Fig 1.
Schematic overview of conventional insertional mutagenesis, TIM, and TIM-tagging in C. reinhardtii.
(A) Conventional insertional mutagenesis to create mutants randomly. Donor DNA encoding a selectable marker is introduced into the cell; it then inserts randomly into the nuclear genome, causing a large insertion, deletion, and/or reorganization at the site of integration. (B) Targeted insertional mutagenesis (TIM) to disrupt a specific gene of interest. Based on insertional mutagenesis, but a CRISPR/Cas9 RNP is included in the transformation step and short gene-specific homology arms are added on both ends of the donor DNA. The RNP creates a DSB at the target site, which facilitates the incorporation of donor DNA at this site. The homology arms further increase the probability that the donor DNA will insert at this site, but they are not critical for success. Transformants expressing the selectable marker are screened by gene-specific PCR analysis to identify those in which the target gene has been disrupted. (C) TIM-tagging to specifically tag a gene of interest, as illustrated here for C-terminal tagging. Based on TIM, but the RNP target site is placed close to and upstream of the intended tagging site. The donor DNA starts with an essential homology arm to promote precise integration via homology-directed repair (HDR) at this end of the donor DNA, and then continues with the rest of the gene, including the tag sequence, followed by a selectable marker. A second homology arm at the 3’ end of the donor DNA is optional but might be helpful. In the final result, the allele contains the tag sequence, and is followed by the selectable marker. Following selection of transformants, cells that have integrated the tag sequence are identified by gene-specific PCR analysis. Green star: start codon; red star: stop codon; black and white boxes: homology arms on donor DNA or sequence on gene matching homology arms.
Fig 2.
Schematic overview of TIM-based C-terminal tagging strategy.
The target gene is illustrated at top, with the protospacer adjacent motif (PAM) shown as a blue bar and the stop codon as a red asterisk; the target RNP-cut site is located upstream of the stop codon. Donor DNA is designed to include, successively, a “left” homology arm (L) homologous to sequence upstream of the target cut site, gene sequence downstream of the target cut site, including a PAM with silent mutation (white bar on donor DNA), the desired tag sequence (yellow oval), the stop codon (red star), 3’ end non-coding sequence (light grey box), and the drug-resistance gene (green oval). The RNP can cut the wild-type gene, creating a DSB, but not the donor DNA or the tagged gene due to the silent mutation at the PAM site. When, with the help of the left homology arm, the donor DNA is inserted into the cut site without any modification, a tagged wild-type gene is formed as illustrated here. This gene carries the silent mutation at the PAM site. The drug-resistance cassette from the donor DNA is located downstream of the gene. A portion (indicated by a red line) of the original gene remains immediately downstream of the drug-resistance cassette. Following selection of colonies by growth on antibiotic-containing agar, clones are screened by PCR using a 5’-end primer (P1) complementary to sequence upstream of the donor DNA region and a 3’-end primer (P2) complementary to sequence on, or downstream of, the tag. In the latter case (shown here), product amplified from the wild-type gene will be smaller than that from the tagged gene. The sizes of the expected PCR products for the specific experiment shown in Fig 5 are indicated between the primers.
Fig 3.
Schematic overview of TIM-based N-terminal tagging strategy using a single RNP.
The 5’ end of the gene is shown as a grey box. The start codon is shown as a green star. The RNP target cut site is located downstream of the start codon. The PAM is shown as a blue bar on the gene. The donor DNA begins with a drug-resistance gene (green oval), then sequence homologous to the 5’ end of the target gene and continuing through the start codon, then sequence encoding the desired tag (yellow oval), and finally sequence homologous to that of the target gene beginning immediately after the start codon, including the RNP PAM modified with a silent mutation (white box), and continuing downstream for at least 20 bp (the “right” homology arm). The RNP can cut the wild-type gene, creating a DSB, but not the donor DNA or tagged gene if the donor DNA end is used as template to repair the DSB (upper diagram of “successfully targeted gene”). The new 5’ end created by the DSB also can be used as a template for repair, resulting in a tagged allele with a wild-type PAM sequence (lower diagram of “successfully targeted gene”). In these tagged alleles, the drug-resistance cassette is upstream of the gene. A portion (indicated by a red line) of the original gene remains immediately upstream of the drug-resistance cassette. To screen for positive clones, a 5’-end primer (P1) is designed to be complementary to the tag sequence (as illustrated here) or to sequence upstream of the tag; the 3’-end primer (P2) is designed to be complementary to sequence downstream of the homology arm.
Fig 4.
Schematic overview of TIM-based N-terminal tagging strategy using two gRNAs.
The 5’ end of the gene is shown as a grey box, the start codon is indicated by the green star, and the PAMs are indicated by blue bars. RNP1 cuts upstream of, but as near as possible to, the start of the gene, while RNP2 cuts downstream of the start codon. The donor DNA is identical to that of Fig 3 except that it starts with a left arm (L) homologous to genomic sequence just upstream of the RNP1-cut site. In the ideal situation, the DSB created by RNP1 and RNP2 will be repaired by homologous integration of the donor DNA facilitated by the left and right homology arms, resulting in a wild-type gene tagged at its N-terminus, and a drug-resistance gene just upstream of the tagged target gene. During integration, the 3’ end of the donor DNA or the genomic DNA’s new 5’ end created by the DSB can be used as a template, creating either a tagged allele with a mutated RNP2 PAM sequence (top diagram of “successfully targeted gene”) or a tagged allele with a wild-type RNP2 PAM sequence (bottom diagram of “successfully targeted gene”), respectively. Following selection of colonies on antibiotic-containing medium, colonies are screened by PCR using primer pairs P1/P2 and P3/P4 to identify those in which the insertion has occurred as planned. The sizes of the expected PCR products for specific experiments of Figs 8 and 9 are shown between the primers.
Fig 5.
Initial PCR screening of transformants from the LF5-HA C-terminal tagging experiment.
DNA from the wild-type strain (W), 94 transformants, and the plasmid pLF5HA, which contains the LF5-HA gene, were used as templates for PCR screening with primer pairs P1 and P2 as illustrated in Fig 2. The expected size for the wild-type product was 455 bp (marked with red * on the right side of the gels and below the bands). In theory, one small fragment (563 bp) and one large fragment (3576 bp) would be predicted as products resulting from PCR amplification of the tagged gene in this experiment, yet only the small fragment was observed (marked with green arrows on the right side of the gels and below the bands). There are several likely reasons for this. First, we used a PCR elongation time optimal for 563-bp products, so there would not have been enough time for the polymerase to amplify products as large as 3576 bp. Second, to facilitate screening of a large number of clones, we used crude DNA as template, and in our experience it is more difficult to amplify longer PCR products from crude DNA. Third, due to the high GC content of C. reinhardtii DNA, it is harder to amplify long PCR products than short ones. The presence of the wild-type product indicates that the cell line was not edited at the target site, and the antibiotic-resistance gene has inserted into the genome somewhere else. The absence of the wild-type product indicates that the cell line was edited at the RNP-cut site. Among these 94 cell lines, 11 yielded products of the size expected from the tagged gene, suggesting that the donor DNA was inserted as desired. The PCR products from these 11 strains were sequenced for further confirmation: 10 had the sequence expected for HDR; one (here marked with a red S underneath the band) did not (S1 Appendix). Among the 10 clones with the expected sequence, eight expressed LF5-HA (see Fig 6); only two (here marked with a red P) did not. Among the eight clones that expressed LF5-HA, all but one (here marked with a red U) had a complete 3’ UTR (see Fig 7). The complete absence of PCR product or product of a size other than that expected from the wild-type or tagged gene indicates that a more complicated editing event occurred. These are likely to be lf5 mutants. Six products (L1-L6), the sizes of which range from 1235 bp to 1808 bp, are different from the large fragment (3576 bp) expected from an ideal integration of donor DNA. L1 to L6 were the products of alleles in which only a part of the donor DNA, consisting mainly of the drug-resistance cassette, had integrated into the gene (S1 Appendix). Nucleic acid markers (M) (NEB 100 bp DNA ladder, catalog number NO551), from top to bottom, are 1517, 1200, 1000, 900, 800, 700, 600, 500/517, 400, 300, 200, and 100 bp.
Fig 6.
LF5-HA protein is expressed and localized normally in 8 out of 10 PCR-positive clones.
(A) Representative western blots of whole-cell extracts of cells of wild-type strain g1 (WT), the LF5-null mutant lf5-2, and 10 positive clones from the initial PCR screen. The blots were probed with anti-LF5 antibody (top panel) or anti-HA antibody (middle panel). Eight out of 10 PCR-positive clones expressed near normal levels of LF5-HA. ATP synthase β subunit (βF1-ATPase) was probed as a loading control (lower panel). The HA-tagged LF5 migrates slightly slower than non-tagged LF5. In these blots of whole-cell extracts, the anti-LF5 antibody labels a broad, diffuse band in the LF5 region; this band is present in lf5-2, indicating that it is non-specific. (B) Representative immunofluorescence microscopy images of cells of the wild-type strain g1 (WT) and one of the HA-tagged strains. Cells were probed with anti-HA antibody (shown in green in merged images) and anti-acetylated tubulin antibody (shown in magenta in merged images). The HA signal is localized to the flagella with concentration at the proximal end of the flagellar shaft, which is the typical localization pattern for LF5 [14]. Scale bar: 10 μm.
Fig 7.
Seven out of the eight strains that express LF5-HA have an intact LF5 3’ UTR and have normal flagellar length.
(A) Agarose gels showing the PCR results to check the integrity of the 3’ UTR on the eight strains that expressed near normal levels of LF5-HA. W: wild-type g1 genomic DNA as template. Plasmid: plasmid pLF5HA, which contains the LF5-HA gene, as control. Transformants: genomic DNA from the eight HA-tagged strains. Wild-type product (1674 bp) migrates slightly faster than product from the HA-tagged strains (1785 bp). Mating-type-plus-specific gene FUS1 was used as a positive control. M: NEB 1 kb DNA ladder (catalog number NO552) on the upper gel and NEB 100 bp DNA ladder (catalog number NO551) on the lower gel. (B) Flagellar length is normal in the seven strains that express LF5-HA and have a complete LF5 3’ UTR. g1 is the wild-type strain in which the editing experiment was done. When compared with g1, only the null mutant lf5-2 shows significant difference (P<0.05) as marked by *.
Fig 8.
PCR screening of transformants for insertion of the 3’ end of the NAP1L1 mNeonGreen-3xFLAG donor DNA at the RNP2 cut site.
Crude DNA preparations from 96 transformants were used as templates for PCR amplification with primer pair P1/P2 as illustrated in Fig 4. Since P1 is complementary to the tag sequence, no PCR product is expected for the wild-type gene. The expected size for product from a correctly tagged gene is 450 bp. The presence of product of a size other than 450 bp indicates that more complicated editing occurred. Sixteen clones (labeled A8-H10) yielded product of the size expected for a correct edit (green arrows below the bands and on the right side of the gels); the PCR products from these clones were sequenced for further confirmation (S3 Appendix). Eight of the 16 (marked with green SD) had a sequence that was exactly as expected if the donor DNA was used as the repair template, whereas four (marked with green SE) had a sequence that was exactly as expected if the endogenous DSB end was used as the repair template. In the remaining four (red SX1-SX4), indel or point mutations had been introduced. Ten clones subsequently found to express NAP1L1 tagged with mNeonGreen-3xFLAG (see Fig 11) are marked with a green P. M1: NEB 1 kb DNA ladder (catalog number NO552). M2: NEB 100 bp DNA ladder (catalog number NO551). The band sizes from top to bottom for M1 are 10, 8, 6, 5, 4, 3, 2, 1.5, 1, and 0.5 kb. The band sizes from top to bottom for M2 are 1517, 1200, 1000, 900, 800, 700, 600, 500/517, 400, 300, 200, and 100 bp.
Fig 9.
PCR screening of transformants for insertion of the 5’ end of the NAP1L1 mNeonGreen-3xFLAG donor DNA at the RNP1 cut site.
Crude DNA preparations from the same 96 transformants and in the same order as those analyzed in Fig 8 were used as templates for PCR amplification with primer pairs P3/P4 as illustrated in Fig 4. Because P4 is complementary to sequence in the antibiotic-resistance gene, no PCR product is expected for the wild-type gene. The expected product size for correct editing at the RNP1 target site is 542 bp. However, the PCR products are around 1–1.5 kb. A product of 1350 bp is expected if the RNP1 target site was not cut and the donor DNA inserted cleanly at the RNP2 cut site. To explore this possibility, we sequenced product (here marked S1-S10 below the bands) from 10 clones (labeled at the top of the lanes as in Fig 8) that had yielded product of the correct size with primer pair P1/P2. The results (S4 Appendix) confirmed that the RNP1 target site was not cut and the donor DNA had inserted into the RNP2 cut site, but with modification of its 5’ end. M1: NEB 1 kb DNA ladder (catalog number NO552). M2: NEB 100 bp DNA ladder (catalog number NO551). The band sizes from top to bottom for M1 are 10, 8, 6, 5, 4, 3, 2, 1.5, 1, and 0.5 kb. The band sizes from top to bottom for M2 are 1517, 1200, 1000, 900, 800, 700, 600, 500/517, 400, 300, 200, and 100 bp.
Fig 10.
Schematic overview of TIM-based N-terminal tagging of NAP1L1.
Two RNPs were used as described in Fig 4. However, the gene was cut by RNP2 only, resulting in an insertion of the donor DNA at that site. A portion (indicated by a red line) of the original gene remains immediately upstream of the drug-resistance cassette. The sizes of the expected PCR products for specific experiments of Figs 8 and 9 are shown between the primers.
Fig 11.
NAP1L1-mNeonGreen-3xFLAG protein is expressed in 10 transformants.
Whole-cell lysates from the wild-type parental strain (WT) and 10 transformants (labeled as in Fig 8), each of which had a precise insertion at the 3’ end of the donor DNA with the exception of clone F9, were probed with anti-FLAG antibody and anti-F1 beta ATPase antibodies for loading control. All transformants expressed FLAG-tagged protein at the expected mass (66 kDa) for NAP1L1-mNeonGreen-3xFLAG.