Nucleic Acids Research, 2020 1

doi: 10.1093/nar/gkz1225

DNA targeting by Clostridium cellulolyticum

CRISPR–Cas9 Type II-C system
Iana Fedorova 1,*,† , Anatolii Arseniev2,† , Polina Selkova1 , Georgii Pobegalov2 ,
Ignatiy Goryanin1 , Aleksandra Vasileva1 , Olga Musharova1 , Marina Abramova2 ,

Downloaded from https://academic.oup.com/nar/advance-article-abstract/doi/10.1093/nar/gkz1225/5707196 by guest on 19 January 2020

Maksim Kazalov2 , Tatyana Zyubko1 , Tatyana Artamonova2 , Daria Artamonova1 ,
Sergey Shmakov1,3 , Mikhail Khodorkovskii2 and Konstantin Severinov 1,4,*
1
Skolkovo Institute of Science and Technology, Center of life sciences, Skolkovo, Russia, 2 Peter the Great St.
Petersburg Polytechnic University, Saint Petersburg, Russia, 3 National Center for Biotechnology Information, National
Library of Medicine, National Institutes of Health, USA and 4 Center for Precision Genome Editing and Genetic
Technologies for Biomedicine, Institute of Gene Biology, Russian Academy of Sciences, Moscow, Russia

Received July 20, 2019; Revised December 16, 2019; Editorial Decision December 17, 2019; Accepted December 20, 2019

ABSTRACT CRISPR arrays during the infection. The CRISPR array is

transcribed into a pre-crRNA, which is processed further to
Type II CRISPR–Cas9 RNA-guided nucleases are short mature crRNAs containing a single spacer and flank-
widely used for genome engineering. Type II-A Sp- ing repeat sequences. Complementary pairing between cr-
Cas9 protein from Streptococcus pyogenes is the RNA spacer segment and the invader genome allows Cas
most investigated and highly used enzyme of its nucleases to specifically recognize foreign targets and de-
class. Nevertheless, it has some drawbacks, includ- grade them, thus preventing the spread of the infection.
ing a relatively big size, imperfect specificity and re- The crRNAs with investigator defined spacer sequences
striction to DNA targets flanked by an NGG PAM se- allow one to guide Cas nucleases to virtually any desirable
quence. Cas9 orthologs from other bacterial species target. Because of their relative simplicity, single-subunit
may provide a rich and largely untapped source of Cas nucleases of Type II CRISPR–Cas systems form the
biochemical diversity, which can help to overcome basis of multiple genome editing applications. Since 2013
Type II CRISPR-based instruments are used for genome
the limitations of SpCas9. Here, we characterize Cc-
modification and transcription regulation in eukaryotic, in-
Cas9, a Type II-C CRISPR nuclease from Clostridium cluding human, cells (1). Alongside with eukaryotic genome
cellulolyticum H10. We show that CcCas9 is an active editing, there is a large demand for genome engineering
endonuclease of comparatively small size that rec- of microorganisms useful in biotechnology and several ef-
ognizes a novel two-nucleotide PAM sequence. The ficient CRISPR-based methods of bacterial genome edit-
CcCas9 can potentially broaden the existing scope of ing have been developed (2–4). Most of these genome edit-
biotechnological applications of Cas9 nucleases and ing approaches rely on the use of the SpCas9 protein, the
may be particularly advantageous for genome editing most investigated to date effector nuclease from Streptococ-
of C. cellulolyticum H10, a bacterium considered to cus pyogenes Type II-A CRISPR–Cas system (5). Despite
be a promising biofuel producer. high DNA cleavage efficiency, SpCas9 has several limita-
tions due to its large size, a strict requirement for an NGG
PAM (protospacer adjacent motif essential for target DNA
INTRODUCTION
recognition) and imperfect specificity.
CRISPR–Cas systems are bacterial and archaeal immune Bioinformatic searches for Cas9 orthologs and their sub-
systems that protect their hosts from invaders such as plas- sequent biochemical characterization reveal nucleases with
mids or bacteriophages. The immune mechanism is based different properties, which can broaden Cas9 proteins appli-
on the function of Cas ribonucleoprotein effector com- cation. Thus, SaCas9 from Staphylococcus aureus and Cj-
plexes composed of Cas nucleases and CRISPR RNAs (cr- Cas9 from Campylobacter jejuni, two small size Cas9 or-
RNAs). crRNAs are encoded in CRISPR arrays consist- thologs with PAM requirements 5 - NNGRRT-3 and 5 -
ing of repeats and intervening unique spacers. Some spac- NNNNRYAC-3 , respectively, were shown to be active in
ers are derived from invader’s DNA and are introduced into human cells (6,7). In 2014, Fonfara et al. using bioinfor-

* To
whom correspondence should be addressed. Tel: +79062713200; Email: [email protected]
Correspondence may also be addressed to Konstantin Severinov. Tel: +79854570284; Email: [email protected]
†
The authors wish it to be known that, in their opinion, the first two authors should be regarded as Joint First Authors.


C The Author(s) 2020. Published by Oxford University Press on behalf of Nucleic Acids Research.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which
permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
2 Nucleic Acids Research, 2020

matics approaches detected a Type II-C system CRISPR– coli DH5alpha strain and plated to media supplemented
Cas in Clostridium cellulolyticum genome but no functional with 100 ␮g/ml ampicillin. The plates were incubated
characterization of this system was performed (8). The at 37◦ C. Eighteen hours after transformation >50 000
mesophilic cellulolytic bacterium C. cellulolyticum is con- colonies were washed off the plates, and the plasmid library
sidered to be a promising biofuel producer since it can was extracted by Qiagen Plasmid Maxi kit (Qiagen 12162).
directly convert plant biomass to lactate, acetate, ethanol HTS analysis of the library showed representation of 15716
and hydrogen (9). Fast and efficient approaches of C. cel- PAM variants. The library plasmid map is presented in the
lulolyticum genome engineering will be required to im- Supplementary Table S1. Competent E. coli Star cells car-

Downloaded from https://academic.oup.com/nar/advance-article-abstract/doi/10.1093/nar/gkz1225/5707196 by guest on 19 January 2020

prove the fermentation properties of this microorganism. To rying pACYC184 CcCas9 locus or an empty pACYC184
date, several CRISPR–Cas-based strategies were applied to vector were transformed with 7N PAM plasmid libraries
change the C. cellulolyticum genome, all of them relying on and plated to 100 ␮g/ml ampicillin and 25 ␮g/ml chloram-
SpCas9 due to the lack of any information about the host phenicol containing agar plates. After 16 h, cells were har-
CRISPR–Cas system (PAM requirements, guide crRNAs vested and DNA was extracted using Qiagen Plasmid Maxi
sequences, protospacer length etc.) (10–12). Studying of C. kit (Qiagen 12162). PAM-containing sequences were PCR
cellulolyticum Type II-C CRISPR–Cas system and, in par- amplified using M13 f and M13 r primers and sequenced
ticular, its effector Cas9 nuclease, could facilitate genome using Illumina platform with pair-end 150 cycles (75 + 75).
modification of this bacterium and provide an additional
small-size Cas9 effector for biotechnology or biomedicine.
Here, we demonstrate that C. cellulolyticum H10 CcCas9 Bacterial RNA sequencing
protein is an active RNA-guided nuclease, which efficiently E. coli DH5alpha carrying pACYC184 CcCas9 locus were
introduces double-stranded breaks in DNA targets flanked grown 16 h at 37◦ C in LB (Luria Bertani) medium sup-
by two-nucleotide 5 -NNNNGNA-3 PAM. To facilitate plemented with 25 ␮g/ml chloramphenicol. Bacteria were
further application of CcCas9 in biotechnology, we deter- resuspended in TRIzol (ThermoFisher, 15596026). Total
mined the main features of this CRISPR–Cas system, such RNA was purified using Direct-Zol RNA kit (Zymo re-
as crRNA and tracrRNA sequences, the range of tempera- search, R2051). RNA was DNase I (Zymo research) treated
tures required for in vitro activity and created a nickase ver- and 3 dephosphorylated with T4 PNK (NEB, M0201).
sion of CcCas9, which could be suitable for C. cellulolyticum Ribo-Zero rRNA Removal Kit (Gram-Negative Bacteria)
H10 genome editing by a single-nick-assisted homologous kit (Illumina, 15066012) was used to remove ribosomal
recombination (11). RNA. HTS samples were prepared using NEBNext Mul-
tiplex Small RNA Library Prep Set for Illumina (NEB,
MATERIALS AND METHODS E7300). The library was sequenced using Illumina platform
Plasmids cloning with pair-end 150 cycles (75 + 75).

The entire predicted CRISPR–Cas Type II-C system lo-

cus of C. cellulolyticum including flanking regions (100 nt RNA sequencing analysis
upstream of putative tracrRNA coding sequence and 180 HTS results of RNA sequencing were aligned to the refer-
nt downstream of the last DR) was PCR amplified with ence plasmid pACYC184 CcCas9 locus using BWA aligner
primers locus F and locus R using C. cellulolyticum H10 (13). Determined coordinates of 5 and 3 RNA ends were
genomic DNA (DSMZ 5812) as a template. The result- used to reconstruct the full-length RNA sequences. The re-
ing fragment was inserted into XbaI and HindIII digested sulting fragments were analyzed using Geneious 11.1.2. Fil-
pACYC184 vector using NEBuilder HiFi DNA Assem- tered 40–130 nt-length sequences were used to generate the
bly Cloning Kit (NEB, E5520). To obtain pET21a CcCas9 alignment.
plasmid, CcCas9 coding sequence was PCR amplified
with CcCas9 F and CcCas9 R primers using C. cellu-
lolyticum H10 genomic DNA as a template. The resulting In vitro DNA cleavage assays
fragment was inserted into XhoI and NheI digested pET21a
vector by NEBuilder HiFi DNA Assembly Cloning Kit DNA cleavage reactions were performed using the recom-
(NEB, E5520). The vectors maps are presented in the Sup- binant CcCas9 protein and linear dsDNA targets. The reac-
plementary Table S1. tion conditions were: 1× CutSmart (NEB, B7204) buffer, 1
mM DTT, 30 nM DNA, 400 nM CcCas9, 2 ␮M crRNA, 2
␮M tracrRNA. Samples were incubated at an appropriate
Plasmid transformation interference screening
temperature for 20 min (unless otherwise stated). Further,
Randomized 7N plasmid libraries carried a protospacer se- 4× loading dye containing 10 mM Tris–HCl, pH 7.8, 40%
quence flanked by seven randomized nucleotides (Supple- glycerol, 40 mM EDTA, 0.01% bromphenol blue, 0.01%
mentary Table S1). To create the library the ssDNA oligo xylene cyanol was added to stop the reaction. Reaction
Library f containing randomized nucleotides was double- products were analyzed by electrophoresis in 1.5% agarose
stranded through single stage PCR with Library r primer gels or, where indicated, in 1× TBE polyacrylamide gels.
(Evrogen). This fragment was assembled with PUC19 frag- Pre-staining with ethidium bromide or post-staining with
ment synthesized through PCR using primers PUC19 F SYBR gold stain (ThermoFisher, 11494) was used for visu-
and PUC19 R by NEBuilder HiFi DNA Assembly Cloning alization of bands on agarose or polyacrylamide gels, cor-
Kit (NEB, E5520). The mix was transformed to Escherichia respondingly.
 Nucleic Acids Research, 2020 3

Downloaded from https://academic.oup.com/nar/advance-article-abstract/doi/10.1093/nar/gkz1225/5707196 by guest on 19 January 2020

Figure 1. Organization of Clostridium cellulolyticum H10 CRISPR–Cas Type II-C locus. (A) A scheme of the C. cellulolyticum H10 CRISPR–Cas locus.
DRs (direct repeats) are shown as black rectangles, spacers are indicated by rectangles of different colors. The tracrRNA coding sequence is shown as a
brown rectangle. The cas genes are labeled. The direction of transcription is indicated with black arrows. Mapping of small RNAs reads revealed by RNA-
seq is shown at the top of CRISPR array in blue. A sequence of typical mature crRNA is expanded below with the DR part shown in bold typeface. (B)
Domain organization of the CcCas9 protein. (C) In silico co-folding of C. cellulolyticum H10 CRISPR–Cas Type II-C system DR and putative tracrRNA.
The DR sequence is colored in red, the tracrRNA sequence is colored in green. The cleavage sites introduced during crRNA maturation are indicated with
red arrows. Co-folding was performed using Geneious software, free energy of structure shown is −80.50 kcal/mol.

All in vitro DNA cleavage reactions were performed at Computational sequence analysis
45◦ C unless otherwise stated. For testing the activity of Cc-
For PAM screens results analysis, Illumina reads were fil-
Cas9 at different temperatures a mix of CcCas9 protein
tered by requiring an average Phred quality (Q score) of
with in vitro transcribed crRNA–tracrRNA in the cleavage
at least 20. Resulting reads were mapped against the cor-
buffer, and the DNA substrates, also in the cleavage buffer,
responding reference sequence using BWA (13). All un-
were first incubated separately at the chosen temperature for
mapped reads were discarded from the analysis. The degen-
10 min, combined, and incubated for additional 10 min at
erate 7-nucleotide region was extracted from the sequences.
same temperature.
For interference PAM screens analysis, depleted PAM se-
For in vitro PAM screens, 100 nM linear DNA 7N PAM
quences were determined by comparing the number of each
library was incubated with 400 nM CcCas9, 1 ␮M cr-
PAM counts for CRISPR CcCas9 sample and control. The
RNA and 1 ␮M tracrRNA. Reactions without crRNA were
representation of unique PAM in both samples, as well as
used as negative controls. The reactions were performed at
PAM representation of initial 7N library was >15 000 PAM
45◦ C for 20 min. Reaction products were separated by elec-
variants. WebLogo was used to generate a logo based on 887
trophoresis in agarose gels. Uncleaved DNA fragments were
of statistically significantly (one-sided Pearson chi-square
extracted from the gel using Zymo Clean Gel Recovery kit
test with a P-value < 10−12 ) depleted PAM sequences (listed
(Zymo research, D4007). HTS libraries were prepared us-
in Supplementary File S2). In case of in vitro PAM de-
ing Ultra II DNA library prep kit (NEB, E7646). Samples
termination screens 16 364 and 16 363 unique PAM se-
were sequenced using MiniSeq Illumina with single-end 150
quences were found, respectively, for the depleted and con-
cycles. All RNAs used in this study are listed in Supplemen-
trol samples. Depletion values of PAM sequence positions
tary Table S2.
4 Nucleic Acids Research, 2020

Downloaded from https://academic.oup.com/nar/advance-article-abstract/doi/10.1093/nar/gkz1225/5707196 by guest on 19 January 2020

Figure 2. Determination of CcCas9 PAM sequence using plasmid transformation interference screening. (A) A scheme of the bacterial interference screen
experiment. Escherichia coli cells carrying the CcCas9 locus were transformed by PUC19-based library carrying the protospacer sequence flanked by seven
randomized nucleotides and plated on ampicillin containing plates. The presence of an interference-proficient PAM decreases the frequency of plasmids
with this PAM among ampicillin-resistant colonies. Comparison of PAM representation in CcCas9 locus carrying cells and in control cells without the
CcCas9 locus reveals depleted PAM sequences and allows one to deduce the PAM consensus. (B) Clostridium cellulolyticum CRISPR–Cas Type II-C system
PAM sequence logo determined by plasmid transformation interference screening.

were counted according to (14). The frequencies of each tegrity of recombinant protein was confirmed by mass spec-
PAM variants in depleted and control samples were pro- trometry.
cessed by R script. The frequencies of PAM variants were
also used for PAM wheel construction.
RESULTS AND DISCUSSION
Recombinant protein purification Clostridium cellulolyticum H10 CRISPR–Cas II-C system:
locus organization
For recombinant CcCas9 purification competent E. coli
Rosetta cells were transformed with pET21a CcCas9 plas- The C. cellulolyticum H10 type II-C CRISPR–Cas locus
mid and grown till OD600 = 0.6 in 500 ml LB media sup- was bioinformatically found by Fonfara et al. in 2014 but
plemented with 100 ␮g/ml ampicillin. The target protein up to date there is no information about the activity of
synthesis was induced by the addition of 1 mM IPTG. Af- this system. The CRISPRFinder tool (https://crispr.i2bc.
ter 18 h of growth at 22◦ C, cells were centrifuged at 4000g, paris-saclay.fr/Server/) revealed an array composed of nine
the pellet was resuspended in lysis buffer containing 50 36-bp DRs (direct repeats) interspaced by 31-bp spacers
mM Tris–HCl pH 8.0 (4◦ C), 500 mM NaCl, 1 mM ␤- in the proximity of the cas genes operon (Figure 1A). A
mercaptoethanol and 10 mM imidazole supplemented with Blast search using spacer sequences as queries revealed no
1 mg/ml lysozyme (Sigma) and cells were lysed by sonica- matches to sequences from publicly accessible databases.
tion. The cell lysate was centrifuged at 16 000g (4◦ C) and The C. cellulolyticum H10 cas genes comprise the CcCas9
filtered through 0.45 ␮m filters. The lysate was applied to 1 effector nuclease gene and the adaptation module com-
ml HisTrap HP column (GE Healthcare) and CcCas9 was posed of cas1 and cas2 genes. Being a II-C type Cas nucle-
eluted by imidazole gradient in the same buffer without ase, CcCas9 has a relatively small size (1021 amino acids or
lysozyme. After affinity chromatography, fractions contain- 118 kDa) compared to the widely used SpCas9 (1368 amino
ing CcCas9 were applied on a Superose 6 Increase 10/300 acids/158 kDa). Alignment of the CcCas9 amino acid se-
GL (GE Healthcare) column equilibrated with a buffer con- quence with the previously characterized small-size Type
taining 50 mM Tris–HCl pH 8.0 (4◦ C), 500 mM NaCl, II-A SaCas9 protein from S. aureus shows the presence of
1 mM DTT. Fractions containing CcCas9 monomer were all domains necessary for nuclease activity (Figure 1B, Sup-
pooled and concentrated using 30 kDa Amicon Ultra-4 cen- plementary Figure S1). Upstream of cas genes, we identi-
trifugal unit (Merc Millipore, UFC803008). Glycerol was fied a putative tracrRNA-encoding sequence with an anti-
added to final concentration of 10% and samples were flash- repeat partially complementary to DRs. In silico co-folding
frozen in liquid nitrogen and stored at −80◦ C. Purity of of part of DR with the putative tracrRNA predicts a stable
CcCas9 was assessed by denaturing 8% PAGE and the in- secondary structure (Figure 1C).
 Nucleic Acids Research, 2020 5

Downloaded from https://academic.oup.com/nar/advance-article-abstract/doi/10.1093/nar/gkz1225/5707196 by guest on 19 January 2020

Figure 3. In vitro cleavage of DNA targets by CcCas9. (A) Scheme of the DNA library used for in vitro PAM screening experiment. Cleavage of 7N DNA
library with CcCas9 generates DNA products shortened by 50 bp. The location of the cleavage site is shown with red arrows. (B) Analysis of depletion
of PAM library sequences after in vitro cleavage. (C) Single-nucleotide substitutions in the 5th position of PAM prevent DNA cleavage by CcCas9. An
agarose gel showing the results of electrophoretic separation of cleavage products of targets with PAM sequences shown at the top is presented. The +/–
signs signify, correspondingly, whether cleavage was or was not observed. Bands corresponding to cleaved and uncleaved DNA fragments are indicated.
(D) Wheel representation of in vitro PAM screen results for fifth, sixth and seventh nucleotide positions of PAM. Nucleotide positions from the inner to
outer circle match the PAM positions moving away from the protospacer. For a given sequence, the area of the sector in the PAM wheel is proportional
to the relative depletion in the library. (E) In vitro cleavage of different 20-bp target sites on a linear DNA fragment by CcCas9. The PAM sequences
corresponding to each target are shown in the table. The +/– signs signify, correspondingly, whether cleavage was or was not observed. The conserved
G at the fifth position is indicated by green color. Below, a gel showing results of in vitro cleavage of targets with indicated PAMs is presented. White
arrows indicate positions of poorly visible bands. Above, a scheme showing the relative positions of the targeted DNA sites on the linear DNA fragment
is presented.

Figure 4. A scheme of the CcCas9 DNA–cleavage complex. DNA is shown in blue, crRNA in green and tracrRNA in black.
6 Nucleic Acids Research, 2020

Downloaded from https://academic.oup.com/nar/advance-article-abstract/doi/10.1093/nar/gkz1225/5707196 by guest on 19 January 2020

Figure 5. Activity of CcCas9 at different temperatures. (A) CcCas9 was incubated with tracrRNA, crRNA and a 2.7 kb plasmid DNA (above) or a 921 bp
linear DNA fragment (below) containing a target sequence at indicated temperatures for 10 min. Products were separated by agarose gel electrophoresis.
(B) CcCas9 was incubated with tracrRNA, crRNA and plasmid or linear DNA as in panel A. Cleavage efficiency (in per cent) was calculated as a ratio of
intensity of staining of cleaved bands to the combined intensity of cleaved and uncleaved bands. Mean values and standard deviations obtained from three
independent experiments are shown.

The entire CRISPR–Cas locus of C. cellulolyticum H10 tion initiation within each repeat has been shown experi-
with adjacent non-coding sequences likely containing pro- mentally (16). Each C. cellulolyticum H10 crRNA contains
moters was cloned into E. coli pACYC184 plasmid vector 23–26 nt of spacer sequence and 24–28 nt of DR. The tracr-
for heterologous expression. Although E. coli cells carry a RNA coding sequence is also expressed, generating variably
CRISPR–Cas system of their own, it belongs to a different sized, 70–107 nt, products. In the natural host, the length of
class (type I-E), relies on different kinds of crRNAs, and mature crRNAs and tracrRNA could be slightly different
is inactive at least at laboratory conditions (15). Thus, no from those obtained during heterologous expression in E.
influence of resident CRISPR–Cas on the function of C. coli.
cellulolyticum H10 CRISPR–Cas is expected. To determine
the polarity of C. cellulolyticum H10 CRISPR array tran-
scription and confirm the tracrRNA sequence, small RNAs Determination of CcCas9 PAM by DNA interference screen-
present in E. coli heterologously expressing C. cellulolyticum ing
H10 CRISPR–Cas locus were sequenced. We found that the Given robust expression of C. cellulolyticum crRNAs in E.
CRISPR array is actively transcribed in the orientation op- coli, we performed a bacterial interference screen to de-
posite to the cas genes transcription and mature crRNAs termine the CcCas9 protospacer adjacent motif (PAM) se-
corresponding to every spacer in the array could be detected quence (Figure 2A). Based on the knowledge about organi-
(Figure 1A). This could be due to efficient processing of zation of known Cas9-guide RNAs–target DNA complexes
pre-crRNA or, alternatively, due to transcription from in- and the direction of C. cellulolyticum CRISPR array tran-
ternal promoters embedded into the repeat sequence, as has scription, we designed a plasmid-based PAM library carry-
been observed in some Type II-C systems (16). Indeed, we ing a 30-bp protospacer sequence matching the first spacer
noted that the terminal nine nucleotides of C. cellulolyticum in the C. cellulolyticum CRISPR array flanked at one side
H10 DRs have a sequence similar to bacterial extended −10 with seven randomized nucleotides (Figure 2A). E. coli cells
promoter consensus element, as is also the case for Neisse- carrying a compatible plasmid with the CcCas9 locus or an
ria meningitidis CRISPR–Cas II-C system, where transcrip- empty vector were transformed with the library and plated
 Nucleic Acids Research, 2020 7

Downloaded from https://academic.oup.com/nar/advance-article-abstract/doi/10.1093/nar/gkz1225/5707196 by guest on 19 January 2020

Figure 6. The influence of temperature on Clostridium cellulolyticum H10 CRISPR–Cas II-C locus interference. A plasmid library composed of five
members carrying protospacer matching the first spacer in the CRISPR array and flanked by 5 -ACAGGTA-3 (PAM 1), 5 -CGGTGTA-3 (PAM 2),
5 -TGAAGAA-3 (PAM 3), 5 -ATTGGAA-3 (PAM 4), and 5 -TTCATAT-3 (no PAM) sequence was transformed in E. coli cells carrying a plasmid with
the C. cellulolyticum H10 CRISPR–Cas II-C locus or a control plasmid. Cells were plated on LB media supplemented with ampicillin and chloramphenicol
and grown for 18 h at 37◦ C (left panel) or 22◦ C (right panel). The plasmid DNA was extracted from grown colonies and HTS was used to estimate the
representation of each library member. The pie charts showing PAM representation in colonies formed are shown below. Each colored sector represents a
fraction of corresponding PAM sequence.

on a medium that only allowed the growth of cells carrying PAM at the 3 -flank, in agreement with results obtained dur-
both plasmids. High-throughput sequencing of the targeted ing in vivo interference screening (Figure 3B), although A at
protospacer region amplified from plasmids extracted from the 7th position was less conserved comparing to G at the
pooled transformant colonies revealed depletion of 887 out 5th position.
of 16 384 library members in cells carrying the CcCas9 locus To validate CcCas9 PAM sequence preferences, single-
compared to control cells (Supplementary File S2). Most of nucleotide substitutions in the deduced consensus PAM se-
depleted variants had a 5 -NNNNGNA-3 sequence, indi- quence were introduced and individually tested for cleav-
cating that CcCas9 prefers purines at positions 5 and 7 of age efficiency (Figure 3C). The results confirmed the im-
the non-target DNA strand downstream of the protospacer portance of a G at the fifth position and a less strict prefer-
(Figure 2B). ence for an A at the seventh position (Figure 3C). To further
investigate CcCas9 PAM sequence preferences, in particu-
In vitro cleavage of DNA by CcCas9 lar, to identify individual sequences representing functional
PAMs and the relative activity of each sequence, we used
Based on the interference screening experiments results we the PAM wheel approach developed by Leenay et al. (18)
proceeded to reconstitute CcCas9 DNA cleavage in vitro. A for results visualization. The PAM wheel confirmed the 5 -
recombinant CcCas9 was purified (Supplementary Figure NNNNGNA-3 motif with a moderate preference for an A
S2) and tested for its ability to cleave linear DNA PAM li- at the seventh position but also revealed a slight bias for an
braries containing a target site flanked with seven random- A in addition to G in the fifth position (Figure 3D).
ized nucleotides (Figure 3A). Since C. cellulolyticum H10 We next tested CcCas9 DNA cleavage activity on dif-
was isolated from decayed grass in a compost pile (17), we ferent targets flanked by the 5 -NNNNGNA-3 consensus
first performed DNA cleavage reactions at 33◦ C, the re- PAM as well as 5 -NNNNGNN-3 PAM sequences (Fig-
ported optimal growth temperature (17), but did not detect ure 3E). Several 20-bp target sites with CcCas9 PAM in
any cleavage. The change of reaction temperature to 45◦ C a 1592-bp PCR fragment of human grin2b gene were se-
led to observable library DNA cleavage. Uncleaved DNA lected, the corresponding crRNAs synthesized, and in vitro
fragments as well as a negative control (original DNA PAM cleavage reactions were performed with recombinant Cc-
library incubated with DNA cleavage reaction components Cas9 charged with these crRNAs. Control crRNAs recog-
in the absence of crRNA) were sequenced using the Illumina nizing sequences flanked by PAMs with no G at the fifth po-
platform. Comparison of PAM variants representation in sition were also tested. As can be seen from Figure 3E, Cc-
experimental and control samples allowed us to determine Cas9 did not recognize targets flanked by control sequences
PAM sequences depleted in the presence of the CcCas9 ef- with substitutions of G at the fifth position. On the other
fector complex. The analysis revealed that recombinant Cc- hand, the CcCas9 nuclease recognized and cleaved not only
Cas9 in complex with in vitro synthesized tracrRNA and targets with 5 -NNNNGNA-3 consensus PAM, but also
crRNA was able to cleave DNA targets with ‘NNNNGNA’ targets flanked by 5 -NNNNGNN-3 sequences, confirm-
8 Nucleic Acids Research, 2020

ing that 5 -NNNNGNN-3 PAMs are functional. Similar phenicol and grown for 18 h at either 22 or 37◦ C. Plas-
results were obtained when in in vitro DNA cleavage by mid DNA was purified from colonies formed at each tem-
CcCas9 was performed using a supercoiled plasmid carry- perature and HTS of PAM-containing regions was per-
ing the cloned grin2b gene fragment (Supplementary Figure formed to determine the changes in representation of li-
S3). The cleavage efficiency of CcCas9 on different DNA brary members (Figure 6, Supplementary Table S3). Anal-
targets varied significantly, which is likely a combination ysis of HTS results showed the decrease in the frequency of
of contributions by protospacer sequences and by identity 5 -NNNNGNA-3 PAM-containing plasmids in cells carry-
of ‘N’ nucleotides in the PAM. Overall, based on plasmid ing the CcCas9 locus due to interference and correspond-

Downloaded from https://academic.oup.com/nar/advance-article-abstract/doi/10.1093/nar/gkz1225/5707196 by guest on 19 January 2020

transformation interference screening results and in vitro ing increase of the ‘no PAM’ plasmid representation at
DNA cleavage data, we conclude that CcCas9 recognizes a 37◦ C as well as at 22◦ C compared to control (Supplemen-
two-nucleotide 5 -NNNNGNA-3 PAM, with requirement tary File S1, Supplementary Figure S6). The observed ef-
for an A in seventh position being not very stringent. To fect was stronger in colonies formed at 37◦ C than at 22◦ C.
the best of our knowledge, this PAM is distinct from PAM Plasmids with different 5 -NNNNGNA-3 PAM sequences
sequences of known Cas9 nucleases. showed different depletion levels. Thus, the temperature de-
Experiments described above were conducted using the pendence of C. cellulolyticum CRISPR–Cas II-C system
tripartite system composed of CcCas9, crRNA and tracr- can be observed in bacteria as well as in vitro.
RNA. To simplify the CcCas9 DNA-cleavage process, we
sought to design sgRNA, a single guide RNA where crRNA CONCLUSION
is fused to tracrRNA. Several sgRNA variants were tested,
but none were active in vitro (Supplementary Figure S4). Despite the extensive use of Cas9 nucleases for genome en-
Thus, the CcCas9 DNA minimal cleavage system to date gineering, to date, only several Cas9 orthologs can be con-
consists of three components: CcCas9 nuclease, tracrRNA sidered as well-characterized. Given the diversity of Type
and crRNA (Figure 4). Additional studies might reveal the II CRISPR–Cas systems, Cas9 orthologs can show sig-
requirements for a functional sgRNA sequence in this sys- nificant variations in PAM requirements, specificity and
tem. other biochemical properties. In this work, we functionally
One of the possible applications of CcCas9 is genome characterized CRISPR–Cas system from Clostridium cel-
modification of its host, C. cellulolyticum. To facilitate fur- lulolyticum H10. When introduced in E. coli, the C. cel-
ther use of CcCas9 for editing of C. cellulolyticum via the lulolyticum CRISPR–Cas system shows high levels of cr-
single-nick-assisted HR strategy proposed by Xu et al. (11), RNA expression, as well as interference against plasmid
we generated a CcCas9 nickase version by mutating the transformation. The C. cellulolyticum Cas9 effector, Cc-
aspartic acid D8 to alanine in the active site of CcCas9 Cas9, is a Type II-C endonuclease and thus has a rela-
RuvC nuclease domain. The incubation of D8A CcCas9 tively small (compared to other Type II effector proteins)
mutant with a double-stranded DNA target in the presence molecular weight. This nuclease in complex with tracr-
of crRNA and tracrRNA led to cleavage of only one DNA RNA and crRNA actively cleaves DNA targets flanked
strand, as expected (Supplementary Figure S5). by two-nucleotide PAM sequence 5 -NNNNGNA-3 . Most
other small Type II-C Cas9 effectors have more complex
PAM requirements, i.e. NmeCas9, CjeCas9 and GeoCas9
Activity of CcCas9 at different temperatures require, 5 -NNNNGNTT-3 , 5 -NNNNRYAC-3 and 5 -
Based on the initial observations showing that DNA cleav- NNNNCNAA-3 , respectively (7,19–20). The simple, two-
age by CcCas9 is temperature-dependent, we decided to de- nucleotide PAM of CcCas9 may thus be considered as an
termine the dependence of its nuclease activity on temper- advantage for future biotechnology applications. Whereas
ature. Incubation of CcCas9, crRNA, tracrRNA and plas- further studies are needed to check the ability of CcCas9 to
mid carrying a protospacer flanked by consensus PAM se- edit eukaryotic genomes, we envision that the CRISPR–Cas
quence 5 -ACAGGTA-3 at different temperatures led to ef- system characterized here potentially can be conveniently
ficient cleavage of the target in a temperature range of 25– used as an instrument for C. cellulolyticum H10 genome en-
45◦ C with maximal cleavage at 40◦ C (Figure 5A and B). Cc- gineering.
Cas9 cleavage of a linear DNA fragment carrying the same
target site showed similar temperature activity profile. DATA AVAILABILITY
Given the observed differences in CcCas9 in vitro DNA
Raw sequencing data have been deposited with the Na-
cleavage efficiency at room temperature and at 37◦ C, we
tional Center for Biotechnology Information Sequence
compared the CcCas9 CRISPR–Cas II-C system interfer-
Read Archive under BioProject ID PRJNA554628.
ence activity at 22◦ C and 37◦ C. To this end, we used an
equimolar mixture of five PUC19-based plasmids carrying
a protospacer matching the first spacer in the CRISPR ar- SUPPLEMENTARY DATA
ray and flanked by 5 -ACAGGTA-3 , 5 -CGGTGTA-3 , 5 - Supplementary Data are available at NAR Online.
TGAAGAA-3 and 5 -ATTGGAA-3 CcCas9 PAM vari-
ants and a 5 -TTCATAT-3 sequence as a ‘no PAM’ control.
ACKNOWLEDGEMENTS
This 5-members PAM library was transformed into com-
petent E. coli cells carrying pACYC184 CcCas9 locus plas- We thank the Center for Precision Genome Editing and
mid or pACYC184 vector as a control. Cells were plated Genetic Technologies for Biomedicine, IGB RAS, for the
on LB medium supplemented with ampicillin and chloram- equipment
 Nucleic Acids Research, 2020 9

FUNDING 9. Desvaux,M. (2005) Clostridium cellulolyticum: model organism of

mesophilic cellulolytic clostridia. FEMS Microbiol. Rev., 29, 741–764.
Ministry of Science and Higher Education of the Rus- 10. Xu,T., Li,Y., Shi,Z., Hemme,C.L., Li,Y., Zhu,Y., Van Nostrand,J.D.,
sian Federation Subsidy Agreement 14.606.21.0006 He,Z. and Zhou,J. (2015) Efficient genome editing in Clostridium
[RFMEFI60617X0006]. Funding for open access cellulolyticum via CRISPR–Cas9 nickase. Appl. Environ. Microbiol.,
81, 4423–4431.
charge: grant 075-15-2019-1661 from the Ministry of 11. Xu,T., Li,Y., He,Z., Van Nostrand,J.D. and Zhou,J. (2017) Cas9
Science and Higher Education of the Russian Federation. nickase-assisted RNA repression enables stable and efficient
Conflict of interest statement. None declared. manipulation of essential metabolic genes in Clostridium
cellulolyticum. Front. Microbiol., 8, 1744.

Downloaded from https://academic.oup.com/nar/advance-article-abstract/doi/10.1093/nar/gkz1225/5707196 by guest on 19 January 2020

12. Pyne,M.E., Bruder,M.R., Moo-Young,M., Chung,D.A. and
REFERENCES Chou,C.P. (2016) Harnessing heterologous and endogenous
1. Cong,L., Ran,F.A., Cox,D., Lin,S., Barretto,R., Habib,N., Hsu,P.D., CRISPR–Cas machineries for efficient markerless genome editing in
Wu,X., Jiang,W., Marraffini,L.A. et al. (2013) Multiplex genome Clostridium. Sci. Rep., 6, 25666.
engineering using CRISPR/Cas systems. Science (New York, N.Y.), 13. Li,H. and Durbin,R. (2009) Fast and accurate short read alignment
339, 819–823. with Burrows-Wheeler transform. Bioinformatics, 25, 1754–1760.
2. Jiang,W., Bikard,D., Cox,D., Zhang,F. and Marraffini,L.A. (2013) 14. Maxwell,C.S., Jacobsen,T., Marshall,R., Noireaux,V. and Beisel,C.L.
RNA-guided editing of bacterial genomes using CRISPR–Cas (2018) A detailed cell-free transcription-translation-based assay to
systems. Nat. Biotechnol., 31, 233–239. decipher CRISPR protospacer-adjacent motifs. Methods, 143, 48–57.
3. Ronda,C., Pedersen,L.E., Sommer,M.O. and Nielsen,A.T. (2016) 15. Pougach,K., Semenova,E., Bogdanova,E., Datsenko,K.A.,
CRMAGE: CRISPR optimized MAGE recombineering. Sci. Rep., 6, Djordjevic,M., Wanner,B.L. and Severinov,K. (2010) Transcription,
19452. processing and function of CRISPR cassettes in Escherichia coli.
4. Jiang,Y., Qian,F., Yang,J., Liu,Y., Dong,F., Xu,C., Sun,B., Chen,B., Mol. Microbiol., 77, 1367–1379.
Xu,X., Li,Y. et al. (2017) CRISPR-Cpf1 assisted genome editing of 16. Zhang,Y., Heidrich,N., Ampattu,B.J., Gunderson,C.W., Seifert,H.S.,
Corynebacterium glutamicum. Nat. Commun., 8, 15179. Schoen,C., Vogel,J. and Sontheimer,E.J. (2013)
5. Jinek,M., Chylinski,K., Fonfara,I., Hauer,M., Doudna,J.A. and Processing-independent CRISPR RNAs limit natural transformation
Charpentier,E. (2012) A programmable dual-RNA-guided DNA in Neisseria meningitidis. Mol. Cell, 50, 488–503.
endonuclease in adaptive bacterial immunity. Science (New York, 17. Petitdemange,E., Caillet,F., Giallo,J. and Gaudin,C. (1984)
N.Y.), 337, 816–821. Clostridium cellulolyticum sp. nov., a cellulolytic, mesophilic: species
6. Ran,F.A., Cong,L., Yan,W.X., Scott,D.A., Gootenberg,J.S., from decayed grass. Int. J. Syst. Evol. Microbiol., 34, 155–159.
Kriz,A.J., Zetsche,B., Shalem,O., Wu,X., Makarova,K.S. et al. (2015) 18. Leenay,R.T., Maksimchuk,K.R., Slotkowski,R.A., Agrawal,R.N.,
In vivo genome editing using Staphylococcus aureus Cas9. Nature, Gomaa,A.A., Briner,A.E., Barrangou,R. and Beisel,C.L. (2016)
520, 186–191. Identifying and visualizing functional PAM diversity across
7. Kim,E., Koo,T., Park,S.W., Kim,D., Kim,K., Cho,H.Y., Song,D.W., CRISPR–Cas systems. Mol. Cell, 62, 137–147.
Lee,K.J., Jung,M.H., Kim,S. et al. (2017) In vivo genome editing with 19. Ciaran,M.L., Thomas,J.C. and Gang,Bao (2016) The Neisseria
a small Cas9 orthologue derived from Campylobacter jejuni. Nat. meningitidis CRISPR–Cas9 system enables specific genome editing in
Commun., 8, 14500. mammalian cells. Mol. Ther., 24, 645–654.
8. Fonfara,I., Le Rhun,A., Chylinski,K., Makarova,K.S., 20. Harrington,L.B., Paez-Espino,D., Staahl,B.T., Chen,J.S., Ma,E.,
Lecrivain,A.L., Bzdrenga,J., Koonin,E.V. and Charpentier,E. (2014) Kyrpides,N.C. and Doudna,J.A. (2017) A thermostable Cas9 with
Phylogeny of Cas9 determines functional exchangeability of increased lifetime in human plasma. Nat. Commun., 8, 1424.
dual-RNA and Cas9 among orthologous type II CRISPR–Cas
systems. Nucleic Acids Res., 42, 2577–2590.