4. Discussion
LASSO technology enables massively multiplexed capture of large DNA fragments for sequencing and/or expression cloning. A key factor for LASSO capture efficiency is the quality of the LASSO probe library in terms of probe sequence identity and probe representation. A previous assembly protocol we developed presented a number of drawbacks that resulted in a poor quality of the LASSO library with only ~10% of correctly assembled LASSO probes. The critical phase was the self-circularization by ligation of the LASSO probe precursor, wherein EcoRI digested probe ends were intramolecularly ligated to each other. Shukor S et al. [34] noted that when attempting to generate thousands of probes in a single reaction by this manner, there exists a strong possibility that intermolecular ligations would manifest as mismatched probe arms on a mature LASSO probe. The presence of the discordant probes in the mature LASSO libraries was responsible for a reduction of the efficiency, likely due to the depletion of PCR reactants used up for the amplification of low molecular weight unspecific DNA amplicons arising in the post capture PCR (Supplementary figure 1 table a).
To improve the quality of the LASSO probe libraries we developed a different LASSO assembly methodology that leads to the same mature LASSO probe configuration but uses the cre recombination of a custom plasmid (pLASSO) to supply the linker of the mature LASSO probe. The assembly process herein described starts with the cloning of the pre-LASSO library in the pLASSO vector and E.coli transformation. The multiplication of a pLASSO library in E.coli, reduces the possible skewing of the different LASSO probes in the library. The plasmid library was then subjected to Cre recombination of the two loxp sites oriented head-to tail in pLASSO resulting in the excision of a DNA minicircle containing the LASSO precursor in its final configuration (Figure 2).
The LASSO library we assembled targeted the same E. coli ORFs of our previous work [33] and the LASSO probes had identical design but displayed superior capture performance. This observation is in agreement with the higher percentage (~46%) of concordant probes present in the E.coli LASSO library. The median RPKM for targeted ORFs versus untargeted ORFs was much higher than produced with the previous LASSO assembly methodology - especially for shorter ORFs (~8 times higher). This finding suggests that a better quality of the LASSO probe library results in a higher capture efficiency and in the reduction of undesirable low molecular weight amplicons in post capture PCR (Supplementary Figure 1)
With a model system, we showed that the sensitivity of LASSO capture potentially allows for the massive parallel capture of DNA targets at the whole human genome scale. We also evaluated the ability of the new LASSO probes to capture two individual full-length ORFs from a total human cDNA. The genes β-actin and glyceraldehyde 3-phosphate dehydrogenase were captured thereby verifying that this new LASSO production method provides comparable capture efficacy to the first-generation method using a self-circularization reaction. Future evaluation of this method is necessary to evaluate the breadth of this LASSO technology for the creation of human protein libraries.
As novel long-read sequencing technologies emerge, there is an increasing need for novel target enrichment methods that allow highly multiplexed selection of kilobase-sized DNA. We expect that LASSO probes can find immediate applications for targeted construction of long-read sequencing libraries. LASSO probes can also be used for the rapid and inexpensive production of pooled ORFeome libraries that can be expressed using standard vectors for functional screening applications.