A descriptive guide for absolute quantification of produced shRNA pseudotyped lentiviral particles by real-time PCR

Gene silencing techniques, including RNA interference methodologies, are widely used in reverse genetics to study the role of specific genes in biological processes. RNA interference has become easier to implement thanks to the RNAi Consortium (TRC), which has developed libraries of short hairpin RNA (shRNA) sequences in pseudotyped lentiviral particles capable of targeting most genes in the human and mouse genomes. However, a problem is the lack of a simple method to titrate the homemade lentiviral particle product, making it difficult to optimize and standardize shRNA experiments. Here we provide a guide describing a quick, non-laborious and reliable method for the titration of TRC pseudotyped lentiviral particles that is based on the detection and measurement of viral RNA using quantitative PCR. Our data demonstrate that purified linearized shRNA plasmids represent more suitable standards than circular or unpurified linearized plasmids. We also show that for precise absolute quantification, it is important to determine suitable plasmid and viral cDNA concentrations in order to find the linear range for quantification, as well as to reduce inhibition and primer dimer amplification. Finally, we show that the lentivirus concentration impacts the level of knockdown in transduced cells. Primers utilized in this non-functional titration can potentially be applied to functional titration of proviral DNA copies or transgene expression, overcoming problems arising from the absence of fluorescent reporter genes in TRC plasmids.


BACKGROUND
The RNAi Consortium (TRC) has developed genome-scale short hairpin (sh)RNA libraries targeting human and mouse genes [1], facilitating lentiviral-based RNA interference (RNAi). Quantification of lentiviruses is important for developing robust and reproducible RNAi protocols. Two types of methods have been elaborated to quantify viral titers: functional titrations, determining the concentration of viral particles needed to transduce a cell; non-functional titrations, determining the number of viral particles secreted by producer cells.
In the context of functional titration, if a fluorescent reporter gene, e.g., green fluorescent protein [2], is transduced alongside the sequence of interest, the functional titer can be determined by flow cytometry. However, this requires a fluorescent marker and cannot distinguish single from multiple proviral DNA integration sites within the host genome. An alternative is to estimate the number of integrated proviral DNA copies per cell by quantitative (q)PCR [2]. Because proviral copies can be inserted into different chromatin regions, the transgene expression can vary [2]. Quantification of the expressed transgene by reverse transcription (RT)-qPCR instead of the proviral DNA copies gives a closer estimation of the transgene expression [2].
Functional titrations require transduction, which is time-consuming, and efficiencies depend on cell type and transgenes [2,3]. Non-functional approaches thus present an easier alternative for determining the titer of How to cite this article: Mournetas V, Pereira SM, Fernig DG, Murray P. A descriptive guide for absolute quantification of produced shRNA pseudotyped lentiviral particles by real-time PCR. J Biol Methods 2016;3(4):e55. DOI: 10.14440/jbm.2016.142 a lentiviral particle batch for optimization and standardization. These methods mostly rely on the measurement of either the expressed viral protein p24 or of viral RNA. p24 measured by enzyme-linked immunosorbent assay has reliability issues, due to the presence of a variable amount of p24 originating from defective particles and non-particle-associated p24 [4]. Alternatively, several groups have measured viral RNA, either directly by qPCR [5] or indirectly by RT-qPCR [2,4,6,7]. Even if viral RNA-based titrations also measure defective particles to some extent [6], it has been shown that mainly full-length viral RNA transcripts are incorporated into pseudotyped lentiviral particles [4]. To our knowledge, only one study has described PCR primers suitable for TRC plasmid quantification [7]. However, these primers were within the lentiviral capsid sequence, which is not integrated into the host genome after transduction, limiting their use to non-functional titration.
Here we develop a descriptive guide for absolute titration of homemade TRC shRNA pseudotyped lentiviruses by RT-qPCR using TRC shRNA plasmids as standards. Plasmid amplification and purification followed by pseudotyped lentiviral production take less than 5 d, and then 2 d are needed to determine absolute titers (Fig. 1).
We demonstrate that the conformation of the plasmid and its purity (circular, linearized or purified linearized) affect qPCR efficiency, and that absolute quantification requires both the plasmid and viral cDNAs to be within an appropriate concentration range. Designed PCR primers are suitable for functional titration via the quantification of integrated proviral DNA copies and/or expressed transgene. Moreover, we show that lentivirus concentration correlates positively with the level of knockdown in transduced cells. Figure 1. From plasmid amplification to pseudotyped lentiviral particle quantification. Main steps in the timeline of pseudotyped lentiviral particle production, starting with plasmid amplification and ending with pseudotyped lentiviral particle absolute quantification.  (Fig. 2).

TIP:
To ensure reproducibility, we advise the determination of standard stability during storage as time and temperature can significantly affect quantification [10].

4.3.
Transiently co-transfect cells using Sigma Calcium Phosphate Transfection kit by adding plasmids of the second-generation packaging system at 2 µg/well and at a ratio of 4:2:1 of TRC transfer vector [1] (Table S1): Packaging vector psPAX2: Envelope vector pM2D.G.

CAUTION:
Manipulating the amount of plasmids (transfer, packaging or envelope) can affect the non-functional titer without altering the functional one [11].

5.2.
Freeze cell medium aliquots in crushed dry ice.

Stored at -80°C.
NOTE: Medium from non-transfected HEK-293TN cells was used as a negative control for absolute quantification of the pseudotyped lentiviruses.

CAUTION:
The workable concentration range for qPCR is a key parameter: low viral cDNA concentrations tend to overestimate the absolute number of pseudotyped lentiviruses due to the presence of primer dimers, whereas high viral cDNA concentrations can inhibit the PCR, resulting in the number of lentiviruses being underestimated.

8.5.
Run qPCR products on a 2.5% (w/v) agarose gel with the HyperLadder V as molecular size marker to verify product sizes.
NOTE: qPCR product specificity is validated by comparing qPCR product melting peak and gel electrophoresis of the viral cDNA templates with NRT, NTC and negative controls.
Lentiviral particle absolute quantification 9. Construct standard curves 9.1. Use purified linearized plasmids as standards for absolute quantification.

9.3.
Use the concentrations yielding the best technical replicates in terms of their similarity of signal and the absence of primer dimers (checked using melting peak data) to generate standard curves.

NOTE:
A factor of 2 is applied to N to take into account that plasmids are double-stranded while viral RNA and cDNA are single-stranded. [21]. A slope of -3.32 is equivalent to E = 1 (or 100%).

11.
Use the same standard curves to establish the pseudotyped lentiviral particle absolute titer as follows:

NOTE:
Each lentivirus carries two RNA copies implying a theoretical ratio of ½ [22].

TIP:
To better estimate the absolute titer, more than one cDNA concentration was utilized to calculate the final quantification.
Cell culture, transduction and knockdown analysis NOTE: This section uses hESCs as a model cell line but other cell lines could be used instead.

12.3.
Change culture medium every two days.

13.3.
Add lentiviruses to the cell medium.

NOTE:
Concentrations of 18 pg viral RNA/cell (C1) and 36 pg viral RNA/cell (C2) were used in the present study.

CAUTION:
For different shRNAs, equivalent amounts of extracellular RNA may not equate to equivalent concentrations of the respective lentiviruses.

13.5.
Incubate cells for two more days.

13.7.
Add puromycin to the medium at a concentration of 0.5 μg/ml for selection.
At day 5: 13.8. Lyse cells for RNA extraction using TriZOL reagent.

Knockdown analysis
CAUTION/TIP: RT efficiency can be impacted by different factors such as the type of RT enzymes used [16,17], the amount of RNA [17], RNA quality (integrity [18,19] and purity [17]) and the priming strategy [23].

14.4.
Relative quantification: extract raw qPCR data and calculate the relative quantification using the method [27].

NOTE:
Raw qPCR data were extracted and analyzed with Bio-Rad CFX Manager 3.1 software in the present study. Statistical analysis was performed using IBM SPSS Statistics 21 software.

Convert viral RNA concentration into number of viral particles if need be.
14.5.2. Plot the knockdown percentage as a function of the absolute number of viral particles (plot each target gene independently).

14.5.3.
Use a logarithmic fitting curve to describe the correlation.

Optimization of the qPCR standard
To obtain a precise and accurate measurement of the absolute titer of pseudotyped lentiviral particles, real-time qPCR standards have to be as reliable and efficient as possible. According to our results, when using plasmids as standard, the purified linearized form was the most appropriate, with a qPCR efficiency (E) of 84% and 73% in the given examples ( Fig. 2A and 2B). For circular plasmids, E was 37% and 43%, suggesting that the circular shape had a negative effect on the PCR reaction. The PCR reaction was also inhibited when the linearized plasmid was not purified (E = 38% and 32%).
The workable concentration range for qPCR also needs to be established by: (1) verifying the absence of primer dimers (especially for the low cDNA concentrations, from 0.3 pg/μl in the present case) using qPCR product melting peaks (Fig. 2C) and gel electrophoresis ( Fig. 2D); (2) checking if the highest cDNA concentration(s) lower E. Our data demonstrated that removing the two highest tested plasmid concentrations of 5 ng/µl and 1 ng/μl improved E by 19% ± 8% ( Fig.  2B-2E, Table 1, Fig. S2).

Specificity of the real-time qPCR
Viral RNA should be DNase-treated prior to RT to avoid quantification bias due to contaminant plasmids from transfected HEK-293TN cell medium. The real-time qPCR must contain the following controls: (1) cDNA from viral RNA of untransfected HEK-293TN cell medium as negative control; (2) DNase-treated viral RNA as NRT and (3) NTC. The qPCR melting peaks, as well as gel electrophoresis of the controls were compared to those from samples for assessing product specificity (Fig. 3A-3F). None of the tested cDNA concentrations from non-transfected HEK-293TN cells gave a specific amplification of our product of interest (Fig. 3D-3F). Calculated efficiencies of six different purified and linearized shRNA plasmids. A: with the both highest tested concentrations; B: without the highest concentration of 5 ng/reaction; C: without the both highest tested concentrations of 5 and 1 ng/reaction. or without the both highest tested concentrations of 5 and 1 ng/reaction (blue). Linear regression curves are given with their correlation coefficient R 2 .
The blue curve gives a slope of -3.52 which is the closest to -3.32, the ideal slope resulting in an amplification efficiency of 100%.

Absolute quantification of the pseudotyped lentiviral particle
Similarly to the plasmids, our data demonstrate that it is very important to determine the workable concentration range suitable for viral cDNA amplification. Low viral cDNA concentrations resulted in non-specific amplification of primer dimers (Fig. 3A-3C), leading to an overestimation of the absolute number of pseudotyped lentiviral particles (Fig. 3G). High viral cDNA concentration led to PCR reaction inhibition and, therefore, to the underestimation of the absolute number of pseudotyped lentiviral particles (Fig. 3H). Because the suitable concentration range for titration was limited (usually between 1 and 7.5 ng/reaction maximum), calculating the cDNA amplification efficiency was not relevant.

shRNA lentiviral concentration and knockdown, a correlation
The highest viral RNA concentration triggered higher knockdowns (Fig. 4A), but no correlation was apparent between RNA concentration and knockdown efficacy. This may be due to the measurement of lenti-viral shRNA concentration in terms of extracellular RNA, which may be confounded by contributions from cellular RNAs. Thus, for different shRNAs, equivalent amounts of extracellular RNA may not equate to equivalent concentrations of the respective lentiviruses. In contrast, when the conversion into absolute number of viral particles was done using the method described above (Fig. 1), there was a positive correlation with the knockdown efficacy (Fig. 4B). Target genes were expressed at different levels in hESCs in terms of qPCR Ct range: OCT4 (21.5 < Ct < 26.5, control cells; 26 < Ct < 39, knockdown cells) was more highly expressed than NRP1 (27.5 < Ct < 31.5, control cells; 27 < Ct < 33, knockdown cells), which was highly expressed than PLXNB1 (30.5 < Ct < 36.5, control cells; 33 < Ct < 39, knockdown cells). Consequently, the correlation was stronger when each gene was analyzed individually (Fig. 4C-4E). The differences in knockdown efficiency might be linked to the expression level of each gene [28] and/or to the shRNA sequences themselves [29]. Moreover, the correlation seemed to be more logarithmic than linear, which is consistent with published results of expressed transgene quantification by RT-qPCR in mesenchymal stem cells [30]. Table 2 recapitulates the main sensitive points for achieving a reliable absolute quantification of produced lentiviral particles. Use suitable qPCR controls (e.g. NRT, NTC*)

Construct the qPCR dissociation curves
Run agarose gels to verify qPCR product sizes

CONCLUSIONS
Many studies have demonstrated that shRNA lentiviral-based approaches provide efficient and stable knockdown in various cell types. The present guide is a simple way to non-functionally quantify the absolute number of produced pseudotyped TRC shRNA lentiviral particles in a batch. The quantification can be used prior to transduction for protocol standardization and optimization, such as finding the best producer cell line or investigating the optimal harvesting period. Moreover, this non-functional titration rapidly provides insight into the knockdown efficacy that can be achieved, without doing a functional titration. The same primer pair can potentially be used for functional titration of both the integrated proviral DNA copies and the expressed transgene level in mouse and human cells. Table S1. shRNA plasmid references and sequences. Table S2. PCR primer references and sequences. Figure S1. Partial pLKO.1-puro plasmid sequence. Figure S2. Effect of the highest standard concentrations on qPCR efficiencies.

Supplementary information
Supplementary information of this article can be found online at http://www.jbmethods.org/jbm/rt/suppFiles/142.