Volume 8, Issue 2, Pages 332-341 (August 2003) Development of a sensitive assay for detection of replication-competent recombinant lentivirus in large-scale HIV-based vector preparations  Paul Escarpe, Nathalie Zayek, Peggy Chin, Flavia Borellini, Romain Zufferey, Gabor Veres, Veronique Kiermer  Molecular Therapy  Volume 8, Issue 2, Pages 332-341 (August 2003) DOI: 10.1016/S1525-0016(03)00167-9 Copyright © 2003 The American Society of Gene Therapy Terms and Conditions

FIG. 1 RCL assay principle. The RCL-permissive cell line C8166-45 is exposed to the vector sample for 7 days to maximize the chances of RCL infection and amplification. The culture is then diluted regularly for six passages to allow RCL amplification. Samples of culture supernatant are collected at each passage and assayed for viral replication by measure of the p24 capsid protein concentration. Molecular Therapy 2003 8, 332-341DOI: (10.1016/S1525-0016(03)00167-9) Copyright © 2003 The American Society of Gene Therapy Terms and Conditions

FIG. 2 Sensitivity of the RCL assay. (A) Spike and recovery of the positive control. Six 30-ml unpurified vector samples, each containing 1.5 × 107 transducing units of lentiviral vector (i.e., 1.1 × 1010 vector particles or 930 ng p24), were spiked with 1 TCID50 (10 fg p24) of the RCL positive control and assayed for RCL. The p24 concentration of culture supernatant samples collected at each passage is shown. (B) Amplification of the positive control in the absence of vector. Six RCL assay cultures were set up without vector samples, inoculated with 1 TCID50 of RCL positive control, and processed as in A. (C) Assay of the vector sample in the absence of positive control. Three 30-ml vector samples, identical to those tested in A but not spiked, were assayed for RCL as in A. Molecular Therapy 2003 8, 332-341DOI: (10.1016/S1525-0016(03)00167-9) Copyright © 2003 The American Society of Gene Therapy Terms and Conditions

FIG. 3 RCL testing of a large-scale vector production lot. (A) Test results. One 14-liter production lot (LPR25) was tested for RCL. The vector concentration in this lot was 3.4 × 106 transducing units/ml (587 ng p24/ml). A 460-ml sample of the bulk unpurified lot of vector was split into 14 32-ml parts, each containing 1.1 × 108 TU of vector. Fourteen replicate RCL assays were performed with 32-ml cultures. The p24 concentrations in samples collected at each passage of the cultures are shown. (B) Representative spike/recovery controls. As a control for the testing, six samples identical to the 32-ml test samples were spiked with 1 TCID50 of the positive control virus and assayed for RCL. One of six cultures became productively infected. The growth kinetics of the virus in this culture is shown by the open squares. The filled squares stand for a representative culture that did not become productively infected. (C) Representative positive control. Six additional cultures were set up similarly without vector sample and inoculated with 1 TCID50 of the positive control virus. In this case also, one of the six cultures became productively infected with the positive control; the growth kinetics of the positive control in that culture is shown by the open triangle. The filled triangles stand for a representative culture that did not become productively infected. Molecular Therapy 2003 8, 332-341DOI: (10.1016/S1525-0016(03)00167-9) Copyright © 2003 The American Society of Gene Therapy Terms and Conditions

FIG. 4 (A) Sensitivity of the VSV G PCR detection method. A cell clone, 52G, containing approximately 7 copies of the VSV G gene was mixed with C8166-45 cells in proportions ranging from 10 to 0.0001% 52G cells. Genomic DNA was extracted from the mixed-cell populations and analyzed by quantitative PCR. PCRs were performed in triplicate with 1 μg of DNA per reaction (the equivalent of 1.4 × 105 cells). The PCR readout expressed in VSV G copies per reaction, or 1.4 × 105 cells, is plotted against the number of VSV G-containing cells per reaction. The average of six independent DNA extractions is shown. The threshold of detection was determined to be 10 copies per reaction and is indicated on the graph. (B, C, and D) Detection of VSV G-containing cells in three RCL assays. During the testing of three large-scale vector lots described in Table 3, C8166-45 cells from some of the RCL test cultures were collected at each passage. VSV G sequence in genomic DNA extracted from these cells was quantified by PCR. The average of three PCRs per sample is shown and expressed as VSV G copies per PCR (1.4 × 105 cells). Each curve represents the evolution of VSV G levels over time in one test culture. The assays were set up with the following amounts of vector per culture: (B) vector lot LPR38, 1.66 × 108 TU (2.8 μg p24), (C) vector lot LPR28, 1.1 × 108 TU (18.9 μg p24), (C) vector lot LPR42, 2.2 × 108 TU (10.4 μg p24). Molecular Therapy 2003 8, 332-341DOI: (10.1016/S1525-0016(03)00167-9) Copyright © 2003 The American Society of Gene Therapy Terms and Conditions