Materials and Methods

Efficient and modular CRISPR‐Cas9 vector system forPhyscomitrella patens

2. MATERIALS & METHODS2.1. Protospacer sequence design and ligationFor each editing experiment, entry vectors were linearized with BsaI. The CRISPOR online software ( (Haeussler et al., 2016) was used to design protospacers for each editing experiment using P. patens (Phytozome V11) and S. pyogenes (5′ NGG 3′) as the genome and PAM parameters, respectively. Protospacers were chosen based on high specificity scores and low off‐target frequency. The chosen protospacer for a given gene and its reverse complement were then constructed to have 4 nucleotides added to their 5′ ends such that, when annealed, they create sticky ends compatible with BsaI‐linearized entry vectors (Figure ​(Figure2b).2b). These were synthesized as oligonucleotides (Table S1) and annealed together using a PCR machine (500 pmol of each, 10 µl total volume with PCR machine setting: 98°C for 3 min, 0.1°C/s to oligo Tm, hold 10 min, 0.1°C/s to 25°C). The final product was ligated into BsaI‐linearized entry vector using Instant Sticky‐End Ligation Master Mix (New England Biolabs) following the manufacturer's recommendations.Open in a separate windowFigure 2(a) A plasmid map of the entry vector containing the U6 promoter to drive expression of the sgRNA flanked by Gateway att sites. (b) The DNA sequence of the U6 promoter‐sgRNA junction within the entry clone separated by inverted BsaI sites. Upon digestion of the entry clone with BsaI, unique vector overhangs allow for directional ligation of custom oligonucleotides containing the protospacer sequence. (c) Plasmid maps of the destination vectors pMH‐Cas9‐gate, pMK‐Cas9‐gate, and pZeo‐Cas9‐gate for hygromycin, G418, and Zeocin selection in plants, respectively. (d) Entry vectors are shown that were generated based on the plasmid shown in (a) with modified att sites enabling compatibility with Multisite Gateway reactions2.2. Polymerase III promoter assayTo build a CRISPR/Cas9 vector system for P. patens, we wanted to first assess RNA polymerase III promoter efficiency. The NLS‐4 moss line contains a transgene that codes for nuclear‐localized GFP fused to GUS (NLS‐GFP‐GUS) as described in Bezanilla, Pan, and Quatrano (2003). To target NLS‐GFP‐GUS using a rice U3 promoter, we ligated protospacer oligos into pENTR‐OsU3‐sgRNA (a gift from Devin O’Connor) to create pENTR‐OsU3‐sgRNA‐NGG. To target NLS‐GFP‐GUS using a P. patens U6 promoter, we removed the OsU3 promoter from pENTR‐OsU3‐sgRNA‐NGG using an AscI and SalI digest. We subsequently amplified the PpU6 promoter from wild‐type P. patens Gransden strain using primers with AscI and SalI sites (Table S1), digested with AscI and SalI, and ligated into linearized pENTR‐OsU3‐sgRNA‐NGG using Sticky‐End Ligation Master Mix (New England Biolabs) following the manufacturer's recommendations. These entry vectors were recombined with pH‐Ubi‐Cas9 (Miao et al., 2013) using an LR clonase reaction to create the final expression constructs, pH‐Ubi‐Cas9‐OsU3‐NGG and pH‐Ubi‐Cas9‐PpU6‐NGG.Prior to imaging, we removed the labels from the plates such that the images were acquired by a blinded observer and drew a grid on the bottom of the plates to act as guides for counting. We counted 7‐day‐old plants and recorded the presence or absence of nuclear fluorescence for each plant using a fluorescence stereomicroscope (Leica MZ16FA), equipped with the following filter: excitation 480/40, dichroic 505 long pass, emission 510 long pass.2.3. U6 promoter/sgRNA entry vector constructsTo generate pENTR‐PpU6‐sgRNA‐L1L2, the first Gateway entry vector for the P. patens vector system, we amplified the PpU6 and sgRNA fragments with two separate PCRs. For the PpU6 fragment, we used a forward primer containing an AscI site and a reverse primer containing two inverted BsaI sites at the 5′ ends (Table S1). Similarly, for the sgRNA fragment we used a forward primer containing two inverted BsaI sites and a reverse primer containing a SalI site (Table S1). The two fragments were then ligated using an overlap extension PCR and ligated into pGEM/T‐Easy (Promega). Positive clones were digested with AscI and SalI, and the dropout was subsequently ligated into an AscI‐ and SalI‐linearized pENTR‐PpU6‐sgRNA‐NGG plasmid.To generate the six entry vectors compatible with Multisite Gateway (Invitrogen) for multiplexing experiments, we amplified the sgRNA expression cassette from pENTR‐PpU6‐sgRNA‐L1L2 using primers (Table S1) containing different Multisite Gateway attachment sites (attB) and subsequently recombined with the Multisite Gateway pDONR221 plasmid set (Invitrogen) using a BP clonase reaction following the manufacturer's recommendations.2.4. Cas9/sgRNA destination and expression constructsTo generate the three destination vectors, we purified a fragment containing the Cas9 and Gateway cassette from pH‐Ubi‐Cas9 (Miao et al., 2013) digested with StuI and PmeI. This fragment was ligated into linearized pMH, pMK, and pZeo vectors by blunt‐end ligation to create pMH‐Cas9‐gate, pMK‐Cas9‐gate, and pZeo‐Cas9‐gate. Sequences are available on AddGene ( All of the Cas9/sgRNA expression vectors used in this study were generated using Gateway (for one sgRNA) or Multisite Gateway (for multiple sgRNAs) to recombine the entry vectors and destination vectors just described (Invitrogen).2.5. Homology‐directed repair constructsTo generate pENTR‐R4R3‐stop (the “stop cassette”), we amplified 360 bp of the plasmid pBluescriptSK(+), including the multiple cloning site, with attB4r and attB3r Gateway primers (Table S1). The forward primer also contained three stop codons in each frame. We subsequently cloned the PCR fragment into pDONR221‐P4rP3r (Invitrogen) using a BP clonase reaction following the manufacturer's recommendations.To generate the mEGFP and mRuby2 tagging entry vectors, we amplified mEGFP (Vidali et al., 2009) and mRuby2 (Lam et al., 2012) coding sequences using forward and reverse primers (Table S1) that contained attB4r and attB3r sites, respectively. The forward primers (Table S1) for both mEGFP and mRuby2 also contained a BamHI site. These PCR products were subsequently cloned into pDONR221‐P4rP3r (Invitrogen) using a BP clonase reaction to create pENTR‐R4R3‐mEGFP‐C and pENTR‐R4R3‐mRuby‐C. mEGFP‐pGEM, a vector described by Vidali et al. (2009), contains BamHI and BglII sites flanking the mEGFP coding sequence. We generated mRuby2‐pGEM, a vector constructed in the same way as mEGFP‐pGEM (Vidali et al., 2009). We digested these vectors with BamHI and BglII and the resulting fragments were ligated into BamHI‐digested pENTR‐R4R3‐mEGFP‐C and pENTR‐R4R3‐mRuby‐C to create pENTR‐R4R3‐2XmEGFP‐C and pENTR‐R4R3‐2XmRuby‐C, respectively. We linearized the resulting 2X constructs with BamHI and ligated the BamHI/BglII fragments from mEGFP‐pGEM and mRuby2‐pGEM to create the 3X constructs, pENTR‐R4R3‐3XmEGFP‐C and pENTR‐R4R3‐3XmRuby‐C, respectively. To create the mEGFP and mRuby2 N‐terminal constructs, the process above was repeated except the attB3r primers (Table S1) did not contain stop codons.2.6. DNA donor templatesWe used the three‐fragment Multisite Gateway cloning system (Invitrogen) to generate the final homology‐directed repair constructs. For Pp3c22_1100, we amplified 2 fragments of approximately 800 bp upstream and downstream of the Pp3c22_1100 stop codon. For Pp3c16_8300, we amplified 2 fragments of approximately 800 bp upstream and downstream of the expected start codon. For both genes, we cloned the upstream fragments into pDONR221‐P1P4 and the downstream fragments into pDONR221‐P3P2 using a BP clonase reaction. To create the final homology‐directed repair DNA donor plasmids, the resulting pENTR vectors from the BP reaction underwent an LR clonase reaction with the second‐position tagging vector (pENTR‐R4R3‐mEGFP‐C for Pp3c22_1100 and pENTR‐R4R3‐mRuby‐N for Pp3c16_8300) and the destination vector, pGEM‐gate (Vidali et al., 2009).To restore efficient homology‐directed repair while tagging Pp3c16_8300, we performed site‐directed mutagenesis on the entry vector containing the 3′ homology fragment (pENTR‐L3L2‐3c16‐3Arm) to create pENTR‐L3L2‐3c16‐3Arm‐mut by altering the third nucleotide from the PAM (5’ NGG 3’) within the protospacer. We repeated the LR reaction using this third‐position entry vector to generate pGEM‐3c16‐mRuby‐HDR‐mut.2.7. Moss tissue culture and transformationWe propagated moss tissue weekly by light homogenization and subsequently plated on 10‐cm petri dishes to maintain the protonemal growth stage. Dishes contained 25 ml PpNH4 growth medium (103 mM MgSO4, 1.86 mM KH2PO4, 3.3 mM Ca(NO3)2, 2.72 mM (NH4)2‐tartrate, 45 µM FeSO4, 9.93 µM H3BO3, 220 nM CuSO4, 1.966 µM MnCl2, 231 nM CoCl2, 191 nM ZnSO4, 169 nM KI, and 103 nM Na2MoO4) with 0.7% agar covered with cellophane disks. Plants were grown in daily cycles of 16‐hr light/8‐hr dark with 85 µmol photons m−2 s−1. For transformation, protoplasts were transformed with 30 µg of each DNA construct using PEG‐mediated transformation protocol (as described in Augustine, Pattavina, Tuzel, Vidali, and Bezanilla (2011)). Plants were allowed to regenerate on plant regeneration media (PRMB) for 4 days (described in (Wu & Bezanilla, 2014)) atop of cellophane disks. Depending upon the selection cassette present on the expression vector, we subsequently moved plants to PpNH4 growth media containing either hygromycin (15 µg/ml), G418 (20 µg/ml), or Zeocin (50 µg/ml). Plants were not selected for homology‐directed repair DNA donor vectors. After 7 days on selection, we moved plants to PpNH4 media without antibiotics for maximal growth, except for plants that were transformed to compare the efficiency of the rice U3 and moss U6 promoters—those plants remained on selection and were imaged after 2 weeks. All other plants were allowed to grow for 2–3 weeks until tufts were 0.5–1 cm in diameter for DNA extraction.2.8. DNA extraction and genotypingWe extracted DNA from plants that were 3–4 weeks old (0.5–1 cm in diameter) using the protocol as described in Augustine et al. (2011). For editing experiments, we used PCR primers (Table S1) surrounding the expected Cas9 cleavage site (~300–400 bp on each side). For homology‐directed repair experiments, we used PCR primers outside of the homology region to avoid amplification of residual DNA donor template. To perform PCR, we used Q5 polymerase (New England Biolabs) using the manufacturer's recommendations.2.9. T7 endonuclease assayTo detect CRISPR edits, we amplified a 0.5 to 1 kb genomic region flanking the potential CRISPR editing site by PCR. The PCR product from each candidate plant was mixed with a roughly equivalent amount of wild‐type PCR product of the same locus. The mixture was denatured and annealed in a PCR machine and subsequently digested with 1 µl of T7 endonuclease (New England Biolabs) following the manufacturer's recommendations. We examined the digest on a 1% agarose gel.2.10. Laser scanning confocal microscopyFor confocal imaging, moss protonemal tissue was grown in microfluidic imaging device as described in Bascom, Wu, Nelson, Oakey, and Bezanilla (2016). The imaging device is filled with half‐strength Hoagland's medium (4 mM KNO3, 2 mM KH2PO4, 1 mM Ca(NO3)2, 89 μM Fe citrate, 300 μM MgSO4, 9.93 μM H3BO3, 220 nM CuSO4, 1.966 μM MnCl2, 231 nM CoCl2, 191 nM ZnSO4, 169 nM KI, 103 nM Na2MoO4) and kept at room temperature on the benchtop with overhead fluorescent lights. Images were acquired on a Nikon A1R confocal microscope system with a 1.49 NA 60x oil immersion objective (Nikon Apo TIRF 60x Oil DIC N2) at room temperature. 488 nm laser illumination was used for mEGFP excitation. Emission filter was 525/50 nm for mEGFP. Image acquisition was controlled by NIS‐Elements software (Nikon). Images were processed using NIS‐Elements software (Nikon): advanced denoising with regression and other parameters set to default.

Article TitleEfficient and modular CRISPR‐Cas9 vector system forPhyscomitrella patens


CRISPR‐Cas9 has been shown to be a valuable tool in recent years, allowing researchers to precisely edit the genome using an RNA‐guided nuclease to initiate double‐strand breaks. Until recently, classical RAD51‐mediated homologous recombination has been a powerful tool for gene targeting in the mossPhyscomitrella patens. However, CRISPR‐Cas9‐mediated genome editing inP. patenswas shown to be more efficient than traditional homologous recombination (Plant Biotechnology Journal, 15, 2017, 122). CRISPR‐Cas9 provides the opportunity to efficiently edit the genome at multiple loci as well as integrate sequences at precise locations in the genome using a simple transient transformation. To fully take advantage of CRISPR‐Cas9 genome editing inP. patens, here we describe the generation and use of a flexible and modular CRISPR‐Cas9 vector system. Without the need for gene synthesis, this vector system enables editing of up to 12 loci simultaneously. Using this system, we generated multiple lines that had null alleles at four distant loci. We also found that targeting multiple sites within a single locus can produce larger deletions, but the success of this depends on individual protospacers. To take advantage of homology‐directed repair, we developed modular vectors to rapidly generate DNA donor plasmids to efficiently introduce DNA sequences encoding for fluorescent proteins at the 5′ and 3′ ends of gene coding regions. With regard to homology‐directed repair experiments, we found that if the protospacer sequence remains on the DNA donor plasmid, then Cas9 cleaves the plasmid target as well as the genomic target. This can reduce the efficiency of introducing sequences into the genome. Furthermore, to ensure the generation of a null allele near the Cas9 cleavage site, we generated a homology plasmid harboring a “stop codon cassette” with downstream near‐effortless genotyping.

Login or Signup to leave a comment
Find your community. Ask questions. Science is better when we troubleshoot together.
Find your community. Ask questions. Science is better when we troubleshoot together.

Have a question?

Contact or check out our support page.