G Model

ARTICLE IN PRESS

PBA-9998; No. of Pages 4

Journal of Pharmaceutical and Biomedical Analysis xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Journal of Pharmaceutical and Biomedical Analysis journal homepage: www.elsevier.com/locate/jpba

Short communication

A direct qPCR method for residual DNA quantification in monoclonal antibody drugs produced in CHO cells Musaddeq Hussain ∗ BioProcess Development, Biologics and Vaccines Research, Merck Research Laboratories, Kenilworth, NJ, USA

a r t i c l e

i n f o

Article history: Received 5 February 2015 Received in revised form 3 March 2015 Accepted 5 March 2015 Available online xxx Keywords: CHO (Chinese hamster ovary) cells Host residual DNA Monoclonal antibody KAPA protease Direct qPCR method

a b s t r a c t Chinese hamster ovary (CHO) cells are the host cell of choice for manufacturing of monoclonal antibody (mAb) drugs in the biopharmaceutical industry. Host cell DNA is an impurity of such manufacturing process and must be controlled and monitored in order to ensure drug purity and safety. A conventional method for quantification of host residual DNA in drug requires extraction of DNA from the mAb drug substance with subsequent quantification of the extracted DNA using real-time PCR (qPCR). Here we report a method where the DNA extraction step is eliminated prior to qPCR. In this method, which we have named ‘direct resDNA qPCR’, the mAb drug substance is digested with a protease called KAPA in a 96-well PCR plate, the protease in the digest is then denatured at high temperature, qPCR reagents are added to the resultant reaction wells in the plate along with standards and controls in other wells of the same plate, and the plate subjected to qPCR for analysis of residual host DNA in the samples. This direct resDNA qPCR method for CHO is sensitive to 5.0 fg of DNA with high precision and accuracy and has a wide linear range of determination. The method has been successfully tested with four mAbs drug, two IgG1 and two IgG4. Both the purified drug substance as well as a number of process intermediate samples, e.g., bioreactor harvest, Protein A column eluate and ion-exchange column eluates were tested. This method simplifies the residual DNA quantification protocol, reduces time of analysis and leads to increased assay sensitivity and development of automated high-throughput methods. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Many therapeutic proteins like monoclonal antibodies (mAb) are manufactured in CHO cells as host. The host cell derived impurities pose safety concerns and the regulatory agencies have defined acceptable levels of such impurities in the protein drug [1–3]. Therefore, it is imperative for the manufacturer to remove the host cell derived protein and DNA impurities from the protein drug and demonstrate the level in the purified drug substance. In Merck Research Laboratories, in cases of biologic drugs under development where the patient daily dose is unknown or too early to be fixed, the acceptance criteria of allowable host residual DNA is often set as ≤1.0 pg of DNA per mg of drug substance protein. As a result, use of a very sensitive method of DNA detection and quantification is needed, such as, quantitative real-time PCR (qPCR) [4]. However, the small amount of host residual DNA needs to be extracted from the protein drug sample prior to qPCR. Although the qPCR is a very sensitive method, the efficiency of DNA extraction impacts

∗ Tel.: +1 908 740 6434. E-mail address: [email protected]

the sensitivity and usefulness of the qPCR based host residual DNA quantification method. Therefore, a qPCR method free of a DNA extraction step is highly desirable. Here, such a method for host residual DNA is described where the DNA extraction step is eliminated; instead, the drug protein sample is digested with a specific protease after which the protease is heat denatured and the digest subjected to qPCR. The method, called direct resDNA qPCR method, is highly sensitive with high precision and accuracy over a wide analytical range. 2. Materials and methods 2.1. Materials Four different mAb drug substances were used for this study; two were IgG1 mAb and the other two were IgG4 mAb formulated in buffer containing histidine, trehalose or sucrose, and polysorbate 80 with pH close to 6.0. Total DNA from CHO cells was extracted using QIAamp DNA Mini kit from Qiagen (Valencia, CA). The qPCR oligonucleotide primers and probe were custom synthesized at Applied Biosystems (Foster City, CA). The KAPA Express Extract kit containing KAPA proteinase and buffer was purchased from KAPA

http://dx.doi.org/10.1016/j.jpba.2015.03.005 0731-7085/© 2015 Elsevier B.V. All rights reserved.

Please cite this article in press as: M. Hussain, A direct qPCR method for residual DNA quantification in monoclonal antibody drugs produced in CHO cells, J. Pharm. Biomed. Anal. (2015), http://dx.doi.org/10.1016/j.jpba.2015.03.005

G Model

ARTICLE IN PRESS

PBA-9998; No. of Pages 4

M. Hussain / Journal of Pharmaceutical and Biomedical Analysis xxx (2015) xxx–xxx

2

Biosystems (Boston, MA). TaqMan Universal PCR Master Mix with AmpErase UNG (Cat# 4304437) was from Life Technologies (Warrington, UK). MicroAmp Fast 96-well Reaction Plate and MicroAmp optical adhesive film from Invitrogen was used for PCR using

2.5. Data analysis Precision of determination of DNA quantity was measured from three replicate PCR wells and expressed as %RSD. The accuracy was determined by measuring the DNA spike recovery and expressed as %Recovery from the following calculation:

DNA quantity in spiked sample − DNA quantity in unspiked sample × 100 Spike amount AB 7500 Fast Real-time PCR system. PCR-grade water, molecular biology grade 1× TE buffer, pH 8.0 and 10× TE buffer, pH 8.0 were purchased from Sigma–Aldrich. MicroAmp 96-well Full Plate Cover was from Applied Biosystems (Part# N8010550). 2.2. KAPA digestion KAPA is a novel thermostable protease that has been designed for optimal tissue lysis and sample DNA preservation. However, it can be denatured by heating at 95 ◦ C [5]. The KAPA protease digestion of the sample was performed in PCR plate suitable for AB 7500 Fast Real-time PCR system. A serial dilution of the CHO DNA standard from 1E6 fg/1.0 ␮L to 0.1 fg/1.0 ␮L was prepared in 1× TE buffer in a 96-well sterile (ELISA) plate with cover. Each KAPA protease digestion reaction in the PCR well contained 3.0 ␮L PCRgrade water, 2.0 ␮L mAb sample (∼100 ␮g protein), 1.2 ␮L of 10× TE buffer, pH 8.0, 0.8 ␮L (0.8 Unit) of KAPA protease, and 5.0 ␮L of CHO DNA as spike or 5.0 ␮L of water for unspiked samples with a total volume of 12.0 ␮L. The plate was covered with MicroAmp 96well Full Plate Cover and placed in a BioRad T100 Thermal Cycler and incubated at 56 ◦ C for 60 min followed by 95 ◦ C for 10 min and held at 12 ◦ C. At this point the plate was ready for qPCR. 2.3. DNA extraction by AutoMate Express AutoMate Express is an automated DNA extractor from Applied Biosystems which uses pre-filled and sealed cartridges (PrepSEQ Express reagent cartridges, Cat# 4467071) and methodology involving proteinase K digestion of the sample, binding DNA to a coated magnetic resin, washing of the resin to remove undesirable compounds, and elution of DNA from the resin in a low-salt buffer. AutoMate Express instrument was used with PrepSEQ ResDNA software version 1.0.0 set up for 60 min lysis and 75 ␮L elution volume for 200 ␮L sample. 2.4. CHO qPCR The Alu repetitive elements were the target of the qPCR for quantification of CHO DNA. The sequence of the primers and probe were the following: forward primer, 5 -GGTCCTGAGTTCAATTCCCA-3 ; reverse primer, 5 -ATCTGCACACCAGAAGAGGG-3 ; and the probe, FAM-CAACTACATGGTGGCTCACAAC-TAMRA. A 10× stock of primers and probe was prepared containing 10 ␮M of each primer and 2.5 ␮M probe and was stored at −20 ◦ C in small portions in amber microtubes. Prior to qPCR, a PCR mix was prepared so that each reaction well contained 15 ␮L of 2× TaqMan Universal Master Mix and 3 ␮L of the 10× stock of primers and probe. The total qPCR reaction volume per well was 30 ␮L and the 96-well PCR plate was covered with optical adhesive sheet and placed in a AB 7500 Fast Real-time PCR system equipped with 96-well heating block and with AcuSEQ software version 2.0. The PCR cycling conditions were: 2 min at 50 ◦ C, one cycle; 10 min at 95 ◦ C, one cycle followed by 40 cycles each consisting of 15 s at 95 ◦ C and 1 min at 57 ◦ C. Each experiment contained at least three no-template controls (NTC) and a dilution series of CHO DNA in triplicate for standard curve.

3. Results and discussion In laboratories engaged in releasing purified drug substance in Merck, it is necessary to demonstrate presence of ≤1.0 pg of host residual DNA per 1.0 mg of the drug substance protein which requires a very sensitive method of DNA quantification like the qPCR. The full-length mAb, however, cannot be directly added to qPCR reaction and analyzed for DNA because of precipitation of protein during heating cycles of PCR resulting in a white opaque gel in the PCR well. One method, which is the most commonly used method for host residual DNA quantification, is the extraction of total DNA from the sample prior to qPCR analysis. There are many different extraction methods and kits currently available for use [6–8]. Success of such a method is dependent on the efficiency of extraction of presumably very small amount of DNA present in purified drug substance. Another method could be the digestion of protein to small fragments so that it remains in solution during PCR and there are no precipitates. In this case, the protease has to be removed or denatured effectively for successful PCR. We tried this approach with proteinase K, trypsin, papain and FABRICATOR enzymes with results showing complete loss of recoverable DNA. In continuation of such efforts, a protease called KAPA (sold in KAPA Express Extract kit) has recently been tried successfully and the results are described below. We tested the tolerance of the qPCR reaction to the KAPA buffer, along with other buffers, and heat denaturation of the KAPA enzyme. We added 5.0 ␮L of the following buffers to final 30 ␮L PCR reaction containing standard CHO DNA in different PCR wells: 1× KAPA buffer (10× supplied with the KAPA Express Extract kit); 1× TE (10 mM Tris, 1 mM EDTA, pH 8.0); Tris–NaCl (10 mM Tris, 5 mM NaCl, pH 8.0); 1× PBS (9.0 g/L NaCl, 144 mg/L KH2 PO4 , 795 mg/L Na2 HPO4 , pH 7.4). Results showed that the 1× KAPA, 1× TE and Tris–NaCl buffers had no deleterious effect on qPCR reaction. To test the heat denaturation of the KAPA enzyme, we added KAPA enzyme after heating at 95 ◦ C for 5, 10, 15 or 30 min to PCR reactions and found that heating for 10 or more minutes denatured KAPA enzyme completely (data not shown). We then tested for the efficiency of KAPA enzyme by adding 0.8 units of KAPA enzyme in TE buffer to different amounts of IgG1-1 drug substance protein spiked with about 2400 fg CHO standard DNA (shown in Table 1) in a 12 ␮L reaction volume and incubated at 56 ◦ C for 60 min for digestion followed by 95 ◦ C for 10 min for denaturation. After the qPCR, it was noticed that the PCR well containing 350 ␮g protein had some white precipitates while the other wells containing 0–250 ␮g protein were clear. Table 1 Effect of drug substance protein concentration on the direct resDNA qPCR method efficiency. IgG1-1 protein

Observed DNA, mean (fg)

Observed DNA, RSD (%)

Spike recovery (%)

0 ␮g 50 ␮g 100 ␮g 150 ␮g 250 ␮g 350 ␮g

1963.12 2086.91 1856.49 1473.56 978.68 366.85

1.64 3.75 2.37 2.05 4.89 28.99

81 86 76 60 40 15

Please cite this article in press as: M. Hussain, A direct qPCR method for residual DNA quantification in monoclonal antibody drugs produced in CHO cells, J. Pharm. Biomed. Anal. (2015), http://dx.doi.org/10.1016/j.jpba.2015.03.005

G Model

ARTICLE IN PRESS

PBA-9998; No. of Pages 4

M. Hussain / Journal of Pharmaceutical and Biomedical Analysis xxx (2015) xxx–xxx

3

Table 2 Direct resDNA qPCR method with IgG1-1 drug substance spiked with CHO DNA standards. Spike level (fg)

Expected DNA, mean (fg)

Observed DNA, mean (fg)

Observed DNA, RSD (%)

Spike recovery (%)

5E+06 5E+05 5E+04 5E+03 5E+02 5E+01 5E+00 5E−01

5,024,506.50 530,080.85 46,945.29 5185.08 450.74 49.89 5.72 0.49

2,956,119.00 321,521.22 31,517.94 3540.02 360.36 32.28 5.13 0.38

6.56 8.42 14.22 21.77 17.44 16.74 30.26 4.34

59 61 67 68 80 65 90 77

The expected DNA values were back calculated values for the standards.

Table 3 Direct resDNA qPCR method with IgG1-2 drug substance spiked with CHO DNA standards. Spike level (fg)

Expected DNA, mean (fg)

Observed DNA, mean (fg)

Observed DNA, RSD (%)

Spike recovery (%)

5E+06 5E+05 5E+04 5E+03 5E+02 5E+01 5E+00 5E−01

5,187,359.93 509,807.32 55,004.54 4511.08 467.22 40.39 6.80 0.40

5,888,229.50 543,356.56 46,455.90 5512.22 552.36 52.48 5.74 1.05

6.90 5.47 4.68 13.38 2.47 7.27 10.57 46.51

114 107 84 122 118 130 84 263

The recovery of spiked CHO standard DNA was 60% with 150 ␮g protein and higher with 50 or 100 ␮g protein (Table 1). Further evaluation of the method was conducted with 50–100 ␮g mAb protein. The experiment shown in Table 1 demonstrated that digestion of 50–100 ␮g IgG1-1 drug substance with KAPA protease followed by heat denaturation and qPCR, could be a good residual DNA method with about 80% spike recovery. In order to assess the range of the direct resDNA qPCR method and to develop an easily workable methodology, an experiment was conducted with IgG1-1 50 mg/mL drug substance (2.0 ␮L or 100 ␮g protein per reaction) in 27 wells of the 96-well PCR plate as described in Section 2.2 with serially diluted DNA standards spiked in triplicate to 24 wells and other three wells used as unspiked control. The results, shown in Table 2, clearly demonstrated that the method had a linear range from 0.5 fg to 5E6 fg for 100 ␮g of drug substance with good precision at