Harnessing CRISPR-Cas adaptation for RNA recording and beyond

Fig. 1. Overview of the mechanism and classification of CRISPR-Cas systems. (A) Schematic of the CRISPR-Cas adaptive immunity. During the adaptation stage, foreign DNA (gray helical lines) or RNA (pink lines) can be captured and integrated into CRISPR array, by Cas1-Cas2 and (Cas6-)RT-Cas1-Cas2 complexes, respectively. Acquired nucleic acids are converted to new spacers and recorded in a leader-proximal order, creating immunological memories. In the expression stage, spacers in CRISPR array are transcribed to pre-crRNA, and processed into crRNAs (crRNA biogenesis). The effector RNP complex are assembled crRNAs with expressed effector Cas proteins. At the final interference stage, effector complexes target against revisiting foreign (viral) DNAs and their transcripts through sequence-specific cleavage complementary to the sequence of crRNA. The image is adopted and modified from Lee et al. Trends in Biochemical Sciences, 2022 (8). (B) Classification and gene structure of CRISPR-Cas systems. In adaptation associated cas genes, there are well conserved genes, Cas1 and Cas2. Some accessory genes exist like Cas4 in type I, II and V, or reverse transcriptase (RT) in type III and VI, or Csn2 in…

Fig. 2. Structural architectures of CRISPR integrases and related proteins. (A) Structure of E. coli type I-E DNA-unbound (apo) Cas1-Cas2 complex (Protein Data Bank, PDB, ID: 4P6I) (50). (B) Structure of E. coli type I-E prespacer DNA-bound Cas1-Cas2 complex (PDB ID: 5DS4) (52). (C) Structure of Thiomicrospira Cas6-RT-Cas1-Cas2 complex (PDB ID: 7KFU) (75). Two of the four Cas6-RT domains are invisible. (A–C) The arrangement models of the color-coded subunits and DNA (in B) are represented above the crystal (in A and B) and cryo-EM (in C) structures of the complexes. (D) Structure of Marinomonas mediterranea (MMB-1) truncated Cas6 domain (PDB ID: 6DD5) (76). (E) Structure of group II intron reverse transcriptase bound to an RNA template base paired with a DNA primer (PDB ID: 6AR3) (82).

Fig. 3. Domain structure and Phylogenetic distribution of the RT-fused Cas1 and its related proteins. (A) Schematic of the domain organization of HIV-1 RT, TeI4c group II intron RT, A. platensis type III-B RT-Cas1, MMB-1 type III-B RT-Cas1, and E. coli type I-E Cas1. Conserved RT motifs (1 to 7) are indicated in black boxes. Conserved motifs in mobile group II intron and non-LTR-retrotransposon RTs (0 and 2a) are labeled in red boxes. The YXDD sequence found in motif 5 represented with two aspartic acid residues is indicated by arrow. The X/Thumb domains commonly found in HIV-1 and group II intron RTs are indicated. Amino acid numbers are indicated below the bars. D, DNA binding domain. En, endonuclease domain. The image is taken from Silas et al. Science, 2016 (69). (B) Phylogenetic tree of Cas1 associated with RTs. The analysis is reconstructed with 148 Cas1 proteins. The identified clades with numbers are named and colored according to the RT-associated clade. The figure is taken and modified from Toro et al. Scientific Reports, 2017 (73).

Fig. 4. In vivo transcriptional recording by RT-Cas1-Cas2. (A) Comparison between RNA-seq and Record–seq. RNA-seq captures the transcriptomes from subject cells at each point in time, providing a transient snapshot of cellular events. By contrast, transcriptional records in CRISPR arrays provide molecular records that can be used to reconstruct transcriptional events that occurred over time only by a one-shot analysis. The image is adopted and revised from Schmidt et al. Nature, 2018 (88). (B) Schematic of experimental platform for the noninvasive interpretation of transcriptional recordings using F. sac RT-Cas1-Cas2 in bacterial cells. The E. coli cells transformed to express RT-Cas1-Cas2 proteins and harbor plasmids with CRISPR arrays are orally gavaged into mice. Stress response patterns by bacterial strains as sentinel cells are stored in CRISPR arrays in an inheritable and time-ordered way. Eventually, the recorded information can be reconstructed by sequencing the CRISPR arrays from microbial samples without any effort for the time-course harvest. The figure is taken and modified from Schmidt et al. Science, 2022 (101).