Cytoreductive conditioning intensity predicts clonal diversity in ADA-SCID retroviral gene therapy patients

Aaron R. Cooper, Georgia R. Lill, Kit Shaw, Denise A. Carbonaro-Sarracino, Alejandra Davila, Robert Sokolic, Fabio Candotti, Matteo Pellegrini and Donald B. Kohn

Data supplements

Article Figures & Data


  • Figure 1.

    Cancer-related genes are targeted and dysregulated, causing expansions. (A) Frequently targeted genes. Gene name size corresponds to frequency with which the gene’s transcription start site was the nearest transcription start site to an integration site. Gene names in red are associated with human or murine leukemias.38,39 (B) Integration clusters near MECOM among all patients. ISs in the sense and antisense orientation relative to MECOM are denoted by arrows. (C) A lower proportion of ISs was near the MECOM CIS region defined in the present analysis than in a retroviral WAS trial. The proportion of total integrations within the MECOM CIS area called by the kernel convolution approach was quantified for data sets from other clinical gene therapy trials. A Fisher exact test was used to compare the proportions, and P values are indicated. Error bars represent Gaussian approximation of a binomial 95% confidence interval.

  • Figure 2.

    Clonal composition and dynamics over time. Each colored area represents a unique integration site over time, and the total height of the colored areas at a given time point indicates the VCN as measured by ddPCR.

  • Figure 3.

    Clonal diversity indicated by integration sites among patients. (A) Shannon indices calculated using the numbers of unique integration sites and their readcounts over time among patients. (B) Chao2 estimates of total integration sites in patients calculated by repeated sampling and sequencing of integration sites in PBMC DNA samples among patients. Each point represents a unique PBMC DNA sample that was sampled 4 times.

  • Figure 4.

    Clones with integrations near MECOM, near sets of leukemia-related genes and near a set of retrovirally tagged cancer genes (RTCGD) are more abundant on average than other clones. (A) Clonal abundance estimation from sequence readcounts for clones with integrations near MECOM and LMO2. (B) Integration site-specific clonal tracking via ddPCR of clones with integrations near MECOM and LMO2.

  • Figure 5.

    High frequency and dominance of clone with integration on chr21 in patient 301. (A) IS-specific ddPCR clonal tracking of dominant clone in patient 301. (B) Total and IS-specific ddPCR VCN measurement in sorted lineages. (C) NKG2C+ population expansion in patient 301 compared with another ADA trial patient and 2 healthy donors. Whole blood was stained with CD3, CD19, CD11b, CD56, and NKG2C antibodies and analyzed via flow cytometry. A CD56-FITC vs NKG2C-PE plot is shown for CD3CD19CD11b cells gated for characteristic lymphoid forward and side scatter. (D) EBV copy number correlates with NK cell count in patient 301. EBV copy number was determined in PBMC samples via ddPCR for the BALF5 gene.

  • Figure 6.

    TCR diversity by high-throughput sequencing. (A) Chao2 estimates of total CDR3 rearrangements in PBMC samples. (B) CDR3 lengths for rearrangements involving V segment 12. (C) Linear regression of Chao2 estimates of integration site and CDR3 sequence total diversities. Dotted lines represent 95% confidence interval, and P value indicates probability that slope is zero.

  • Figure 7.

    Correlation of integration site and TCR rearrangement diversity with conditioning, cell dose, and age at treatment. (A-D) Dotted lines represent 95% confidence interval. (A-F) P value indicates probability that the observed values resulted from a relationship with slope zero.


  • Table 1.

    Top 20 CISs

    CIS window (hg19)No. ISsGenes in CISP
    Chr11:33785000-34075000155LMO2, FBXO3, CAPRIN1<2.002e-07
    Chr15:74955000-75560000134C15orf39, GOLGA6C, PPCDC, SCAMP5, RPP25, COX5A, FAM219B, MPI, SCAMP2, ULK3, CPLX3, LMAN1L, MIR4513, CSK, CYP1A2, CYP1A1, EDC3<3.359e-07
    Chr1:234570000-235225000118LOC100506810, LINC00184, IRF2BP2, LOC100506795, TARBP1<1.180e-07
    Chr12:11705000-12180000108ETV6, LOC338817<1.983e-07
    Chr20:52150000-52730000104ZNF217, SUMO1P1, MIR4756, BCAS1<4.352e-07
    Chr21:43255000-43735000100C21orf128, UMODL1, ZNF295-AS1, ZBTB21, ABCG1, C2CD2, TFF3, PRDM15<7.754e-07
    Chr6:36940000-3729000095PIM1, TBC1D22B, TMEM217, FGD2, MTCH1<1.558e-07
    Chr2:64785000-6545500094SERTAD2, LOC339807, AFTPH, LOC400958, SLC1A4, CEP68, RAB1A, ACTR2<1.115e-07
    Chr2:113310000-11372500093SLC20A1, FLJ42351, CKAP2L, IL1A, CHCHD5, IL1B, POLR1B, IL37<1.115e-07
    Chr6:2640000-303000092WRNIP1, MYLK4, SERPINB1, MIR4645, MGC39372, SERPINB9, SERPINB6, LINC01011, NQO2, HTATSF1P2<1.558e-07
    Chr9:20275000-2059000091MIR4473, MIR4474, MLLT3<2.379e-07
    Chr2:43030000-4351000090ZFP36L2, LOC100129726, THADA<1.115e-07
    Chr18:60445000-6104000090BCL2, KDSR, PHLPP1<3.502e-07
    Chr17:40330000-4073500087STAT5A, STAT5B, STAT3, GHDC, PTRF, HCRT, KCNH4, ATP6V0A1, MIR5010, NAGLU, HSD17B1, COASY, MLX, PSMC3IP, FAM134C<3.412e-07
    Chr5:75515000-7632000087S100Z, F2RL1, CRHBP, F2R, NCRUPAR, F2RL2, IQGAP2, SV2C<1.502e-07
    Chr7:2350000-304000084GNA12, AMZ1, TTYH3, IQCE, BRAT1, MIR4648, LFNG, CARD11, CHST12, EIF3B, SNX8<1.725e-07
    • IS, insertion site.

  • Table 2.

    Gene ontology (GO) enrichment analysis of genes in CISs

    GO termFDR
    Phosphorus metabolic process5.04E-07
    Phosphate metabolic process5.04E-07
    Protein amino acid phosphorylation6.24E-07
    Intracellular signaling cascade3.47E-05
    Lymphocyte activation2.41E-04
    Leukocyte activation8.17E-04
    Negative regulation of signal transduction0.001166391
    Cell activation0.001251908
    Enzyme linked receptor protein signaling pathway0.0016406
    Programmed cell death0.001732176
    Positive regulation of cellular biosynthetic process0.001827324
    Positive regulation of biosynthetic process0.00281247