiving an empty AAV9 injection. While mean MTRasym values in the AAV9-LRP group were not substantially different between days 15 and 45, (Right) individual mice within the AAV9-LRP group (squares) demonstrated diverse changes in MTRasym values between both time-points. The corresponding values in mice receiving empty AAV9 vectors (circles) were consistent over time.

• 

  Figure 4. Representative maps of MTRasym at 60 days after tail vein injection of either saline, empty AAV9, or AAV9-LRP. Myocardial MTRasym values remained low in mice receiving either saline or an empty AAV9 vector. Among mice receiving AAV9-LRP, the average response resulted in a large number of voxels with elevated MTRasym across the heart. However, variability was observed in both the spatial pattern and magnitude of MTRasym elevation among mice receiving AAV9-LRP. This variability can be appreciated by comparing the MTRasym maps generated in mice with the lowest change in MTRasym over baseline levels (Minimum LRP), the group average changes in MTRasym over baseline (Average LRP), and the maximal change in MTRasym over baseline levels (Maximum LRP).

• 

  Figure 5. Systemic administration of AAV9-LRP. (A) Box and whisker plots of mean myocardial MTRasym values reveal a similar absence of CEST contrast 24 hours after intravenous injection of either AAV9 or saline (−1.6 ± 3.9% AAV9-LRP, −2.1 ± 2.74% AAV9-Empty, −0.65 ± 2.3% Saline). In mice receiving AAV9-LRP the mean myocardial MTRasym values increased substantially by Day 60 (4.9 ± 4.3% AAV9-LRP, −1.0 ± 2.3 AAV9-Empty, −1.87 ± 2.1% Saline) and remained elevated at 90 days after injection (3.4 ± 3.1% AAV9-LRP, −0.6 ± 2.3 AAV9-Empty, 0.1 ± 2.3% Saline) in comparison to Day 1. (B) Representative western blots of mouse heart samples using a V5 antibody confirmed expression of LRP-V5 in mice exposed to the AAV9-LRP construct at Day 90. The corresponding full -length gels can be found in Supplemental Figure 2. Co-IP gels demonstrating reduced non-specific binding can be found in Supplemental Figure 2. (*P < 0.05 vs. Day 1 in same group, _P < 0.05 between groups at same time point).

• 

  Figure 6. Measures of ventricular structure and function were unchanged by LRP expression status. In both the direct injection (left) and systemic administration (right) protocols ventricular structure as assessed by septal wall thickness and the ratio of wall thickness to the inner radius (h/r ratio) was not significantly different between hearts in which LRP expression was induced (LRP), empty vector (Empty, E.V.), and saline controls. Ventricular systolic function assessed as peak fractional shortening was similarly unchanged between groups and over time. Anesthesia was maintained using 1.5% isoflurane in oxygen. Mean and standard deviation values for all graphed components can be found in Supplemental Table 3.