LG631 Preclinical Studies

LV Vector Insertion Site Analysis of LG631 - Lentiviral vector expressing P140K MGMT Gene
The insertion site data presented here is consistent with a large body of published data showing that lentiviral vectors do not preferentially insert at transcription start sites. Data below represents analysis of insertions sites from CFU derived from human CD34 cells transduced with research grade lentiviral P140K MGMT. Post transduction cells were either 1) treated with BG+TMZ and cultured in vitro (Table 1 or Table 2) transplanted/engrafted into NOD SCID mice (Table 2).

CFU treated in vitro: Human mobilized peripheral blood CD34+ cells were transduced and cultured in vitro followed by treatment with 10 μM BG + 200 μM TMZ or left untreated. Individual 14 day CFU were then assessed for insertion sites using LAM PCR (per method described by Von Kalle et al). Genomic DNA was isolated from individual CFU using the Gentra Systems DNA isolation kit. One hundred ng of DNA from each fraction was amplified at the genomic-proviral junction sequence by repeat primer extension using an LTR specific 5’ biotinylated primer. Biotinylated product was captured with avidin magnetic beads and the enriched target sequence subjected to second strand synthesis by random hexamer priming. The resulting double stranded DNA was digested with the restriction enzyme Tspl, which cuts within the genomic DNA approximately every 200 base pairs. An oligonucleotide cassette was ligated to the end of the Sse9l-digested fragments and subjected to two rounds of nested PCR using forward primers specific for the ligation cassette and reverse primers specific for the LTR. PCR products were separated on a high resolution agarose gel. Bands were quantified and cut from the gel for sequencing of the unique insertion sites. Our results are shown in Table 1 below. Data assessed relevant to insertional mutagenesis were: biological function of gene, distance from the transcription start site (TSS), and whether the vector inserted into and intron or exon. These data demonstrate that 85% of insertions were in RefSeq genes and all downstream of transcription start sites (82% <10KB away from the TSS) and primarily within introns, decreasing the possibility of activation of a proto-oncogene and disruption of gene function.

Table 1: In vitro LV insertion data

LG631 Vector Insertion Site Analysis in vitro: This table presents data from CFU derived from mobilized peripheral blood CD34 that were either treated in vitro with 10μM BG+200μM TMZ of left untreated. All insertions associated with RefSeq genes were downstream of transcription start sites and located within introns. A unique pattern was not detected in cells treated with BG+TMZ compared to untreated.

CFU derived from NOD SCID mice recipient of P140K MGMT-transduced cells:
Human CD34+ cells from umbilical cord blood were transduced with LV P140K MGMT at an MOI of 50 (two exposures over 48 hours) in media containing 100ng/ml each Flt3, SCF and Tpo with FBS 10%. After transduction, cells were washed and infused into NOD/SCID mice that received non-myeloablative conditioning of 3 Gy and 200μg CD122 anti NK-cell antibody. Mice from Cohorts 1-3 were treated with two cycles of 30 mg/kg and 7.5.BCNU on weeks 4 and 7 post transplant. The remaining cohorts were left untreated. Mice were either treated with 30 mg/kg BG + 7.5 mg/kg BCNU or left untreated. Insertions sites are presented in Table 2. 21 sequences were obtained from 4 mice (2 treated and 2 untreated). As described above, these results showed no tendency for clonal evolution or insertions into known genomic regions selective of proliferation. Of 10 unique insertions, 8 were identified as RefSeq genes. All 10 insertions were downstream of the TSS and all landed in introns but one.

Table 2: NOD SCID mice LV insertion data

LG631 Vector Insertion Site Analysis in vivo: This table presents data from human CFU harvested from NOD SCID transplant recipients in the absence or presence of in vivo selection. Integration site analysis in human CFU harvested from NOD SCID transplant recipients were similar to those obtained from post transduction CFU above. The majority of insertions were in introns and non-coding regions. Some insertions are represented multiple times due to engraftment and repopulation of select clones. We did not detect a unique pattern in mice subjected to drug selection compared to non-drug treated mice.