Supplementary Material Brief Legends

SUPPLEMENTARY MATERIAL– BRIEF LEGENDS

Figure S1.Concentration of three forms of the RM prior to dispensing, determined by ddPCR using five PCR assays.

Figure S2. Homogeneity analysis of three forms of the RM.

Figure S3. Stability analysis of RM1 at 4°C, -80°C, -20°C and 40°C.

Figure S4. Stability analysis of RM2 at 4°C, -80°C, -20°C and 40°C.

Figure S5. Stability analysis of RM3 at 4°C, -80°C, -20°C and 40°C.

Figure S6. Results from PCR and DNA fragment size analysis (MultiNA) for representative samples from validation experiments.

Figure S7. DNA fragment size analysis (agarose gel) of representative samples from validation experiments.

Figure S8. Application of the developed RM to assess stability and robustness of the measurement system used in vector analysis.

Table S1. Ct values for pDNA-EPO and three forms of the RM at nominal concentrations 16 and 160 cp/well in Assays 1–5.

Table S2. Results of PCR analysis in transgene-specific PCR assays of all samples tested in method validation experiments.

SUPPLEMENTARY MATERIAL LEGENDS

Figure S1Concentration of RM1 (a), RM2 (b) and RM3 (c) prior to dispensing, determined by ddPCR using five PCR assays. Five individual aliquots of each form of the RM were analysed using five PCR assays (targeting different ARS or the plasmid backbone), each in replicates of six. The values are the average of individual measurements in each PCR assay; the error bars indicate standard deviation (SD) (n=6). For each form of the RM, the grand mean of all five concentration values from five assays (n=30) with its expanded uncertainty (k=2.06 for degrees of freedom of 25) are shown as a solid and dotted lines, respectively; theseare also shown in the insert tables. RM1 and RM2 were analysed by Assays 2–5 targeting the RS and by the pUC assay (Assay 6). RM3 was analysed by Assays 1–5 for the RS. RM, reference material; ARS, assay reference sequence; RS, reference sequence; U, expanded uncertainty; U (%), relative expanded uncertainty; cp, copies.

Figure S2 Homogeneity analysis of RM1 (a), RM2 (b) and RM3 (c). The values for sub-sample 1 (open diamonds) and sub-sample 2 (filled diamonds) are the average concentration and their SD (n=3) for each sub-sample from one vial. The average concentration (xchar) obtained from characterisation of the whole batch (190 vials) of the particular form of the RM with its expanded uncertainty (see Methods) are shown as the solid and dotted lines, respectively. IQ refers to the concentration from the initial quantification using five PCR assays and its expanded uncertainty (as shown in Figure S1). The p-value in the F-test for RM1, RM2 and RM3 were, respectively, 0.52, 0.14 and 0.16. RM, reference material; SD, standard deviation; cp, copies.

Figure S3 Stability analysis of RM1 at 4°C (a), -80°C (b), -20°C (c) and 40°C (d). At each time point, except the time zero in (a–c), the values are the average concentration values with their SD for three vials, each analysed in triplicate (n=9) by the assay described in Methods. The value for zero time point in (a–c) is the average concentration value with its SD (n=60) obtained from characterisation of the whole batch in the homogeneity study. In (a), the concentration ranges at zero and two week time points overlap, although it is not clearly seen from the graph, and the concentration ranges were both within the final uncertainty of the RM.The short-term stability study at 40°C (d) was performed separately from stability testing at three other temperatures. This study was delayed by six months and, as a result, its zero time point was different from the zero time point for other temperatures. RM, reference material; SD, standard deviation, cp, copies.

Figure S4 Stability analysis of RM2 at 4°C (a), -80°C (b), -20°C (c) and 40°C (d).See the legend to Figure S3 for details. The insert graph shows the concentration values obtained by all six assays at 12 weeks.

Figure S5 Stability analysis of RM3 at 4°C (a), -80°C (b) and -20°C (c). See the legend to Figure S3 for details.

Figure S6 Results from PCR and DNA fragment size analysis (MultiNA) for representative samples from validation experiments.Results for one PCR replicate of each of four samples of blood from one donor in one assay (Assay 2) are shown. (a) Analysis of the transgene in nonviral vector in plasma. (b) Analysis of the transgene in viral vector in plasma. (c) Analysis of the transgene in viral vector in peripheral blood mononuclear cells (PBMC). ‘Positive+RM’ and ‘Negative+RM’ are the same samples as,respectively, ‘Positive contaminated with RM’ and ‘Negative contaminated with RM’. RM, reference material.

Figure S7 DNA fragment size analysis (agarose gel) of representative samples from validation experiments. (a) Analysis of the transgene in nonviral vector in plasma. (b) Analysis of the transgene in viral vector in plasma. (c) Analysis of the transgene in viral vector in peripheral blood mononuclear cells (PBMC).RM, reference material; pDNA (or pD) and AAV are, respectively, plasmid oradeno-associated virus vector incorporating the EPO (erythropoietin) transgene; NEG, negative sample; NTC, no template control; NEC, negative extraction control.

Figure S8 Application of the developed RM to assess stability and robustness of the measurement system used in vector analysis. (a) Longitudinal analysis of PCR positive control over the period of 1.5 years.The values are the average Ct values and the range between the highest and the lowest Ct values for triplicates obtained in independent experiments performed over the period of 1.5 years with various spacing (not reflected in the scale of the X-axis) between them. For most data points the Ct values range falls within the symbol. Note, Assays 3-5 were not performed at the time point 4. (b) Analysis of positive extraction control (PEC; in the context of this study, RM1, RM2 or RM3 at 250 cp/mL blood) using blood from multiple donors. The values are the average Ct and standard error of the mean (SEM) for each assay and each form of the RM pooled for all donors in the experiments to detect the EPO transgene in plasma performed over two months and using two lots of PCR master mixes. The number of donors was eight in the experiments with RM2 and six in the experiments with RM1 and RM3, except for results for Assay 1, which included data from eight donors. As each sample was analysed in PCR in triplicate, but rare replicates did not amplify (Figure 3), the number of individual results used to calculate the average Ct values and their SEM were as follows: n=8 for RM2 for Assays 1–5; n=6 for RM1 and RM3 for Assays 2–5 and n=8 for RM1 and RM3 for Assay 1.Ct, cycle threshold; RM, reference material; EPO, erythropoietin.

Table S1 Ct values for pDNA-EPO and RM1-3 at nominal concentrations 16 and 160 cp/well in Assays 1–5.

#The average Ct value is presented (n=2). The deviation between Ct values for duplicates is not shown for clarity of presentation; for majority of samples (33 out of 40), it was below 1 Ct, for the remaining seven samples, it was between 1.1 and 2.9.

*The amount of the template per PCR well calculated from gravimetric sample preparation.

^Due to an accidental sample mix up, only one replicate was analysed instead of two.

Abbreviations: pDNA-EPO, plasmid incorporating the EPO (erythropoietin) transgene; RM, reference material; cp, copies; Ct, cycle threshold.

Table S2Results of PCR analysisin transgene-specific PCR assays of all samples tested in method validationexperiments.

—Not applicable or not analysed

#Shown are the number of replicates for which amplification was observed out of replicates tested for all donors in method validation experiments.

%One replicate out of 16 returned positive PCR result (Ct 40.6). Since DNA fragment size analysis (Figure 4) confirmed that the amplified product corresponded in size to an amplicon from pDNA-EPO, this replicate was reported as contaminated with pDNA-EPO.Since the second PCR replicate of the sample did not amplify in PCR, the sample was reported negative.

^Positive samples were blood samples spiked with a vector incorporating the EPO transgene (pDNA-EPO or AAV-EPO). Negative samples were not spiked with the transgene. Contaminated samples were mimicked by spiking into positive or negative samples of one form of the RM.

*Number of replicates for Assay 1 was five and for all other assays – three.

Abbreviations: pDNA-EPOand AAV-EPO, plasmid or adeno-associated virus vector, incorporating the EPO (erythropoietin) transgene; RM, reference material; cp, copies; Ct, cycle threshold.

Figure S1Concentration of RM1 (a), RM2 (b) and RM3 (c)prior to dispensing, determined by ddPCR using five PCR assays.

Five individual aliquots of each form of the RM were analysed using five PCR assays (targeting different ARS or the plasmid backbone), each in replicates of six.The values are the average of individual measurements in each PCR assay; the error bars indicate standard deviation (SD) (n=6). For each form of the RM, the grand mean of all five concentration values from five assays (n=30) with its expanded uncertainty (k=2.06 for degrees of freedom of 25) are shown as a solid and dotted lines, respectively; these are also shown in the insert tables. RM1 and RM2 were analysed by Assays 2–5 targeting the RS and by the pUC assay (Assay 6). RM3 was analysed by Assays 1–5 for the RS. RM, reference material; ARS, assay reference sequence; RS, reference sequence; U, expanded uncertainty; U (%), relative expanded uncertainty; cp, copies.

Figure S2Homogeneity analysis of RM1 (a), RM2 (b) and RM3 (c).

The values for sub-sample 1 (open diamonds) and sub-sample 2 (filled diamonds) are the average concentration and their SD (n=3) for each sub-sample from one vial. The average concentration (xchar) obtained from characterisation of the whole batch (190 vials) of the particular form of the RM with its expanded uncertainty (see Methods) are shown as the solid and dotted lines, respectively. IQ refers to the concentration from the initial quantification using five PCR assays and its expanded uncertainty (as shown in FigureS1). Thep-value in the F-test for RM1, RM2 and RM3 were, respectively, 0.52, 0.14 and 0.16.RM, reference material; SD, standard deviation.

Figure S3Stability analysis of RM1 at 4°C (a), -80°C (b), -20°C (c) and 40°C (d).

a b

c d

At each time point, except the time zero in (a–c), the values are the average concentration values with their SD for three vials, each analysed in triplicate (n=9) by the assay described in Methods. The value for zero time point in (a–c) is the average concentration value with its SD (n=60) obtained from characterisation of the whole batch in the homogeneity study. In (a), the concentration ranges at zero and two week time points overlap, although it is not clearly seen from the graph, and the concentration ranges were both within the final uncertainty of the RM.The short-term stability study at 40°C (d) was performed separately from stability testing at three other temperatures. This study was delayed by six months and, as a result, its zero time point was different from the zero time point for other temperatures. RM, reference material; SD, standard deviation, cp, copies.

Figure S4Stability analysis of RM2 at 4°C (a), -80°C (b), -20°C (c) and 40°C (d).

a b

c d

See the legend to Figure S3 for details. The insert graph shows the concentration values obtained by all six assays at 12 weeks.

Figure S5Stability analysis of RM3 at 4°C (a), -80°C (b) and -20°C (c).

a b

See the legend to Figure S3 for details.

Figure S6 Results from PCR and DNA fragment size analysis (MultiNA) for representative samples from validation experiments. Results for one PCR replicate of each of four samples of blood from one donor in one assay (Assay 2) are shown. (a) Analysis of the transgene in nonviral vector in plasma. (b) Analysis of the transgene in viral vector in plasma. (c) Analysis of the transgene in viral vector in peripheral blood mononuclear cells (PBMC). ‘Positive+RM’ and ‘Negative+RM’ are the same samples as, respectively, ‘Positive contaminated with RM’ and ‘Negative contaminated with RM’. RM, reference material.

Figure S7 DNA fragment size analysis (agarose gel) of representative samples from validation experiments. (a) Analysis of the transgene in nonviral vector in plasma. (b) Analysis of the transgene in viral vector in plasma. (c) Analysis of the transgene in viral vector in peripheral blood mononuclear cells (PBMC). RM, reference material; pDNA (or pD) and AAV are, respectively, plasmid or adeno-associated virus vector incorporating the EPO transgene; NEG, negative sample; NTC, no template control; NEC, negative extraction control.

a bc

Figure S8Application of the developed RMto assess stability and robustness of the measurement system used in vector analysis. (a) Longitudinal analysis ofPCR positive controlover the period of 1.5 years.(b) Analysis of positive extraction control(PEC; in the context of this study, RM1, RM2 or RM3 at 250 cp/mL blood)using blood from multiple donors.

The values are the average Ct values and therange between the highest and the lowest Ct values for triplicates obtained in independent experiments performed over the period of 1.5 years with various spacing (not reflected in the scale of the X-axis) between them. For most data points the Ct values range falls within the symbol. Note, Assays 3-5 were not performed at the time point 4.

The values are the average Ct and standard error of the mean (SEM) for each assay and each form of the RM pooled for all donors in the experiments to detect the EPO transgene in plasma performed over two months and using two lots of PCR master mixes. The number of donors was eight in the experiments with RM2 and six in the experiments with RM1 and RM3, except for results for Assay 1, which included data from eight donors. As each sample was analysed in PCR in triplicate,but rare replicates did not amplify (Figure 3), the number of individual results used to calculate the average Ct values and their SEM were as follows: n=8 for RM2 for Assays 1–5; n=6 for RM1 and RM3 for Assays 2–5 and n=8 for RM1 and RM3 for Assay 1.Ct, cycle threshold; RM, reference material..

Table S1 Ct values for pDNA-EPO and RM1-3 at nominal concentrations 16 and 160 cp/well in Assays 1–5.

Template / Nominal amount of template (cp/well) / Amount of template (cp/well)* / Ct#
Assay 1 / Assay 2 / Assay 3 / Assay 4 / Assay 5
RM1 / 16 / 14 / 34.6^ / 34.8 / 33.1 / 34.7 / 36.7
RM2 / 13 / 35.7 / 36.4 / 34.7 / 37.1 / 36.4
RM3 / 14 / 34.9 / 35.3 / 34.3 / 35.8 / 36.2
pDNA-EPO / 16 / 35.2 / 33.7 / 32.8 / 33.6 / 35.0
RM1 / 160 / 160 / 30.9 / 31.1 / 29.7 / 31.2 / 33.6
RM2 / 136 / 32.0^ / 32.3 / 29.6 / 32.0 / 33.6
RM3 / 160 / 31.8 / 32.3 / 30.8 / 31.6 / 33.0
pDNA-EPO / 160 / 31.7 / 30.5 / 29.4 / 30.0 / 31.8

*The amount of the template per PCR well calculated from gravimetric sample preparation.

^Due to an accidental sample mix up, only one replicate was analysed instead of two.

Abbreviations: pDNA-EPO, plasmid incorporating the EPO (erythropoietin) transgene; RM, reference material; cp, copies; Ct, cycle threshold.

Table S2Results of PCR analysisin transgene-specific PCR assays of all samples tested in method validationexperiments.

Extraction kit / Extraction matrix / Full sample name^ / Expected result / Experimental PCR results# / number of donors
Assays 1 - 5 / Assay 1 / Assay 2 / Assay 3 / Assay 4 / Assay 5
Negative samples
QB / plasma / negative (extracted) / No Ct / 1/16% / 0/16 / 0/16 / 0/16 / 0/16 / —
Miniprep / PBMC / negative (extracted) / No Ct / 0/6 / 0/6 / — / — / — / —
blank / NEC (extracted) / No Ct / 0/12 / 0/8 / — / — / — / —
— / — / PCR NTC (not extracted) / No Ct / 0/14 / 0/12 / 0/6 / 0/6 / 0/6 / —
Positive samples
QB / plasma / negative contaminated with RM1 / Ct (+) / 15/15 / 9/9 / 9/9 / 9/9 / 9/9 / 5 or 3*
negative contaminated with RM2 / Ct (+) / 15/15 / 15/15 / 15/15 / 15/15 / 15/15 / 5
negative contaminated with RM3 / Ct (+) / 15/15 / 9/9 / 9/9 / 9/9 / 9/9 / 5 or 3*
positive (pDNA) / Ct (+) / 15/15 / 15/15 / 15/15 / 15/15 / 15/15 / 5
positive (pDNA) contaminated with RM1 / Ct (+) / 9/9 / 9/9 / 9/9 / 9/9 / 9/9 / 3
positive (pDNA) contaminated with RM2 / Ct (+) / 9/9 / 9/9 / 9/9 / 9/9 / 9/9 / 3
positive (pDNA) contaminated with RM3 / Ct (+) / 9/9 / 9/9 / 9/9 / 9/9 / 9/9 / 3
positive (AAV) / Ct (+) / 15/15 / 9/9 / 9/9 / 9/9 / 9/9 / 5 or 3*
positive (AAV) contaminated with RM1 / Ct (+) / 9/9 / 9/9 / 9/9 / 9/9 / 9/9 / 3
positive (AAV) contaminated with RM2 / Ct (+) / 9/9 / 9/9 / 9/9 / 9/9 / 9/9 / 3
positive (AAV) contaminated with RM3 / Ct (+) / 9/9 / 9/9 / 9/9 / 9/9 / 9/9 / 3
Miniprep / PBMC / negative contaminated with RM1 / Ct (+) / 12/12 / 12/12 / — / — / — / 3
negative contaminated with RM2 / Ct (+) / 12/12 / 12/12 / — / — / — / 3
negative contaminated with RM3 / Ct (+) / 12/12 / 12/12 / — / — / — / 3
positive (AAV) / Ct (+) / 12/12 / 12/12 / — / — / — / 3
positive (AAV) contaminated with RM1 / Ct (+) / 12/12 / 12/12 / — / — / — / 3
positive (AAV) contaminated with RM2 / Ct (+) / 12/12 / 12/12 / — / — / — / 3
positive (AAV) contaminated with RM3 / Ct (+) / 12/12 / 12/12 / — / — / — / 3

—Not applicable or not analysed

#Shown are the number of replicates for which amplification was observed out of replicates tested for all donors in method validation experiments.

^Positive samples were blood samples spiked with a vector incorporating the EPO transgene (pDNA-EPO or rAAV-EPO). Negative samples were not spiked with the transgene. Contaminated samples were mimicked by spiking into positive or negative samplesof one form of the RM.

*Number of replicates for Assay 1 was five and for all other assays – three.

Abbreviations: QB, QIAamp DNA blood midi kit, Miniprep,QIAprep spin miniprep kit,pDNA-EPO and AAV-EPO, plasmid or adeno-associated virus vector, incorporating the EPO (erythropoietin) transgene; RM, reference material; cp, copies; Ct, cycle threshold.