Improved LbCas12a variants with altered PAM specificities further broaden the genome targeting range of Cas12a nucleases

LbCas12a-RVRR combines four PAM-altering substitutions — G532R, K538V, Y542R and K595R — the same mutation set found at the analogous positions of AsCas12a-RVR and AsCas12a-RR, transferred onto LbCas12a. Combining the RVR and RR mutations this way abolishes nuclease activity entirely in the AsCas12a and FnCas12a orthologs, but LbCas12a-RVRR and MbCas12a-RVRR retain considerable activity; LbCas12a-RVRR remains compatible with the TTTC, TATC, TTCC and TCCC PAM sequences optimal for LbCas12a-WT, -RVR and -RR, respectively, though its activity on TCTC is comparatively low.

On PAM sequences with a fixed -3 thymine (TTTN), LbCas12a-WT is most active on TTTV (V = A, C or G) and retains lower but considerable activity on TTTT, whereas LbCas12a-RVRR is active on all four TTTN sequences, including TTTT, with its highest activity on TTTA. On PAM sequences with a fixed -3 adenine (TATN), LbCas12a-RVR is robust only on TATV, while LbCas12a-RVRR is active on all four TATN sequences, with higher activity when position -1 is a purine (TATA or TATG).

A cytosine at PAM position -3 is disfavored by LbCas12a-WT and LbCas12a-RVR, but LbCas12a-RR shows robust activity on TCTN sequences; LbCas12a-RVRR shows only moderate activity on TCTV and negligible activity on TCTT. LbCas12a-RVRR is also active on TGTN PAM sequences — a motif on which LbCas12a-WT, -RVR and -RR all show decreased or negligible activity — reaching its highest activity with an A or C at position -1 (TGTA or TGTC), making TGTN the only PAM class newly accessible through the RVRR combination rather than inherited from either parent variant.

Across all positions tested, LbCas12a-RVRR shows activity on 15 of the 16 possible TNTN PAM combinations, the sole exception being TCTT, giving it a consensus PAM resembling TNTN with a residual preference for a -3 thymine and, in some combinations, tolerance for a -2 cytosine. LbCas12a-RVRR additionally shows robust activity on the TACC, TTCC and CTCC PAM sequences associated with the RVR and RR parent variants.

Because the rarity of the "TTT" motif limits the standard EGFP-disruption assay to only two testable targets for Cas12a nucleases, a HEK293 reporter line carrying a codon-optimized clover sequence enriched in TTT motifs was generated to evaluate LbCas12a-RVRR activity on 98 genomic targets. LbCas12a-RVRR shows its highest disruption activity on TTTV-PAM targets, moderately lower but still substantial activity on TAYV, TGTV, TTCV, CCCV and TCCV PAMs, and comparatively lower activity on CTCV and TCTV PAMs.

On 17 endogenous targets with TTTV PAM sequences, the RVRR mutations do not reduce LbCas12a's indel-inducing activity relative to LbCas12a-WT, and on 5 targets with the WT-disfavored TTTT PAM, the RVRR mutations significantly increase indel formation. Across 18 targets each with TCTN or TGTN PAM sequences, the RVRR mutations significantly increase indel activity on TGTV and TGTT PAMs and produce a smaller but detectable increase on TCTV and TCTT PAMs.

In GUIDE-seq off-target profiling on seven endogenous TTTV-PAM targets in HEK293 cells, LbCas12a-RVRR cleaves at more off-target sites than LbCas12a-WT at six of the seven targets, the exception being FANCF target 1. LbCas12a-WT's off-target sites typically retain a canonical TTTV PAM or carry only a single T-to-C substitution, whereas LbCas12a-RVRR also cleaves off-target sites whose PAM sequences differ at all three thymidine positions, consistent with its broadened on-target PAM range; restricting the comparison to LbCas12a-WT's own off-target set shows only a slight increase in the off-target-to-on-target read ratio for LbCas12a-RVRR.

Replacing the inactive LbCas12a-WT domain of the dCpf1 cytidine base editor with inactive LbCas12a-RVRR, to create dLbCas12a-RVRR-BE, preserves C-to-T editing efficiency on TTTV-PAM targets relative to dLbCas12a-WT-BE while additionally enabling editing on TNTN-PAM targets that dLbCas12a-WT-BE cannot access. Across 21 predicted off-target sites carrying up to four mismatches, dLbCas12a-RVRR-BE shows no increase in off-target base editing compared to dLbCas12a-WT-BE.

Introducing the D156R mutation — the LbCas12a residue analogous to the activity-enhancing E174R mutation of AsCas12a — into LbCas12a-RVRR produces impLbCas12a, which carries the mutations D156R, G532R, K538V, Y542R and K595R. D156R increases impLbCas12a's average activity approximately 1.3-fold over LbCas12a-RVRR, driven mainly by enhanced activity on TTTV, TCCV, CCCV, TATC and TACV PAM sequences, with the largest individual gains on TACV (1.9-fold) and TATV (1.7-fold) — a smaller enhancement than the roughly two-fold increase E174R confers on AsCas12a for TTTV-PAM targets.

The D156R mutation also broadens impLbCas12a's PAM preference to include NTTV PAM sequences that LbCas12a-RVRR alone does not efficiently target, mirroring the effect of the analogous E174R mutation on AsCas12a and indicating that this activity- and PAM-broadening effect at PAM position -4 is transferable between the As- and LbCas12a orthologs.

impLbCas12a does not exceed the activity of enAsCas12a on targets with NTTV PAM sequences, but it exceeds enAsCas12a's activity on targets with RR- and RVR-related PAM sequences. Because impLbCas12a and enAsCas12a have overlapping but distinct PAM and target preferences, using them together provides access to a combined target space larger than either nuclease offers alone for genome-editing and transcriptome-modulation applications.

An inactive impLbCas12a variant fused to ten tandem SunTag repeats recruits transcription-factor-conjugated antibody fragments to the SPRN gene promoter and activates transcription using either canonical TTTV-PAM targets or PAM-divergent targets accessible only through impLbCas12a's expanded PAM range, reaching activation levels comparable to the equivalent WT-based SunTag system.

The canonical TTTV PAM required by Cas12a nucleases occurs at a theoretical frequency of 3/256 in DNA sequence, compared with 16/256 for the NGG PAM of SpCas9, which limits Cas12a's usable target sites in mammalian genomes. AsCas12a-RVR and AsCas12a-RR raise this theoretical frequency to 6/256 and 9/256 respectively while retaining the canonical TTTV preference of the WT enzyme, and LbCas12a-RVRR further increases the theoretical target frequency to roughly three-fold that of the RR variant, which had previously offered the broadest PAM space among LbCas12a variants.