Hank H. Sun

A COMPARISON OF TWO VIRAL VECTORS FOR GENE DELIVERY INTO MOUSE BRAIN: IS LENTIVIRUS MORE TOXIC THAN ADENO-ASSOCIATED VIRUS? Hank H. Sun¹, Christine M. LaPash², Mark P. Goldberg². Biology Department, Washington University, St. Louis, MO¹; Neurology Department, Washington University Medical School, St. Louis, MO².

Lentivirus (LV) and adeno-associated virus serotype-2 (AAV2) are two vectors useful for transferring genes into the brains of live animals. They infect non-dividing cells, exhibit low toxicity, and show long-term in vivo gene expression. Our lab is currently designing LV and AAV2 vectors for transferring genes and labeling axons in rodents after stroke.

Previously LV and AAV2 constructs (LV-GFP and AAV2-GFP) had been designed to introduce the gene for green fluorescent protein (GFP) into mouse brain cells. Our lab showed that AAV2-GFP was more neuron-specific and more localized in the brain than was LV-GFP. However, the toxicity of these two viruses had yet to be determined.

Based on previous studies, we hypothesized that AAV2-GFP would be less toxic than LV-GFP when injected into the mouse brain. To test this, either AAV2-GFP or LV-GFP was stereotaxically injected into the left motor cortex of adult mice. Cortical saline injections were administered as a negative control. Brains were collected 7, 21, and 35 days post-injection (dpi). The brains were sectioned, and then a TUNEL stain was performed to label cells undergoing apoptosis. TUNEL⁺ cells (TUNEL label that co-localized with a Hoechst⁺ nucleus) were quantified within the cortex of saline- or virus-injected tissue ipsilateral to the injection site. Data suggested that both LV-GFP and AAV2-GFP induced some degree of cell death at all time points with respect to the control. TUNEL labeling showed a significant elevation of apoptosis for both LV-GFP and AAV2-GFP compared with saline (19±5 and 6±1 vs. 1.4±0.4 cells/mm²). There were fewer cells by day 21 (2.8±0.5 and 3.6±0.4 vs. 0.1±0.05 cells/mm²), and apoptosis remained low up to day 35 (3.9±0.5 and 4.5±0.6 vs. 0.16±0.07 cells/mm²), though for both vectors apoptosis was still significantly higher than for the saline control at both time points. To see if viral toxicity was related to transduction efficiency, the total number of TUNEL⁺ cells was normalized to the total number of GFP-expressing (virus-infected) cells for each time point. At 7 and 21 dpi there was no significant difference in apoptotic cell death between LV-GFP and AAV2-GFP (0.35±0.08 vs. 0.31±0.04; 0.07±0.01 vs. 0.09±0.01). At 35 dpi, however, AAV2-GFP showed more TUNEL⁺ cells statistically than did LV-GFP per GFP-infected cell (0.14±0.02 vs. 0.08±0.01), though further experiments may be necessary to confirm this (p<0.04).

This study showed that the injection itself caused some cell death, especially at 7 dpi. Toxicity was also highest at 7 days for both LV-GFP and AAV2-GFP, but was lowered by 35 days. Both viruses were more toxic than saline and were equally toxic at 7 and 21 dpi. But AAV2-GFP may be slightly more toxic than LV-GFP (relative to the number of infected cells) at later time points. Based on this data, either LV or AAV2 may be useful for labeling axons and transferring genes in rodent brain, but extra caution should be exercised when using AAV2 for periods of time greater than 21 dpi.