|
|
Blood, Vol. 107, Issue 11, 4257-4265, June 1, 2006
Proteasome activity restricts lentiviral gene transfer into hematopoietic stem cells and is down-regulated by cytokines that enhance transduction
Blood Santoni de Sio et al.
107: 4257
Supplemental materials for: Santoni de Sio et al
Files in this Data Supplement:
- Table S1. P-values for testing the significance of each variable (fixed effects) when all other
variables are included (PDF, 13.3 KB) -
For the Competitive Repopulation Assay we established a statistical model for % FP, the percentage of SRC expressing the transgene. Such a percentage is the response variable of interest and its distribution may be affected by several variables, including the cytokine treatment (variable “cyt”, + or –), the marker (variable “dye,” YFP or GFP), the cell preparation (variable “batch”) and the xenotransplant model, (variable “tx-mod,” reference or competitively repopulated). Moreover, if there were an interaction between “cyt” and “tx-mod” then the difference in mean %FP (in the logodd scale) between presence and absence of cytokines would change depending on whether the transplant model is reference or competitively repopulated. Otherwise, absence of interaction would mean that the increase in mean %FP (in the logodd scale) from absence to presence of cytokines is the same whether the transplant model is reference or competitively repopulated.
Taking into consideration that each reference mouse generates only one measurement (either %YFP or %GFP), while each competitively repopulated mouse generates two non-independent measurements (%YFP and %GFP), we modelled the presence of within-mouse correlated pairs of observations treating the variable “mouse” as a random effect. According to the statistical literature (see Searle et al., 1992, for a comprehensive treatment), random effects represent specific sources of randomness in addition to natural variability underlying all biological experiments. Since %FP are percentages which are by definition constrained between 0 and 100, we transformed them into their logodds
lofp = log ( %FP/(100-%FP)).
A linear mixed effect model for response variable “lofp” was constructed with fixed effect variables “cyt, ” “dye, ” “batch” and “tx-mod” and random effect variable “mouse. ” The model was fitted using routine lme from the contributed package nlme (Pinheiro et al., 2000) to the free software R33.
Searle SR, Casella G, McCulloch CE. Variance Components. New York: Wiley; 1992
Pinheiro JC, Bates DM. Mixed-Effects Models in S and S-plus. New York: Springer; 2000
- Figure S1. Vector DNA Analysis of Competitively Repopulated Mice BM-derived CFC (JPG, 58 KB)
-
To assess the absolute frequency of vector-positive CFCs for each vector type, individual, well-isolated colonies were plucked 10-14 days after plating and lysed, as previously described16. Two primer sets were used, one specific for GFP (5’ CCCTCGTGACCACCCTGAC 3’ and 5’ TGCTCAGGGCGGACTGGGT 3’) and one specific for YFP (5’ ACCCTCGTGACCACCTTCGG 3’ and 5’ CTTTGCTCAGGGCGGACTGGTA 3’). PCR was carried out as previously described16. Temperature of annealing were 60°C for YFP and 66°C for GFP primer set. PCR products were analysed by agarose gel electrophoresis, yielding a 454 bp band for both primer sets. A) Agarose gel showing a representative analysis of ten colonies from the indicated mouse, scored for YFP (upper panel) and GFP (lower panel). B) Relative proportion of GFP and YFP-positive colonies. Values were calculated from the total number of colonies scored positive for transgene DNA for each mouse, and averaged. For reference, the average relative proportion of GFP and YFP-expressing cells measured in the mouse bone marrow by FACS, calculated from the data in Table 1 (Expr. ratio: expression ratio) is shown. Comparison between the two proportions did not show a statistically significant difference. P-values were calculated by signed rank two-sided tests.
- Figure S2. Cytokine effect on HSPC transduction by LV pseudotyped with different envelopes (JPG, 29.8 KB)
-
Human CD34+ cells were transduced with LV-GFP pseudotyped by the indicated envelope, with (black bar) and without cytokines (grey bar), and analyzed after 2 weeks of culture. A-MLV or RD114/TR pseudotyped LVs were produced using A-MLV envelope-encoding plasmid phCMV-10A1 or the RD114/TR envelope-encoding plasmid phCMV-RD114/TR (Sandrin et al., 2002) and used at a concentration of 3×107 TU/ml, preloaded for 2 hrs on non-tissue culture treated dishes coated with recombinant human Fibronectin Fragment (CH296 — Takara BIO Inc.) following manufacturer instructions.
Each bar represents the average frequency and standard error (two to six experiments performed, depending on the envelope) of GFP expressing-cells in suspension culture (left panel) and CFC assay (right panel). For suspension cultures p= 0.022 for the VSV comparison, 0.058 for the RD114/TR comparison, and 0.097 for the A-MLV comparison, calculated by signed rank two-sided tests. If vector entry were affected by cytokines, we would expect that the enhancement of transduction would be envelope-dependent and possibly abolished by changing it. Since we observed a higher transduction level for the protocol with cytokine stimulation for all pseudotypes tested, we concluded that the cytokines did not modulate a specific entry pathway.
Sandrin V, Boson B, Salmon P, Gay W, Negre D, Le Grand R, Trono D, Cosset FL. Lentiviral vectors pseudotyped with a modified RD114 envelope glycoprotein show increased stability in sera and augmented transduction of primary lymphocytes and CD34+ cells derived from human and nonhuman primates. Blood. 2002;100:823-832
|
|
