concentration-dependent
6,15,34,35
. Our findings are however in accordance with more recent
reports
8,11
that suggest RT is comparably efficient for different assays and template
concentration.
The RT yield showed large variation across tested enzymes and was priming strategy-
dependent (Figure 4). More pronounced differences between yields of various enzymes with
RT primers mixture suggest that especially SuperScript IV and Maxima H- can markedly
benefit from increased number of priming locations
35
. Multiple priming locations in combination
with strand displacement activity, lack of RNase H function, high processivity and strong
template binding affinity enable synthesis of more than just one cDNA copy from one RNA
transcript, thus yields > 100 %.
Differences in RT efficiency between single-cell and bulk templates were minor and
relatively consistent with exception of SuperScript II and eAMV enzymes (Figure 4). The worse
performance observed with SuperScript II and eAMV on the single-cell template could be
expected, as the RNA amount was below the recommended template range by the
manufacturers. Considering this, it is surprising that several popular single-cell RNA-Seq
protocols (CEL-Seq2
27
, Smart-Seq2
20
) are based on SuperScript II, suggesting there is a
potential for their improvement. Apart the discussed high and low efficient RTases, the cDNA
synthesis yield was typically in the range 50 – 80 %, which is in line with earlier studies
8,15–
17,23,34
. Only contradictory result is reported by Levesque-Sergerie et al. on SuperScript II and
III RTases
18
, which may be attributed to imprecise preparation of qPCR standard curves as
discussed by Miranda & Steward
19
.
Our study also reveals the relationship between performance of particular RTases and
their intrinsic biochemical properties (Table 1, Figure 5). The best-performing RTases,
SuperScript IV and Maxima H-, were thermostable, allowing to utilize higher reaction
temperature in the protocols. Previous reports have suggested that destabilization of
secondary RNA structures at elevated temperature leads to more frequent primer
hybridization and stable reverse transcription
6,36
. Our results support these hypotheses.
Interestingly, the majority of RTases employed protocols with the pre-incubation step aiming
to increase the efficiency of primer binding. The exception in our selection was iScript and
SensiScript. Both enzymes achieved relatively lower performance demonstrating the
importance of the pre-incubation step. However, if loss of sensitivity is not an issue, simplified
pipetting protocol, reduced possibility for contamination and errors may be advantageous
factors. RNase H activity was not observed to have any profound effect on the reaction
outcome, as previously reported
11,17
. The important factor for selection of RTase is also its
price. This is relevant especially for single cell RNA-Seq studies, where thousands of cells are
typically analyzed and price for RT is an important part of the budget. From this perspective,
Maxima H- may be a recommendable choice for many researchers for its high performance
and low price (second lowest in our comparison, Table 1). Notably, Maxima H- possesses also
the terminal transferase activity utilized in some RNA-Seq protocols
8–10
(mcSCRB-Seq, STRT-
Seq, SMARTer, Smart-Seq2).
Typically, high-throughput experimental results require reduction of their
multidimensional composition. PCA, the commonly used method, reduces dimensionality by
searching for the largest portions of variation in the data set. Observed variation, however,
arises not only from the biology of the experiment but also from technical factors of the
measurement, including RT
30
. In our model experiment, we hypothesized that higher
.CC-BY-NC-ND 4.0 International licenseunder a
certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available