Polymerase Chain Reaction: Applications in Gene and Cell Therapy Studies

Abstract

The rapidly developing fields of gene and cell therapy allow us a platform to repair or replace defective genes or introduce a missing gene. AAV and lentivirus are common viral vectors used in gene therapy to deliver a DNA payload to a tissue of interest. Recently, self-replicating RNA-based vaccines and therapies are also becoming increasingly popular for gene therapy after the success of SARS-CoV-2 vaccines. Cell therapy is the transplantation of human cells without or with ex vivo modification utilizing CAR-T and stem cell technology. Because PCR allows us to detect transgenes with high sensitivity, we can leverage this technology to quantify the efficacy of a therapy and long-term expression in vivo using both qPCR and RT-qPCR, respectively. PCR provides information that is used to justify first in human dose, toxicological evaluations, efficacy through PK/PD relationships, monitor persistency and shedding as well as biomarker and gene expression quantitation. As evaluation of safety endpoints is critical to drug development, PCR is imperative in the field of clinical pharmacology discovery.

1. Introduction to gene therapy

Countless human diseases are caused by a change or mutation to a single gene. Examples include the CFTR gene in Cystic Fibrosis and the Factor IX protein in hemophilia making these diseases strong candidates for gene therapy. In an effort to treat or cure a disease, gene therapy attempts to introduce a functional protein to the diseased tissue or cells using a viral vector. The most common vector used in gene therapies is AAV (adeno-associated virus), which is a naturally occurring non-pathogenic parvovirus with a single-stranded, linear DNA genome about 4.7 kilobases in size. Once the viral DNA of the wild-type AAV is replaced with the therapeutic DNA also known as the transgene or corrected gene of interest, it is now considered recombinant AAV (rAAV). The rep and cap genes of the AAV serotype are now provided in trans during the production of vector. The packaging limit for AAV is approximately 5 kb although innovations have been made to overcome this. rAAV-mediated gene delivery has proven to have highly efficient transduction as well as long-term gene expression which is what makes this vector a great delivery system for gene therapies.

AAV gene therapy routes of administration can vary by indication or study type. The capsids can be injected directly into the tissue of interest such as intravitreal for eye indications or intramuscularly or intravenously. The typical mechanism of action of an AAV gene therapy is receptor mediated. The capsid will bind to the cell membrane and become endocytosed forming an endosome containing the viral particle. The capsid will escape from the late endosome and translocate to the nucleus where the protein capsid will be uncoated, and the single-stranded genome will be released. The formation of double-stranded DNA will occur in the nucleus, and the rep genes will begin to be expressed allowing for the packaged transgene to replicate inside the cell. AAVs rely on cellular polymerases to replicate its genome and packaged DNA inside the nucleus of transduced cells. This is where qPCR will quantitate the amount of transgene that can be detected inside the cell.

To measure the efficacy of a gene therapy, a biodistribution study is done. This consists of designing a small or large animal study administering the gene therapy and harvesting a panel of tissues to analyze biodistribution of the transgene. Small dissections of tissues from the dosed animals are harvested and processed for qPCR analysis. qPCR (quantitative polymerase chain reaction) is used to quantify the amount of transgene DNA sequences that has been transduced into the tissue of interest as well as peripheral non-targeted tissues. Bioanalysis results are reported as copies of the target DNA per μg of host genomic DNA. qPCR probes are designed to bind and amplify the transgene sequence and a copy number is calculated from the cycle threshold by applying a standardization curve of known copy numbers.

2. Introduction to cell therapy

Cellular therapy is the use of live cells as the drug product in the treatment of disease. The cells or “living drugs” are transplanted either with or without post ex vivo genetic modification and include Chimeric Antigen Receptor (CAR) T-cell therapy, stem cells, Engineered T Cell Receptor (TCR) Therapy, tumor-infiltrating lymphocytes (TILs), and NK cells with the potential for more therapies in the future. T cells are often selected because of the important role they play in cell-mediated immunity and offer a targeted therapy as an alternative to chemotherapy and/or surgery [1]. To highlight how important cell therapy scientific advancements of being able to harness the body’s natural immunity to fight disease have been, we note that they resulted in the Nobel prize in Medicine being awarded to James P. Allison and Tasuku Honjo in 2018 [2]. Their research, along with many others, on how the immune system can be used in defense of cancer, has opened to the doors for clinical trials that have resulted in treatments for individuals who had no other treatment options. The FDA has approved the use of multiple cell therapies including Kymriah®, Yescarta®, Tecartus®, Breyanzi®, Abecma®, Carvykti®, Zyneglo®, and Skysona®.

The CAR-T field specifically is rapidly expanding with a shift in focus from leukemia and lymphoma to solid tumors that are often difficult to target as well as from autologous treatment, where cells are taken from individual patients, to allogenic where donor cells will be used. The latter would allow for a more “off-the-shelf” treatment option, which would allow for earlier intervention of the disease (Parker Institute 2019). Advancement in the cell therapy also includes developing new modalities that are emerging alongside progress in current technology.

Therefore, monitoring the presence of cell therapies and their products is of high importance when utilizing them as a treatment option. Therefore, it is essential to have tools to monitor for the presence of the cellular drug product post-dose, for biodistribution, for potential shedding, and for transgene expression if applicable. These are used to help correlate dose versus efficacy, to evaluate potential toxicity, which would affect safety, for evaluation of off-target distribution and clearance from the body. They are also used to assess potential risks to the patient or community through shedding studies to evaluate the presence of the viral particles. However, as cell therapy treatment potential increases, adverse effects also need to be considered as cytokine release syndrome (CRS) as well as immune effector cell-associated neurotoxicity syndrome (ICANS) can potentially occur [3]. Due to this, it is important to note that immunogenicity is also an important assessment that should be paired with cellular kinetics for full evaluation.

Flow cytometry has historically been used to measure the cellular kinetics of cell therapies and is now often paired or sometimes replaced by quantitative PCR measurement. These two widely used methodologies offer detection strategies that are critical for evaluating the appropriate dose to patients through detecting the cells in relation to the efficacy and potential adverse reactions. A strong correlation has been established between these two assay types when comparing vector copy number by qPCR and CAR expression by flow cytometry. qPCR as a methodology provides measurements used to detect vector or transgene DNA or mRNA by establishing a standard curve for semi-quantitative evaluation where a linear regression is applied to the Ct values and the log concentration of each dilution. Digital PCR applies absolute quantitation and may also be utilized where applicable. For CAR-T, a critical window is 28 days post-dose to measure peak expansion and persistence but needs to be measured in long term as well. While flow cytometry is a reliable method to detect circulating CAR-T cells, it utilizes fresh cells, which can prove difficult if not processed on-site or quickly after. PCR provides readouts at very low levels due to high sensitivity that allow for measurement after the dose and expansion phase to monitor persistence, which expands the time points that we can screen for CAR-T cells [3].

3. Regulated work

Bioanalysis utilizing qPCR, ddPCR, dPCR falls under the supervision of regulatory agencies including the FDA and EMA. As countries approve medical treatments, they release regulatory guidance that is heavily relied on in drug development and clinical testing to determine how testing should occur. Pre-clinical and clinical studies pre-determine the collection and analysis methods that will be used in the evaluation of the efficacy and safety of the drug.

The FDA and the EMA have released both final and draft guidance on cell and gene therapy itself but there are few regulated to validate bioanalysis specifically at this time. There are also ICH guidelines with a focus on nonclinical biodistribution, virus and vector shedding, and oncolytic viruses. However, the validation of the PCR-specific assays still has many areas that are not in focus. The MIQE guidelines combined with published white papers are currently utilized to provide alignment and harmonization across the industry as white papers published by experts in the field also provide direction when conducting regulated work. As draft and final guidance are released for the validation of these PCR assays, harmonization can be achieved.

4. Modalities

Gene therapy is classified into two types based on the cells treated or transgene delivered to. They are somatic and germline gene therapy. Due to ethical reasons and safety/efficacy concerns, viral vectors are considered unacceptable for germline gene therapy (GGT). For GGT only Mitochondrial Replacement Therapy to replace mutated mitochondrial DNA from oocytes with normal mtDNA is permitted in UK [4]. Gene editing tools such as CRISP6-Cas9 are considered as alternatives for GGT for non-mitochondrial GGT [5].

Somatic gene therapy on the other hand allows the use of a wide variety of modalities for gene delivery. This includes both biological and non-biological methods. Among the biological methods viral, non-viral vectors and genome editing tools such as CRISPR-Cas9 are in use. Currently, the viral vectors are the most successful and widely used [6]. The most popular viral vectors in use for gene therapy are adeno-associated virus, adenovirus, lentivirus, retrovirus, oncolytic virus (Herpes Simplex Virus-1). A brief description on the application of these viral and non-viral vectors in gene and cell therapy programs will be discussed.

4.1 Viral vectors

Typically, a viral vector is defined by three components as follows: (1) the transgene; (2) the envelope or capsid that harbors the payload, which also defines antigen recognition and tropism, and (3) the “regulatory cassette” that controls transgene expression with or without integration.

4.2 Adenovirus vectors (Ad vectors)

These are the first viral vectors used in gene therapy [7]. Ad is a non-enveloped virus, with icosahedral protein capsid harboring a 26- to 45-kb linear, double-stranded DNA genome comprising five each of early phase (E1A, E1B, E2, E3, and E4) and late phase genes (L1–L5). These genes are flanked by hairpin-like inverted terminal repeats (ITRs) ranging from 30 to 371 bp in length. E1A is essential for the transcription of other viral genes (E1B, E2-E4), and late phase genes are responsible for viral DNA synthesis and other roles. The Ad viral capsid is composed of hexon, penton, fiber and capsid protein precursors, and viral core proteins. There are about 100 genotypes of Ad viruses. First-generation Ad vectors were generated by replacing the E1A/E1B region with transgene cassettes upto ~4.5 kb. Second-generation Ad vectors were developed by deleting E2A, E2B, or E4 genes that allowed transgene cassettes upto 10.5 kb size. First- and second-generation Ad vectors require engineered cell lines for the deleted genes. Third-generation Ad vectors were developed by deleting all viral sequences except ITRs and the packaging signal (psi). These are called, “helper dependent” and “high capacity” (HCAds) as they can carry up to ~36 kb of the cassette. HCAds also have low immunogenicity [8, 9]. Recently, a PCR-based method termed as Restriction Assembly (RA) has been developed. This method uses long restriction-digested fragments and short PCR products for DNA assembly. This was used successfully to construct E1 or E1/E3 deleted vectors based on simian adenovirus 1 (SAdV-1) [10]. In addition, PCR-based directed evolution and capsid engineering are found to be useful to produce adenovirus vectors of the desired properties. [11]. For example, novel chimeric rAd5 vectors were constructed replacing the seven short hypervariable regions (HVRs) on the surface of the Ad5 hexon protein with the corresponding HVRs from the rare adenovirus serotype Ad48. These vectors lack pre-existing anti-vector immunity against the key hexon surface protein epitopes [12]. Conditionally replicative Ad viruses were constructed (CRAd or rAd) by cloning under tumor-specific promoters to make oncolytic rAd viruses [13].

4.3 AAV vectors

Adeno-associated viruses are currently the leading platform for gene/cell therapy programs. Its genome is ~4.7 kb long single-stranded DNA. For essential genes needed for replication and gene expression, it depends on adenovirus or another virus, which can provide the same support. These functions are provided by the Ad E1, E2a, E4, and VA RNA genes. AAV genome has four ORFs, a) rep encodes the four replication genes, b) cap encodes three viral capsid proteins, c) the third and fourth are nested sub-genomic mRNAs—Assembly Activating Protein (AAP) involved in the shuttling of capsid monomers to the nucleolus for capsid assembly and Membrane Associated Accessory Protein (MAAP) the function of which is not yet clear [14]. The 4.7-kb genome is flanked by 145 base ITRs on both ends of the genome. The ITRs serve as self-priming structures for replication and provide the signal for Rep-mediated packaging.

Several serotypes of AAVs with varied tropism are identified and used in preclinical and clinical studies. The tropism depends on the capsid type [15]. So capsid modification enables the generation of recombinant vectors with tailored properties and tropism. Initial attempts to engineer the capsid and redirecting to the desired tissue were performed by PCR amplification of the capsid sequences of choice and grafting them to the selected AAV serotype genome. DNA shuffling, site-directed mutagenesis, and error-prone PCR are the PCR-based methods utilized in the rational and directed evolution of AAV capsids [16, 17]. For example, mutants were isolated by error-prone PCR and staggered extension protocol of diverse libraries of AAV5 capsids that have enhanced transduction in human hepatocytes and favorable seroreactivities [18]. A robust 2-step PCR method for DNA shuffling of selected serotypes of AAVs to generate capsids of desired immunogenicity levels and tropisms has been developed [19]. For example, capsid DNA from AAV serotypes 1–6, 8, and 9 shuffled and recombined to create a library of chimeric AAVs. From these, two mutants selectively crossed the seizure-compromised BBB and transduced cells in mice [20].

4.4 Lentiviral vectors

Retroviruses are spherical, enveloped with two sense-stranded RNA of ~9.2 kb size. The RNA strands are bound by nucleocapsid proteins. Lentiviruses belong to the HIV-1 type of retroviruses [8]. It has three core genes, namely gag responsible for encoding the structural proteins, pol for reverse transcriptase, protease, and integrase, and env for virus envelope glycoprotein [21]. The genome has regulatory genes (tat for activating viral transcription and rev—for splicing and export of transcripts) as well as four accessory genes (vif, vpr, vpu, and nef). Accessory proteins are not essential for replication. Long terminal repeats are present on two ends. Both the 5′ and 3′ LTRs in wild-type HIV-1 is composed of U3, R, and U5 sequences. The “psi” just after LTR is important for signaling and genome encapsidation.

Lentiviral vectors integrate into host genome with a packaging capacity of ~9 kb. Lentiviral vector systems are improved over the year for safety and lower immunogenicity. They were more successful in ex vivo applications generating CART cells for cancer immunotherapy. The third generation of lentiviral vector carrying the gene of interest contains a promoter (CMV/RSV), part of the 3′U3 LTR. The psi (ψ) packaging signal is followed by the rev response element (RRE) [18]. A novel lentiviral vector (LTR1) described as the fourth-generation vector was designed with a unique genomic structure to prevent transfer of HIV-1 packaging sequences to patient cells. This reduces the total HIV-1 content to ~4.8% of the wild-type genome (Figure 1) [22]. Helper plasmid constructs provide other necessary genes for capsid formation, which do not replicate.

Figure 1.

Adenovirus vector generations. ITR = inverted terminal repeats, ψ = packaging signal, major late promoter (MLP), E1A, E1B, E2A, E2B, E3 and E4 – Early transcription units, L1-L5 – Late transcription units.

4.5 Chimeric antigen receptor T-(thymus) lymphocytes (CAR-T)

CARs are engineered synthetic receptors that function to redirect T cells, to recognize and eliminate cells expressing a specific target antigen. Chimeric antigen receptor (CAR) T-cell therapy is a major step in personalized cancer treatment. In CAR-T cell therapy, a patient’s own T cells (allogenic) are engineered to express a synthetic receptor that binds a tumor antigen and injected back into the patient to attack and kill the cancer cells. Viral vectors are used to encode CARs; the retroviruses integrate their DNA in the genome of the obtained T cells for long-term expression. These viral vectors include retroviruses, lentivirus, adenovirus, and adeno-associated virus [23]. Over the time the CAR-T cell therapy has evolved from first generation to fifth generation. In the first generation, only immunoreceptor tyrosine-based activation motif (ITAM)—(CD3ζ) motifs as the T cell stimulatory molecule within the intracellular domain was present. The second- and third-generation CAR-T cells had one and two costimulatory molecules, respectively. The fourth generation was designed with one costimulatory molecule paired with cytokine expressors (e.g., IL-12) under the control of NFAT transcription factor, hence referred to as T cell redirected for universal cytokine mediated killing (TRUCKs). The fifth generation was also like fourth generation plus intracellular domains of cytokine receptors (e.g., IL-2Rβ) to activate JAK and STAT3/5, stimulate cell proliferation, and enhance its persistence [24, 25]. The most recently FDA approved CAR T-cell therapy, in February 2022, is ciltacabtagene autoleucel (CARVYKTI™) from Janssen Therapeutics Inc. This drug was approved for the treatment of r/r multiple myeloma. It is a genetically modified autologous CAR T-cell therapy directed by B-cell maturation antigen (BCMA).

4.6 Stem cells

Any treatment that involves the use of viable human stem cells including embryonic stem cells (ESCs), iPSCs (induced pluripotent stem cells), and adult stem cells for autologous and allogeneic therapies is considered as stem cell therapy [26, 27]. Application of stem cells for therapeutic purposes could be achieved after the production of iPSCs via the transient over-expression of four transcription factors: OCT4, SOX2, KLF4, and MYC in fully differentiated somatic fibroblast cells [28]. Stem cells from different tissues are produced and applied for the treatment of diseases or impairments related to those organs with different rates of success. Integrating vectors effectively achieve long-term expression of a transgene in the treatment of monogenetic hematopoietic disorders. Accurate vector copy number (VCN) estimation has become increasingly important in stem cell therapies also. Application of both qPCR and dPCR methods was used in the vector copy number analysis of transduced stem cells [29].

4.7 Self-amplifying RNA

Self-amplifying RNA (saRNA) replicon is a type of RNA genome that is unique to certain families of viruses. This is due to the presence of RNA-dependent RNA Polymerase (RdRP) due to which the genomic RNA can make multiple copies of it finally resulting in the production of proteins, from just a single RNA template [30]. Some of the viruses used in the production of saRNA-dependent vaccines are Venezuelan Equine Encephalitis (VEE), Semliki Forest Virus (SFV), and Recombinant Sindbis/Venezuelan Equine Encephalitis Virus (VEE/SINV). A typical saRNA-based transgene construct contains from 5’end a) 5’CSE (Conserved Sequence Element) with 7-methylguanine, b) RdRP complex cassette from an saRNA virus (Alpha virus), c) a transgene/immunogen of interest, d) 3’CSE with polyA tail.

It is also possible to have more than one immunogen in one construct separated by self-splicing peptide sequences such as 2A or internal ribosome entry (IRE) site. However, in the clinical studies involving saRNAs, only RT-qPCR methods need to be developed for PK/biodistribution studies. saRNA constructs are usually delivered via lipid nanoparticles. This requires the formulation of the payloads after construction. It is necessary to monitor the toxicity due to lipid also in these studies (Figure 2).

Figure 2.

Basic structure of a typical saRNA construct. nsp1-nsp4 form RNA-dependent RNA. Polymerase from alpha virus, required for self-replication of RNA. Transgenes (immunogens) are the genes of interest separated by either 2A or IRES sequences. 2A self-cleaves after translation at C-terminal separating the two transgenes. Internal ribosome entry site (IRES) permits translation of second transgene by a different mechanism. 2A/IRES are not required if there is only one transgene. Finally, there will be a 7-methyl guanosine at 5′end and poly a tail at 3′ end.

4.8 RISC (RNA-induced silencing complex)

In diverse group of organisms, dsRNAs were shown to interfere with gene expression in a sequence-specific fashion in cultured cells [31]. This was first identified in plants, as “post-transcriptional gene silencing” (PTGS). RNA interference (RNAi) is a nucleic acid-based immune defense against viruses, transgenes and transposons. dsRNAs present in a cell are converted either through processing or replication, into small specificity determinants of discrete size in a manner analogous to antigen processing. Small interfering RNA (siRNA) and microRNA (miRNA) are small RNAs of 18–25 nucleotides (nt) in length that play important roles in regulating gene expression. They are incorporated into an RNA-induced silencing complex (RISC) and serve as guides for silencing their corresponding target mRNAs based on complementary base-pairing [32]. The pathway of a typical eukaryotic gene silencing goes as follows:

1. Long dsRNAs are cleaved by Dicer, an RNAse III like enzyme, into small RNA duplexes of 21 to 23 nucleotides, called siRNA.

2. The Dicer-siRNA complex interacts with a cellular TAR (Transactive Response) RNA-binding protein (TRBP).

3. Another core protein of RISC complex is Argonaute 2 (Ago2). The Dicer-SiRNA complex binds with the Ago2-associated RISC complex. Some other components need to be identified.

4. After base pairing of the single-stranded guide RNA with the target mRNA, the PIWI (P-element Induced Wimpy) domain of the Ago2 protein is brought into proximity and cleaves the target mRNA resulting in a post-transcriptional gene silencing.

One major difference between siRNAs and miRNAs is that siRNAs are specific to a single mRNA target, whereas the miRNAs bind to multiple targets [33]. Therapeutic approaches based on siRNA involve the delivery of a synthetic siRNA into the target tissue/cells to elicit RNA interference (RNAi), thereby inhibiting the expression of a specific messenger RNA (mRNA) to produce a gene silencing effect. In contrast, miRNA-based therapeutics comprise two approaches: miRNA inhibition and miRNA replacement. Due to very short length, siRNA quantification by regular RT-qPCR is not feasible. For this reason, alternative PCR-based methods such as primer extension (PE), the invader assay, stem-loop RT-qPCR, ligation assay, and crook hairpins are developed and used. A modified form of competitive PCR [34] showed improved results with synthetic siRNA molecules.

5. PCR sample prep

To process tissue matrices for PCR amplification, the DNA and RNA need to be extracted and purified from the cells. Template quality is one of the biggest issues facing industry for bioanalysis. Any contaminant in the sample can cause inhibition and lead to biological non-reproducibility in the assay. There are many variations of nucleic acid extraction protocols that are used within the field. Many vendors provide kits for genomic DNA extraction or total RNA extraction, which are well characterized and used widely for optimal purity. Cells or tissues require pre-processing. The tissues need to be pulverized by mechanical processes such as a metal bead or grinding in liquid nitrogen to homogenize the sample. Proteinase K is an enzyme that can be used to digest peptide bonds to break down proteins that may contaminate your sample. RNase A can be used in DNA purifications as well to rid the sample of any RNA contamination. Once the pre-processing steps are completed, a detergent can be used to lyse the cells. Commercially available kits will call this a lysis buffer. The samples will quickly become viscous once the nucleic acid has broken out of the cells. The DNA will need to be pulled out of this solution in a binding step. It is most common to use a positively charged membrane or bead to bind to the negatively charged DNA molecules. Silica membranes with high concentrations of salt are also very common methods of binding genomic DNA so washing steps can be carried out without the risk of losing DNA. While the DNA is bound, ethanol is used to wash away all of the impurities from the cell lysate while still maintaining conditions to keep the binding strong. Contaminants and enzyme inhibitors should be washed away during these steps to ensure the sample is ready to be used in PCR. Water or TE can be added to elute the DNA off of the bound matrix into a purified sample ready for PCR.

Purifying RNA has the same overall procedure with a few steps catered toward the binding of RNA structure. Phenol-chloroform extractions were considered the gold standard extraction method for many years and are still used today; however, commercially available kits make the process much more high throughput and less risk of contamination. A DNase step would be done after purification of RNA to remove any DNA contamination. This ensures that in a RT-qPCR, all of the DNA that is amplified is from the cDNA synthesis step and not from any genomic DNA contamination from an impure RNA extraction. This is the reason to include a control sample in all RT-qPCR reactions, which excludes the reverse transcriptase in the reaction. If the sample shows any amplification, then it is amplifying genomic DNA contamination. Without the addition of the reverse transcriptase enzyme, cDNA synthesis cannot occur.

6. Standard curve

Quantitative PCR is a relative method, which means to assign absolute quantification to sample, an accurate and reproducible standard curve is required. It is acceptable to use a surrogate for the DNA being quantified in the experiment such as a DNA plasmid containing the sequence of interest. For gene therapy experiments, it is common to use the cis-plasmid used in AAV transfection, which contains the transgene sequence along with the entire sequence to be packaged into the AAV capsid for testing. Linearization of the circular plasmid by restriction digest is required so that the standard DNA most closely mimics that of the sample DNA.

Calculating the DNA concentration of the standard is essential for assay accuracy. Once the DNA concentration is measured using a spectrophotometer-based method such as a GloMax®, a copy number must be calculated in order to generate the standard curve. The equation below is used to convert nanograms per microliter (ng/uL), which is provided by the concentration instrumentation, into copies per microliter using the size of the DNA plasmid in base pairs, the molecular weight of a base pair of DNA (650 Da), and Avogadro’s number.

$$DNA\left(\frac{copies}{\mu L}\right)=\frac{DNA\left(\frac{ng}{\mu L}\right)\times {10}^{-9}\frac{g}{ng}\times 6.022\times {10}^{23}\frac{molecules}{mole}}{N\left(bp\right)\times 650\frac{g}{moleofbp}}$$

This equation can also be used for calculating copies/uL of RNA; however, the average mass of 1 base pair of single stranded RNA (330 g/mol) must be substituted into the equation. The standard curve is a serial dilution that must cover all expected copy numbers of the unknown samples. If a sample is amplified at a Ct lower than the highest point on the standard curve, the sample will have to be diluted to fall within the curve to have a copy number assigned or a new standard curve generated for samples with higher copy numbers. The standard curve should be run in triplicate in the qPCR so that.

7. Probe/primer design for gene therapy studies/inhibition

7.1 AAV titering

PCR is a tool that can be used to both titer viruses and quantitate DNA delivered to cells by AAV vector. AAV titer is one of the limitations in the field of gene therapy. The maximum concentration that can be formulated from a transfection method is about 5E13 genome copies per mL. Human studies require high doses of gene therapies to see efficacious results so routes of administration as well as capsid engineering for more effective delivery methods is a means of overcoming this constraint. Once AAV virus particles are harvested from a transfection in HEK293 cell culture, qPCR will quantitate the amount of AAV in the purified formulation. There is typically a thorough purification process done before the sample is ready for titering, which can include ion-exchange columns, ultracentrifugation, filtering, DNase treatments to get rid of any unpackaged plasmid DNA as well as a proteinase K treatment to open the capsid and release the AAV DNA genome as a template for qPCR. A probe can be designed in any of the AAV genome that persists after packaging. This is commonly the polyA or the ITRs as this is highly conserved between serotypes so a single probe set can be used for titering many viruses.

7.2 Biodistribution

The Food and Drug Administration (FDA) recommends designing qPCR probes to amplify transgene DNA rather than other AAV genome sequences. This is to quantitate the therapeutic DNA directly rather than quantifying AAV genes such as ITRs, rep, or cap gene, which may not directly correlate with the transgene copy number due to impure capsid population. During the transfection process where AAV capsids are assembled and transgene DNA is packaged, it is not uncommon for the DNA payload to become either truncated or missing from some capsids altogether. This is referred to as the empty to full capsid ratio during titering. Empty capsids still contain all of the AAV genes but none of the transgene DNA so designing a probe to anneal to the therapeutic transgene DNA is where the copy number measurement will be.

It is important when designing primers and probes for qPCR transgene amplification that species homology is taken into consideration. In most cases, the transgene of interest is a human sequence that is being used to translate a protein inside the cells for therapeutic effects. In small and large animal studies using rats, mice, dogs, or non-human primates, it must be recognized that these animals carry an endogenous sequence of these genes as well, which may be similar or dissimilar to the human sequence. If there are portions of a gene sequence that is heterologous between species, this is a good candidate for designing a probe against. A thorough probe test must be done to show the probes amplify human sequence and not the endogenous animal sequence to ensure no background copies are counted in the study qPCR readout.

7.3 qPCR inhibition

The occurrence of PCR inhibition can result in inaccurate results, which is of high concern in bioanalysis, biomarker research, and gene expression studies. Known PCR inhibitors include hemoglobin, IgG, humic substances, urea, bile salts as well as laboratory reagents including buffer or collection articles such as swabs. Inhibition of the DNA polymerase, annealing, and fluorescent detection can result in the decreased PCR efficiency. This decreased efficiency will result in the under reporting of the sample copy number in the samples and possibly the standard curve and quality controls as well. When utilizing qPCR for sample analysis, the matrix itself can result in reduction of the fluorescent signal resulting in inaccurate readouts so extraction procedures should be carefully developed to remove inhibitory factors (Sidstedt 2020; Buckwalter 2014).

Methods can be put in place to avoid inhibition and inhibition evaluation can help determine whether it has occurred. In qPCR, when a standard curve and QCs are present, the best practice is to mimic your sample composition where possible with the addition of DNA or RNA and buffer that will be used in sample extraction elution. The use of these will assure that the efficiency of the qPCR was developed with the consideration of the sample type in mind as well as potential inhibitory agents. Also, inhibition controls are often used to evaluate the present of inhibition in individual samples. Spiking a known amount of sequence specific product can be used to standardize the known amount per reaction. When a decreased CT value is shown to be detected, the sample can be reprocessed and reanalyzed. Contamination is huge concern when conducting PCR. While best practices should be employed throughout the process, there should still evaluation steps put in place to evaluate if it has occurred.

8. Data analysis: from Cq (Ct) to copy number

Real-time PCR involves continuous monitoring of the amplification process as it occurs [35]. The amplification at each cycle is monitored fluorometrically. The cycle at which the fluorescence emitted is greater than the background (in exponential phase) is taken to measure template quantity. Hence, it is called cycle quantification (Cq) or cycle threshold (Ct). Under specified conditions, the template copy number is inversely proportional to the Cq/Ct, which is plotted as log of copy number. The real-time PCR platforms are used for both detection and quantification (qPCR). RT-qPCR was originally developed from real-time PCR to amplify distinct nucleic acid sequences for the detection and relative quantification of mRNA levels in human monocyte-derived macrophages [36]. Development of single-Step RT-qPCR and efficient reverse transcriptases have made the RT-qPCR process easier and simpler [37]. In two-step RT-qPCR, the reverse transcription is separated from the real-time PCR assay. This has an advantage when several targets are to be quantified as the efficiency of reverse transcription for each template varies [38]. Though both dye (SYBR) and hydrolysis probes are in use for real-time PCR, due to specificity, multiplicity, and sensitivity, Taqman probes are preferred for quantification.

Though the digital PCR is gaining popularity and applied in several studies, still qPCR/RT-qPCR is considered as the “Gold Standard” technique due to its wide range of quantification, availability of platforms, and less expenditure. Except the conversion of RNA to cDNA by using a reverse transcriptase (RT step), the technology of qPCR and RT-qPCR remained the same. The output data from current real-time PCR platforms are Ct values. In both the gene therapy and cell therapy studies, the reporting units for absolute quantification are as follows,

1. vector genome copy number/μg of host genomic DNA. This is calculated either directly from the sample total DNA concentration or from the copy number of a reference gene [39].

2. Transcript (mRNA) copy number/μg of total RNA in the sample.

In the case of relative quantification, the units are expressed as fraction compared to the level of a selected reference gene. Relative quantification is preferred in pharmacodynamic (PD) studies for expression profiles of several biomarker gene transcripts.

Thus, to report the results in the case of absolute quantification sample DNA/RNA concentration measurement is essential. This is also essential to normalize the nucleic acid (NA) concentration in all samples to rule out the effect of this variation on Ct value.

8.1 Quantification and quality testing of total RNA in purified extracts

For total DNA/RNA quantification, the instruments, Nanodrop/UV-VIS spectrophotometer [25], Qubit, Agilent TapeStation and Promega Glomax, etc., are popularly used in different labs globally. However, specificity of Nanodrop is very less compared to the other three instruments that are fluorescent-based (Qubit & Glomax or electrophoresis based—TapeStation). In addition, Agilent TapeStation and Promega Glomax offer high-throughput technologies more suitable for the analysis of large number of samples. Nanodrop is the easiest way to test the quality of either DNA or RNA by measuring the A260/A280 nm absorbance [40]. As DNA is relatively stable, its quality is not tested once the extraction method in the matrices has been validated. However, the RNA being labile and prone to degradation easily, it is recommended to confirm the sample RNA quality before target quantitation by RT-qPCR. As the % of m-RNA in eukaryotic total RNA extracts in general will be ~5%, the quality of RNA is tested based on the integrity of 28S and 18S rRNA quantities. This is performed by one of several ways such as running on an agarose gel or TapeStation or a fragment analyzer [41].

8.2 qPCR/RT-qPCR data analysis for absolute quantification

A typical PCR-based experimental data analysis for BD of a gene therapy/cell therapy has the steps.

1. Conversion of Ct per reaction to copies per reaction—using a standard curve.

2. Copies per reaction to copies per a reference value.

3. In the case of RNA, calculation of copy number/% of DNA contamination from RT minus reaction data.

The output data from any real-time PCR instrument is Ct/Cq number. Serially diluted standards that produce a linear relationship between Ct values and initial amounts (copy number) of target DNA or RNA are used. From this, the copy number of the same target sequence in study samples based on their Ct values is drawn. In this case, the amplification efficiencies are considered same. In the case of single step RT-qPCR reactions, the standards and any Quality Controls (QC) must be RNA only [42]. GLP/regulated studies require a validated instrument and calculation sheet or a LIMS platform that can perform these calculations. Secondly, the copies/reaction are to be converted to copies per μg of DNA/RNA in the sample or mL of biofluid in clinical studies or mg of fecal material.

8.3 RT-qPCR data analysis for relative quantification of target transcript abundance

Relative quantification in gene and cell therapy involves measurement of changes in the level of sample gene expression based on an external standard or a reference sample in which the results are expressed as a target/reference ratio. A reference gene is completely different from the target gene. So, in most cases the amplification efficiency of the reference and target gene mRNA will be different. The simplest equation to calculate the amplification efficiency from the RT-qPCR data is e = 10^–1/slope [43], where: e = theoretical efficiency, Slope = the slope of the standard curve, plotted with the y axis as Ct, and the x axis as log(quantity). Several mathematical models are developed to calculate the mean normalized gene expression from relative quantification assays. Table 1 below lists some of the available models and their principles [44].

Model name	Details
Standard curve method	Copy numbers quantitated using a standard curve and relative quantitation calculated in comparison with a selected calibrator sample
The comparative Ct method	A mathematical model that calculates changes to relative fold difference between an experimental and calibrator sample, no standard curve required
Pfaffl model	Calculates relative gene expression data while accounting for differences in primer efficiencies using the equation: Gene Expression Ratio = (EGOI) ^ΔCt GOI/(EHKG) ^ΔCt HKG Where E = Efficiency, GOI = Gene of Interest, HKG = HouseKeeping Gene or Reference gene.
Q-gene method, Gentle et al. model, Liu and Saint method, Amplification plot method	Calculate amplification efficiencies and normalize the data.
Data Analysis for Real-time PCR (DART-PCR)	Converts raw data into R 0 values, based upon the theory that fluorescence is proportional to DNA concentration

Table 1.

Mathematical models for relative quantification of RT-qPCR data.

FDA has not yet required validation of qPCR/qRT-PCR assays [45]. Extensive discussion and recommendations for “best practice” for qPCR/qRT-PCR assay design during method development, and validation for gene therapy sample analysis with special focus on AAV-based clinical and non-clinical studies were made recently [46, 47]. These practices are followed for biodistribution analysis in gene/cell therapy studies for vector transgene or its mRNA quantitation by qPCR or RT-qPCR, respectively. Parameters and criteria for a typical AAV gene therapy are shown in Figure 3.

Figure 3.

PCR assay validation criteria for gene/cell therapy studies.

9. Digital PCR

Application of digital PCR (dPCR) is vastly increasing in pharma industry and basic research because of technological progress and increasingly affordable technology. The term “digital PCR” was coined by Vogelstein and Kinzler in 1999 [48]. dPCR involves dispersal of the reaction mixture into thousands of small reaction units and performing PCR with either real-time or end-point data collection in many separate reaction chambers. The partitions are performed either by droplet formation or by microfluidic nanoplate partitions. The essential steps in the data analysis of digital PCR, a) Number of partitions measured; b) Partition volume; c) Copies per partition; and d) dPCR analysis program details. The raw data of the digital PCR provide the number of positive/negative partitions (droplets in case of ddPCR) for each sample. Mean number of target sequences is calculated using the manufacturer’s software applying Poisson distribution co-efficient. This also compensates for if there is more than one copy of template in some partitions. MIQE guidelines [49, 50] listed the information to be reported for the publication of dPCR data. A comparison of the digital and real-time PCR methods has been made (Table 2).

Item	Real time	Digital
Principle	Quantification is performed in real time as the amplification happens after every cycle. Data are collected during exponential (log) phase—to calculate the template copy number.	Partitioning of the PCR reaction into thousands of individual reaction vessels prior to amplification. Microfluidics technology is used for digitization. Mostly, data acquisition is end point. Commercial instruments for real-time digital PCR are not yet very popular.
Quantification	Both absolute and relative quantifications are possible with a set of standards and reference gene, respectively.	Absolute and relative quantifications are possible without a standard curve.
Quantification	Cq/Ct Value is the cycle number of PCR at which the fluorescent intensity is above the set threshold value. A standard curve is constructed with a set of standards with known copy number that gives specific Ct values.	In endpoint dPCR, the copy number is calculated based on the fraction of positive and negative droplets or partitions. A Poisson distribution co-efficient is used to correct the copies per partition using the equation = −ln(1-p), where p = fraction of positive droplets/ partitions. The real-time dPCR reduces the false-positive/false-negative results.
Precision	Varies depending on the target and multiplicity. In general, better at higher copy number in the range.	Higher with more replicates per sample and better than real time at lower concentrations.
Upper Limit of Quantitation (ULOQ)	ULOQ is 10^8 copies for most of the targets. A maximum of 10^8 is reported.	The platforms by the year 2022 offer up to 10^5 copies of ULOQ.
Lower Limit of Quantitation (LLOQ)	Varies (3 and above copies/ rxn) depending on the target.	More consistency was reported at lower concentrations in the range compared to real-time qPCR.
Multiplexing	Higher capability	Limited
Throughput	96 or 384 well format	Only 96 well format. However, some systems can run up to 8 plates at a time.
Run Time	Can go to <90 minutes	2 hours or higher
Inhibition of PCR	Can be monitored with an internal or external positive control	The inhibitors are diluted due to partitioning/digitization.
Cost	More economical, platforms, and trained staff are already available in most of the pharma.	Expensive initial start-up and training on the demerits/troubleshooting is warranted.

Table 2.

qPCR & dPCR—An overview.

9.1 Real-time digital PCR systems

The popular commercial digital PCR platforms are endpoint PCR systems. This has some drawbacks primarily due to false negatives and false positives. The false-negative result happens due to weak reaction efficiency and false-positive result due to water evaporation, cross-contamination, or non-specific amplification. A real-time digital PCR system has been constructed and tested [51]. This integrated system with 3D-printed parts exhibited high accuracy in the quantification of human 18 S ribosomal RNA gene fragment cloned in a plasmid DNA compared to QuantStudio™ 3D dPCR system. In two other studies, the Gnomegen real-time digital PCR system showed superior sensitivity and accuracy compared to the end-point digital PCR system and Quantstudio 5 real-time PCR system in the detection of SARS-COV-2 RNA and in the quantification of epidermal growth factor receptor (EGFR) exon 19 deletion, T790M, and L858R point mutations as well as human epidermal growth factor receptor 2 (HER2) amplification, respectively [52, 53]. Improvement in the integrated real-time digital PCR systems and availability from other manufacturers in the future will be very helpful in the PK-Biodistribution/PD analyses for gene/cell therapy studies and for point mutation detections. Overall, sequencing did not kill PCR. CRISPR gene-edited cell therapies are also showing promising results.