Tilt spectroscopy is an analytical technique that can be used when a rapid, comparable result is needed at each stage of the purification process to provide confidence in the titer at that stage and allow for further processing. Tilt spectroscopy has been used in biological manufacturing for over a decade. Common applications include plasmid concentration and purity, DNA concentration and purity, viral titer, ultrafiltration/diafiltration (UF/DF) monitoring, and empty/full capsid testing. The SoloVPE and FlowVPX systems are designed to ensure that method integrity is maintained at every step of the process.
The SoloVPE system does not rely on a fixed path length to perform sample measurements (illustration 1Conventional UV spectroscopy versus variable path length spectroscopy.). Pathlength is a variable value that captures multiple absorbance readings in less than 1 minute. The concentration is a fixed value and does not require sample dilution as in traditional UV-Vis spectroscopy.
The SoloVPE and FlowVPX systems guarantee linear results based on an R2 value of 0.999 or greater. To collect the data, the fully automated equipment is lowered to set the path length to zero and an algorithm finds 1 Au. The device collects data from up to 10 different path lengths to calculate slope regression. The layer thickness varies from 5 µm to 15 mm, in small steps of just 5 µm.
The Viper AAV application software has evolved into an automated solution specifically designed to calculate the AAV virus titer. For data collection, DNA and protein wavelengths must be specified. Four extinction coefficients are needed to calculate the virus titer concentration. The user would be responsible for entering the extinction coefficient for the 260 DNA. This can be found from the molecular weight or by entering the DNA sequence directly into the software to calculate the extinction coefficient. Test results are displayed on the screen within one minute. There is also a sequence entry for determining extinction coefficients using a DNA sequence.
PTC Therapeutics' Center of Technical Excellence in Gene Therapy has an integrated facility for process development, testing and GMP manufacturing of plasmid DNA and AAV products. PTC has a diverse research platform with a portfolio of six commercial products: five small molecule drugs and a gene therapy product, UpstazaMT(Eladogen exuparvovec).
Upstaza is an AAV2-based gene therapy for the treatment of aromatic L-amino acid decarboxylase (AADC) deficiency that was approved by the European Medicines Agency (EMA) in July 2022. This product marks the fourthlivegene therapy product to obtain regulatory approval in the US or Europe, underscoring its importance to the industry at large. Additional gene therapy candidates are in PTC's therapeutic pipeline.
Typical AAV purification process flow
When AAV is produced on an industrial scale in upstream cell culture, it undergoes a cell lysis and clarification step, followed by capture chromatography (using affinity chromatography or ion exchange chromatography (IEX)). After a capture step, the vector product undergoes a polishing step to separate full from empty capsids using IEX chromatography or another method. After the polishing step, the cleaned AAV goes through a UF/DF step before final formulation and filling finish.
The goal of the process development team is to provide a robust and scalable purification process to efficiently produce AAV. It is important to ensure that Critical Quality Attributes (CQAs) specifications are met for all process parameters. Key CQAs include the AAV genome titer, capsid titer, and capsid full-to-empty ratio.
Currently, PTC's in-house AAV quantification methods include Quantitative PCR (qPCR)/Droplet Digital PCR (ddPCR) to determine genomic titer and Capsid Enzyme Immunosorbent Immunosorbent Assay (ELISA) to determine capsid titer. These are industry standard test methods but have the disadvantage of a turnaround time of up to 2 weeks and a delay to the testing lab.
SoloVPE system evaluation
PTC Therapeutics worked with Repligen to evaluate the CTech SoloVPE system as a preliminary analysis method for quick and easy measurement of AAV titer. The system offers a direct measurement in less than 5 minutes per sample.
The Repligen team worked on site to help PTC Therapeutics set up the method and measure some in-process samples, including affinity elution, polish elution, and some intermediates from the tangential flow filtration step. During baseline, genome titer and capsid titer were measured using the SoloVPE system and these values were compared with genome titer obtained by qPCR and capsid titer obtained by capsid ELISA. The log10 difference was 7.4% for the SoloVPE versus qPCR system and less than 4% versus the ELISA method. The first evaluation showed that the data from the SoloVPE system showed good comparability with the qPCR and ELISA data from the capsid.
To assess the accuracy of the SoloVPE system, further in-process samples from downstream process development were measured, and genome titer and capsid titer using a SoloVPE system were compared to data collected using qPCR and ELISA methods (Figure 2Genome titration data acquired on the SoloVPE system compared to qPCR and capsid ELISA to assess accuracy.). Titer data from the SoloVPE system show excellent comparability with qPCR data in a linear trend with an R2 of 0.9842. Capsid titer also showed greater linear comparability between SoloVPE system data and capsid ELISA data with an R2 value of 0.9952.
The linearity of the SoloVPE system was then evaluated (Figure 3Genome titration data acquired on the SoloVPE system compared with qPCR and capsid ELISA to assess linearity.). A concentrated AAV sample was diluted to 12 different concentrations. Data from the SoloVPE system show excellent correlation with the expected qPCR and capsid ELISA values of the dilutions. The tested concentration range was 4.7 x 1011 - 2.6 x 1013 vg/mL for qPCR and 2.9 x 1012 - 1.6 x 1014 capsid/mL for capsid ELISA.
Then, the repeatability of the SoloVPE system was evaluated. AAV samples were measured in triplicate using the same Fibrette™ and sample aliquoted into the sample vial. The %RSD (relative standard deviation) is less than 2% for the Slope 260 and Slope 280 measurements.
The intermediate accuracy of the SoloVPE system was also analyzed. The AAV sample was measured by two scientists on different days using different fibers, sample tubes and aliquots. Again, the values for Slope 260 and Slope 280 had a small RSD of less than 1%.
After determining the accuracy, linearity, repeatability and intermediate position of the SoloVPE system, potential applications in downstream AAV process development were evaluated.
Figure 4Development of affinity chromatography with the SoloVPE systemshows a load capacity study for affinity chromatography. SoloVPE system technology was used to estimate the genome titer recovered in the elution and it was shown that three times the loading volume results in approximately three times the genome titer in the elution pool. The maximum load capacity of this column has not yet been reached at this higher load.
Alternative potential applications in affinity chromatography for the SoloVPE system include residence time studies, buffer additive studies, and elution buffer pH studies. SoloVPE can also be used in polishing chromatography. The SoloVPE vector titer and total capsid percentage measurement allow for rapid decision making in process development from the AAV polishing step based on total capsid recovery and enrichment to decide which buffer condition is optimal for each purpose.
Other potential applications for using SoloVPE in the polishing step include resin screening, buffer additive screening, loading capacity, residence time, pH/charge conductivity, elution buffer pH conductivity, and gradient/step elution studies .
The SoloVPE system can also be used in the UF/DF process. In this step, purified AAV is concentrated prior to buffer exchange and formulation. It is important to have in-process control for this step to ensure accurate concentration. Monitoring the product title during the process reduces the chance of unexpected product loss.
An experiment to measure genome titer was performed using the SoloVPE system for the starting material and three other samples during either the concentration step or the final product. The concentration factor was calculated using the SoloVPE system and agreed well with the volume reduction calculations. This demonstrates that the SoloVPE system can be used as a new process control method for UF/DF.
How comparable are the results of the SoloVPE system to the industry standard?
JF:There are two industry standards: qPCR and ELISA or ddPCR and ELISA.
If we compare the tilt spectroscopy technique with any of the other methods used, the results are always within the tolerance of this method, which unfortunately is quite high - from ± 20 to 40%. The only thing we can absolutely guarantee is that our method will be faster, more repeatable, and linear. Assessing accuracy against any kind of gold standard remains a challenge. Yan and I are still curious if there is a more accurate technique.
How accurate should the extinction coefficients and finish coefficients be? Is it acceptable to use theoretical values?
YC:If your AAV is wild-type, you can use a previously published AAV extinction coefficient, as all wild-type AAVs have highly conserved capsids.The difference in extinction coefficients between different serotypes can be less than 5%.
For new AAV serotypes that are completely different from existing ones, online tools can be used to calculate AAV extinction coefficients if you know the protein sequence. The calculation of the title depends on the coefficient used.
Have you used this method to measure titers in affinity chromatography? For example, could you use this method to measure titer in clarified crop?
YC:We did not use the SoloVPE system to measure AAV titer at harvest because the SoloVPE system is an A260 and A280 absorbance-based UV method that does not measure a signal provided by host cell protein or DNA.It cannot distinguish where the signal is coming from. There is a lot of host cell protein and DNA in the raw cell culture harvest, which can affect the accuracy of the results. I would not use the SoloVPE system to measure AAV titer in crude cell culture harvest.
JF:Can be used after chromatography.The technique is also unable to tell the difference between partial DNA or DNA. This isn't specific to the SoloVPE system or tilt spectroscopy, by the way - it's a UV issue.
What is the minimum sample volume required to obtain reliable data?
JF:The required sample volume is between 60–100 µL.However, tilt spectroscopy is not a destructive technology, so you keep your sample after the measurement.
How does this technique compare to purified vectors for in-process samples where contamination levels can be significant?
YC:When using this technique after the affinity step to measure an affinity purified sample or a further purified polish sample, impurity levels must be minimal and the SoloVPE system is accurate when making run-by-run comparisons.
However, if you capture your AAV using IEX chromatography, you may have significant contamination and the reading may not be very accurate. This depends on your purification method and the in-process sample you are measuring.
Is the required value for the 260 nm extinction coefficient for the complete capsid with DNA inside or just for DNA? What extinction coefficient values need to be entered into the software?
JF:This is the extinction coefficient at 260 for DNA.This is the only value that we didn't pre-populate in the software for the calculation, as it changes depending on the product you are testing.
We have included the calculator tool in the software that allows you to add your sequence to calculate this value for more accurate result. We found that the other wavelengths for the protein did not change significantly even when different concentrations or different serotypes were tested.
We consider other viruses. However, for a UV method to work, you need extinction coefficients. For example, there are currently no published extinction coefficients for lentiviruses, although we can certainly measure them using the information you provide.
Is the AAV capsid titer specific or is the intact capsid specific?
YC:This depends on the extinction coefficient you are using for the capsid or the specific AAV you are using.If you are working with a traditional wild-type AAV, it should not be AAV-specific. Working with a normal capsid with a different protein sequence and extinction coefficient, the AAV can be very different from traditional AAVs, which can be either capsid-dependent or capsid-specific.
Linearity of the SoloVPE system was checked for genome titers below 4.7?-11vg/ml?
JF:Yes, it does. We arrive at 2.0 × 10-11.With current technology, we wouldn't go any deeper. Louder is never an issue as there is a large linear range.
Realistically anything below 1011does not allow the acquisition of robust linear data, as the signal is very weak and there is no representative change in absorbance along the length of the path. We cannot guarantee measurement accuracy when we drop below 10-11.
Can the SoloVPE system be used for samples with completely unknown titles?
JF:If you go to the system with a blank, we can provide you with a ratio (R) value.However, if you need to calculate vg/mL, you will need available DNA sequence.
However, some of our customers want to make decisions for multiple batches of material based on the R-value alone. When they get to the formulation, I'll put together several different batches of material, and based on the R-value alone, they'll group batches with similar R-values together. and will send this larger batch for analysis. The point is that instead of paying and waiting for ten lots, you might only have to send two or three lots, depending on how closely your R-values are matched.
How would you find the extinction coefficient for AAV?
JF:We use existing publications.Sommer's paper is our primary source for three of the four extinction coefficients we use. For the second, we have to rely on you to provide your molecular weight or DNA sequence.
Are the SoloVPE system and software fully GMP compliant? How is the system and software qualified and what kind of support is provided during qualification?
JF:Our main business is in the GMP environment.We always plan for the future. All of our applications that use tilt spectroscopy have the ability for the software to be 21 CFR Part 11 compliant.
We provide software validation services and work with users to ensure their specific needs are met by IT groups so that we can provide a bespoke configuration for software configuration and also ensure compliance.
How long does it typically take for the SoloVPE system to be installed and validated for process development?
JF:Repligen will not be the bottleneck.Most of the time spent during validation is waiting for our customers to plan for installation and training. The more access we have to work with scientists, the faster this usually happens. Most method development and validation are typically completed within 6 to 12 months. However, there have been cases where we have dramatically improved these timeframes depending on how much access we have to work with the client and their current project needs.
What is the validation process and what validation support does Repligen offer?
JF:Our application specialists have all been trained to work directly with scientists to develop a robust method for each application you are working with and a system that experiment design and standard operating procedures would follow to adapt the method for laboratory use to release .
Our commitment is not just basic user education. We help with the heavy lifting required to define the parameters for method development and then help successfully implement the tests to validate the methods for release.
Can the SoloVPE system be used to test plasmid products received as raw material to verify purity?
JF:As long as the material is clean, the answer is yes.The serotype does not matter. Traditionally, we've considered a slope range of 1.8 to 2.0. We have articles and publications on our website that mention this.
What are the limitations of the SoloVPE system for AAV titer measurements?
YC:As the SoloVPE system is a UV-based method, anything that might disturb the absorbance at 260 and 280 nm could affect the accuracy of the result.Excipients can affect absorption, as can high levels of impurities including host cell protein and nucleic acid in your sample. These factors affect the accuracy of the SoloVPE system results.
Second, it does not distinguish between complete capsids and partial capsids because it does not distinguish specific DNA sequences.
Can the SoloVPE system replace current standard AAV titre assays?
YC:Whether you can use the SoloVPE system to replace the current standard AAV titre assay depends on each company's needs and its own assessment and validation methodology.
In certain cases, such as in-process control of the final UF/DF stage, it has great potential to replace traditional tests due to its speed and accuracy. For the other steps, it depends on the need and validation of each company. It is not intended to replace other standard assays during normal inclusive sample testing. We will be using the SoloVPE system as a complementary method for cross-run comparisons during process development to accelerate our development schedule.
What does the R value measure in the Viper software and what can you use this measurement for?
YC:The R value is the ratio of the 260 slope to the 280 slope of the Solo VPE reading.This value can be used to estimate the percentage of complete capsid in the newly measured AAV process sample. Using the equation provided in Sommer's article, you can use this R value to get a rough estimate of the total percentage of capsid in your sample.
When comparing and evaluating AAV empty capsids against the virulent form, does the optical density measurement provide values that would differ depending on the variant and gene of interest (GOI) in terms of extinction coefficient or slope spectroscopy values?
JF:We didn't see any difference, mainly because the DNA contribution is the star of ratio measurement.The wavelengths of the proteins are skewed and there is a contribution from the capsid, but compared to the DNA we didn't see any significant changes. Values do not necessarily change to the extent that it makes a significant difference in results. We see a difference of 2-5% depending on the concentration range, which doesn't affect the sample too much.
Does this title analysis require the use of Viper software or can it be performed using legacy C Technologies software?
JF:Our Viper platform software is our next generation software.Both software can be used technically - however, the older version of the software makes it more complicated and challenging for scientists to get results due to lack of automation.
The new platform can collect all your data and the software can do all the calculations for you in less than a minute. This would not be possible with older versions of the software.
Does the method need to be qualified for in-process samples?
YC:Personally, we do not qualify the method for in-process sample measurement, as we use it for a run-by-run comparison.However. it depends on the needs of each company.
Joe Ferraiololeads the bioanalytical applications team and is responsible for the SoloVPE variable path spectroscopy system for in-line applications. He has been with the company for over 20 years and has over ten years of experience developing and validating analytical applications. He specializes in UV analysis and leads the development and commercialization of high quality products and flexible solutions that address critical steps in the manufacture of biologics.
Yan Chenis a principal investigator in downstream process development - gene therapy at PTC Therapeutics. He has over 14 years of experience in biology. Yan has worked at several pharmaceutical and biotechnology companies, developing purification processes for biologics, including mAbs, recombinant proteins and viral vectors. His experience includes downstream process development for early and late stage programs, technology transfer and process characterization. He received his PhD in cell biology from New York University.
Director, At-line Bioanalytical Applications,
downstream process development, gene therapy,
Authorship and Conflict of Interests
Posts:All named authors accept responsibility for the integrity of the work as a whole and have consented to the publication of this version.
Disclosure and Potential Conflicts of Interest:Yan Chen announces that he has stock options in PTC Therapeutics. Joe Ferraiolo has no conflicts of interest.
Financing statement:Yan Chen received funding for researching, authoring, and/or publishing this article from PTC Therapeutics.
Article and copyright information
Copyright ©:published byInsights into cell and gene therapyunder the Creative Commons CC BY NC ND 4.0 license certificate which allows anyone to copy, distribute and transmit the article as long as it is duly attributed in the manner indicated below. No commercial use without permission.
assignment:Copyright © 2022 Repligen Corporation. Published byInsights into cell and gene therapysob Creative Commons License Deed CC BY NC ND 4.0.
Article source:This article is based on a webinar that can be foundHere.
Webinar held at:August 10, 2022;Revised manuscript received:September 19, 2022;Release date of:13. October 2022.