The basic principles of nucleic acid amplification technologies (NAT) and definitions of the various techniques are covered in Nucleic Acid-Based Techniques—General 1125. This chapter covers the analytical procedure used to quantify residual DNA in biopharmaceuticals.

Quantification of residual DNA impurities in biopharmaceuticals is based on safety concerns. The cells used to produce biopharmaceuticals can be sources of a range of complex, heterogeneous, and potentially unsafe impurities, and host cell DNA is among these. Much of the safety concern associated with residual DNA in biopharmaceuticals lies in the possibility that host cell DNA, particularly continuous-cell-line DNA, may result in tumors or adverse reactions. Cells used to produce biopharmaceuticals may possibly carry viruses or harbor harmful nucleic acid, and the residual DNA in a given biopharmaceutical product may be infectious. Although animal testing has shown that extraneous DNA can cause tumors or infections, no reports to date have demonstrated this risk in humans. Therefore, some regulatory agencies have allowed a target of 100 pg or less of residual DNA per dose in biopharmaceuticals, and levels up to 10 ng of residual DNA per dose may be considered, depending on the source of the residual DNA and the product's route of administration.

One can address residual DNA in biopharmaceutical processes in two ways: by validating clearance during process validation or by monitoring residual DNA levels by routine testing of the drug substance. The level of concern regarding residual DNA can be tied to the potential source of the residual DNA (e.g., infectious viral DNA) and the route of administration, so the residual DNA specification and procedure for monitoring DNA clearance for a given product should be developed in consultation with regulatory agencies. Regardless of whether routine testing of a drug product is used to determine residual DNA content or whether DNA clearance is demonstrated by process validation, analytical procedures for the quantification of residual DNA are required. The analytical procedures used to determine the residual DNA content of biopharmaceuticals can include hybridization, instrumentation based on DNA-binding protein, quantitative PCR (q-PCR), or other DNA amplification methods. The expectation is that the analytical procedure used to quantify residual DNA in biopharmaceuticals has a detection limit approximating 10 pg per dose. The assays based on hybridization, DNA-binding protein, and q-PCR are typically the techniques of choice because they can meet the sensitivity expectation.

Analysis of residual DNA requires accurate quantification of pg levels of DNA in mg (or larger) quantities of product. The sample itself, whether it is a protein or other chemical entity, can create sample matrix effects that must be overcome in order to yield a useful assay. Protein samples may require only digestion with proteinase (e.g., Proteinase K, Pronase) to allow the analytical method to quantitatively recover the residual DNA. Treating the sample with a detergent may be required to dissociate the residual DNA from the sample matrix. Traditionally, extraction methods based on phenol and chloroform, followed by ethanol precipitation, have been applied to the purification of DNA in molecular biology research. The phenol/chloroform extraction technique may be a useful pretreatment for residual DNA samples prior to analysis. Because of the typically low levels of residual DNA present in samples, quantitative DNA recovery with ethanol precipitation may be difficult. For this reason, a carrier molecule (e.g., glycogen) may be necessary to aid in DNA recovery if this technique is used.

A commercial kit is available1 and has been used successfully for pretreatment of residual DNA samples. The commercial kit uses a chaotrope (sodium iodide) and a detergent (sodium N-lauroyl sarcosinate) to disrupt the association of the DNA with the sample. The DNA is then co-precipitated using glycogen as the carrier molecule in the presence of isopropanol.

Each of these pretreatment techniques may yield acceptable results, or analysts may combine the techniques to obtain acceptable recovery of the residual DNA from the sample. Sample extraction is an extra handling step that may cause the incomplete recovery of the residual DNA or may introduce environmental DNA into the sample, so great care must be taken during any sample manipulations. Addition of DNA-spiked samples in the residual DNA assay is a common practice. A recovery of 80% to 120% of the spiked DNA is an acceptance criterion often applied to residual DNA assays to ensure that the assay yields acceptable results. When sample characteristics (e.g., matrix effects, sample preparation method) make achieving a recovery acceptance criterion of 80–120% impractical, then correcting the observed DNA concentration by the load recovery percentage is also an acceptable approach. During the qualification of a residual DNA assay, some scientists treat the samples with DNase I to degrade the DNA in the sample in order to demonstrate that the assay response was due to DNA and not some other sample component.

The first residual DNA assays were based on DNA hybridization, wherein a DNA probe created from host cell DNA detects and quantifies the amount of complementary DNA present in the product under assay. Double-stranded host cell DNA consists of two complementary strands of DNA that are held together by hydrogen bonding. The double-stranded DNA in the test sample is denatured to single strands and immobilized to a membrane, typically a nitrocellulose or nylon membrane. The sample is probed using host cell DNA that has been denatured and labeled. The host cell DNA probe is not a specific sequence but is prepared by a random labeling procedure during which a radioactive or fluorescent label is introduced into the host cell DNA to produce the probe. When the denatured labeled DNA probe is brought into contact with the membrane-immobilized DNA, the probe will bind to complementary sequences of the host cell DNA. If the probe is radioactive, the membrane is placed against autoradiography film for a sufficient length of time, the film is developed, and a dark spot will be observed where the test DNA was immobilized. If the probe has a fluorescent label, the intensity of the spots is determined using a phosphor- or fluorescence-imaging system. The intensity of the spot is proportional to the amount of probe that was hybridized to the test DNA and therefore is proportional to the amount of residual DNA in the sample. The intensity of the spot can be compared visually with the intensity of spots that correspond to a standard curve yielding semi-quantitative results (i.e., visual quantitation), or the intensity can be determined using an instrument (e.g., densitometer) to create a quantitative value that is compared with the values obtained from the standard curve.

Instrumentation is commercially available for the quantitation of residual DNA in biopharmaceuticals. The instrumentation requires reagents that use DNA-binding protein and antibodies targeted for DNA in a four-step analytical procedure. The first step requires that the DNA be denatured into single-stranded DNA by sample heating. The denatured DNA is mixed with a single reagent that contains DNA-binding protein that is conjugated with streptavidin and a monoclonal anti-DNA antibody that is conjugated to urease. The DNA-binding protein and the monoclonal antibody are specific for single-stranded DNA but do not have any sequence specificity. This liquid phase facilitates the formation of reaction complexes that contain DNA, streptavidin, and urease. During the second step the sample is filtered through a biotinylated membrane that binds to the streptavidin and captures the complexes on the membrane, which is washed to remove any reagents that are not bound to the membrane. During the third step the membrane is inserted into a sensor on the instrument, where the urease in the DNA complex reacts with a urea solution in the sensor, producing ammonia and a change in pH that is detected using a light-addressable potentiometric sensor (LAPS). The change in pH directly correlates with the amount of DNA in the sample. In the fourth step the raw data from the instrument are analyzed using the appropriate software to determine the residual DNA content of the sample.

Real-time q-PCR is a procedure that is well-adapted to fast sample throughput and has applications in many areas of biopharmaceutical manufacture (e.g., copy number detection, virus detection). The technique can quantify the amount of a nucleic acid target sequence in DNA from a variety of samples. The DNA probe used in the analysis is the key to the procedure. The probe has a reporter dye attached to one end and a quencher dye attached to the other end. A DNA primer is also added to the reaction. During the amplification reaction, DNA polymerase I attaches where the DNA primer binds to the single-stranded sample (template) DNA and moves along the sample DNA synthesizing new complementary DNA. While following the template DNA, DNA polymerase I cleaves any complementary DNA in the path. If DNA polymerase I encounters the labeled DNA probe it will cleave the reporter dye from the probe. The reporter dye is released into solution and, in the absence of the quencher dye, can be quantitated as a fluorescent measurement. Repeating the reaction cycle results in an amplification of the fluorescent signal. The number of cycles required for the fluorescent measurement to exceed a threshold value correlates to the amount of starting residual DNA in the sample. By comparing with a standard curve the fluorescence obtained from a sample, analysts can quantify the residual DNA in the sample.

Analysts choosing hybridization, DNA-binding protein, or q-PCR techniques for residual DNA analysis should consider how the assay will be used, the structure of the DNA available (e.g., fragment length), and regulatory issues. The cost of analysis can be significant and should be considered when evaluating an assay format. Traditionally, hybridization assays were performed using 32P-labeled DNA and autoradiography. Because 32P decays quickly, probes prepared with 32P have a limited shelf life, and the precautions necessary for handling radioactive material can be cumbersome.

These issues with 32P labeling may make fluorescence labeling of the hybridization probe a more desirable option. If the hybridization assay is assessed visually, this represents a semiquantitative assay, but if the intensity of the spots is determined using a densitometer or other image system, the results can be quantitative. DNA-binding protein assays and q-PCR give quantitative results. Quantitative assays are typically preferred instead of semiquantitative assays because the results are considered more accurate and precise, which allows better process monitoring and control.

Due to sample interference, a sample pretreatment step is often required to obtain accurate and reproducible results. Pretreatment steps can influence the recovery of DNA, so it is often necessary to design the assay with a spike-recovery control and an acceptance criterion to ensure assay performance. Commercial sources of host cell and vector DNA are typically not available to prepare in-house controls. In-house controls are usually prepared in the laboratory and quantified by UV spectroscopy, using standard techniques employed in molecular biology, to determine the DNA content and purity. Additionally, it is a good practice to evaluate in-house residual DNA controls by agarose gel electrophoresis to demonstrate that the DNA is of a proper size for the assay employed and has not degraded.

The hybridization assay uses genomic and/or vector DNA, labeled randomly throughout the DNA, as the hybridization probe reagent. For this reason the hybridization assay is specific for the source of DNA but is not specific for a given sequence. A synthesized probe, specific for a specific sequence, can be prepared and used in the hybridization assay if this level of specificity is desirable. The DNA-binding protein residual DNA assay is not sequence-specific and hence not specific for the host DNA. Therefore, laboratory personnel should avoid contaminating samples for this assay with environmental DNA before denaturing the DNA; otherwise the DNA result may be falsely elevated. The q-PCR probe is sequence-specific, which creates some special challenges for development of a q-PCR residual DNA assay. The q-PCR-specific sequence must be a stable sequence within a highly conserved region of DNA. The recovery of the probe target sequence must consistently represent the recovery of all the residual DNA. As a guideline, for a DNA fragment to be detected by hybridization, q-PCR, and DNA-binding protein assays, it must have no fewer than 50, 150, and 600 base pairs, respectively. A bioprocess typically may have operations that shear DNA into smaller fragments, and this must be taken into consideration when selecting an assay. Procedures exist to determine whether the DNA fragments in a sample are too small for adequate residual DNA recovery with a given assay. As noted, residual DNA assays are extremely sensitive. Detection limits as low as <1, 3, and 6 pg of DNA per sample have been reported for q-PCR, DNA-binding protein, and hybridization assays, respectively.

Although safety concerns regarding residual DNA impurities are not as prominent as they once were, the levels of residual DNA in any bioprocess remain a key quality attribute and help define the process.