Two Bioinformatic Tools developed for the Identification and research of Anti-CRISPRs
As a natural adaptive immune system of prokaryotes, CRISPR-Cas systems have become popular in recent years due to their successful application in programmable genome editing. Anti-CRISPR (Acr) proteins, on the other hand, have a great potential to enable a more controllable genome editing.
In the past year we developed a software tool, AcrFinder (http://bcb.unl.edu/AcrFinder/) , to help biologists retrieve putative Acr gene candidates and their genomic contexts from the vast amount of sequenced bacterial, archaeal, and viral genomes. AcrFinder combine three well-accepted ideas in one pipeline, which have been used by previous experimental papers to assist in pre-screening genomic data for Acr candidates. We tested AcrFinder on the genomes that contain 16 experimentally characterized Acr-Aca genomic loci/operons, and found that our tool had a 100% recall.
We also built an online database AcrDB (http://bcb.unl.edu/AcrDB) by scanning ~19,000 genomes of prokaryotes and viruses with AcrFinder and further processed with two machine learning-based programs, AcRanker and PaCRISPR. Compared to other anti-CRISPR databases, AcrDB has the following unique features: (i) It is a genome-centric database with a much larger data; (ii) It offers a user-friendly web interface with various functions (iii) It focuses on the genomic context of Acr and Aca homologs instead of individual Acr protein family; and (iv) It collects data with three independent programs each having a unique data mining algorithm for cross validation. AcrDB and AcrFinder will be valuable resources to the anti-CRISPR research community.
icon search Searchable Transcript
Toggle between list and paragraph view.
[00:00:00.480]So today I'm going to focus my talk
[00:00:02.570]on a tool and a database that we've developed recently
[00:00:05.830]in the identification and researching of Anti-CRISPRs.
[00:00:10.440]So I like to give a brief overview of the presentation.
[00:00:14.660]I'd like to first introduce what CRISPR-Cas systems are
[00:00:17.670]what Anti-CRISPR systems are.
[00:00:19.930]And then we'll move to talk more about the AcrFinder tool,
[00:00:23.280]the AcrDB database
[00:00:24.860]as well as our current work and future directions.
[00:00:30.520]or a clustered regularly inter-spaced
[00:00:32.970]short palindromic repeats was first discovered back in 2005
[00:00:36.870]by Yansen et al.
[00:00:38.300]But back then, they didn't know what CRISPR-Cas system did.
[00:00:41.560]It was not until 2007 that the first experimental evidence
[00:00:44.880]that CRISPR-Cas was an adaptive immune system
[00:00:48.760]wa was published.
[00:00:50.350]So the system functions in sweet different processes
[00:00:53.430]the first would be the adaptation process where
[00:00:56.570]the early Cas complexes would pick up foreign
[00:01:00.230]foreign invading a genetic material
[00:01:03.310]and integrate that into the CRISPR array.
[00:01:05.400]And then during the expression process
[00:01:07.300]the CRISPR right will be transcribed
[00:01:08.810]into these CRISPR-Cas RNAs, which would then form
[00:01:11.820]these surveillance complexes that will then smooth out
[00:01:15.300]in the cell and find re-invading foreign genetic material
[00:01:20.700]in interference process and degrading it.
[00:01:23.120]Thus providing an adaptive immune system.
[00:01:26.800]So right now, if you go on PubMed
[00:01:27.633]and you search CRISPR
[00:01:29.300]and you get more than 23,000 publications,
[00:01:31.680]so very hot topic.
[00:01:34.840]So CRISPR-Cas is currently classified
[00:01:37.400]into two major classes.
[00:01:38.900]And a take home message here is that
[00:01:41.080]class one usually have a much more complex Cas complex
[00:01:45.580]compared to a class two
[00:01:47.190]and that can be represented in this figure here.
[00:01:49.710]And this is also a major reason why class two CRISPR-Cas
[00:01:53.710]are more favored by scientists
[00:01:55.290]in the field of genome editing.
[00:01:56.980]The famous Cas nine is actually of Ca
[00:02:00.910]class two CRISPR-Cas systems.
[00:02:04.560]So the Anti-CRISPRs, the anti-CRISPRs were first discovery
[00:02:09.060]by Joseph Bondy,
[00:02:11.190]back in 2013 in pseudomonas phages and prophages.
[00:02:16.510]Anti-CRISPRs are nature's off switch for CRISPR-Cas.
[00:02:19.810]And this gives it a huge potential
[00:02:22.360]to making genome editing more controllable and more safe.
[00:02:27.040]And this is actually also the major reason
[00:02:29.680]for my current interest in anti-CRISPRs.
[00:02:34.190]So, however the nine initially described anti-CRISPR
[00:02:37.690]protein families did not share any common sequence
[00:02:39.900]motifs that could lead to discovery of new
[00:02:43.590]But they did find that phages and coded anti-CRISPR protein
[00:02:48.090]also encoded putative transcriptional regulator known as
[00:02:51.760]anti-CRISPR associated or Aca
[00:02:55.230]This Aca has this helix-turn-helix motif
[00:02:57.440]which makes it much more conserved compared
[00:02:59.210]to the CRISPR compared to the anti-CRISPR proteins.
[00:03:03.170]Thus further studies used is the non-homology based
[00:03:06.670]approach called the guilt by association
[00:03:08.850]bioinformatic approach in the identification
[00:03:11.657]of potential new anti-CRISPR proteins.
[00:03:15.507]So using this guilt by association bioinformatic approach
[00:03:18.820]the the identified putative anti-CRISPR genes
[00:03:22.500]on the basis of their genetic location
[00:03:24.410]which is upstream of this Aca gene.
[00:03:27.290]So this process is clearly represented in this figure here
[00:03:30.710]you would first look for potential Acas
[00:03:33.360]and you will use that Aca to look for potential Acrs
[00:03:36.213]and you use that Acr to find more Acas
[00:03:39.050]and the cycle repeats and repeats.
[00:03:43.470]So by using this method, the first inhibitors to
[00:03:46.600]type two CRISPR-Cas systems were discovered.
[00:03:49.010]And in addition to the guilt by association method
[00:03:51.730]and alternative bioinformatic approach was
[00:03:55.270]used to discover new anti-CRISPR proteins.
[00:03:58.560]So it's called the Self-targeting spacer approach.
[00:04:01.370]Basically if a CRISPR spacer self targets its own genome
[00:04:04.790]then that will mean it will cut itself.
[00:04:07.030]It will cut its own genome causing self self cell death.
[00:04:12.320]So if the cell is somehow still there
[00:04:15.200]then there must be some type of mechanism or complex
[00:04:18.960]in that cell it's stopping this self-destruction process.
[00:04:23.070]So it was postulated that any genome
[00:04:25.330]in which self targeting occurred was likely to carry
[00:04:27.990]an anti-CRISPR gene in that same cell.
[00:04:31.200]So right now, 11 out of the 28 CRISPR systems
[00:04:34.790]were found to have associated anti-CRISPRs.
[00:04:38.880]And as of 2021,
[00:04:43.260]89 published anti-CRISPRs
[00:04:46.430]are available so much more work needs to be done.
[00:04:52.170]So how does this anti-CRRISPR work?
[00:04:54.240]What are their mechanisms to their function
[00:04:56.400]of the inhibiting CRISPR-Cas systems?
[00:04:58.870]Well, we've so far only scratched the tip of the iceberg
[00:05:03.100]but we do know that it functions through two major routes.
[00:05:06.770]This first one will be to exhibit target DNA
[00:05:09.650]binding stopping the Cas complex from binding
[00:05:12.570]to the target DNA.
[00:05:13.720]And the second one would be to stop the
[00:05:16.740]cleavage of the target DNA once the Cas complex are bound
[00:05:20.980]to its target.
[00:05:22.780]Obviously much more studies are needed
[00:05:25.340]and hopefully we'll know more
[00:05:27.640]about the mechanism of this amazing complex in the future.
[00:05:34.400]So to further the studies of anti-CRISPRs
[00:05:37.410]in 2019 our group developed and published an article based
[00:05:40.580]stubborn bioinformatic pipeline.
[00:05:42.780]And using this publication and the pipeline,
[00:05:47.010]we made a tool called AcrFinder
[00:05:49.840]It combines homology search guilt by association,
[00:05:54.220]And it's the first tool to identify anti-CRISPRs
[00:05:56.530]as a genetic operand.
[00:05:57.640]This one was published last year
[00:05:59.900]on the Nucleic Acid Research.
[00:06:02.960]So the pipeline of AcrFinder consists of two major routes.
[00:06:06.410]The first route is the homology search
[00:06:08.190]the traditional homologies search approach.
[00:06:10.890]And the other route indicated by the blue arrow
[00:06:12.600]here is my personal favorite
[00:06:13.830]combines guilt by association and self-targeting spacers.
[00:06:17.860]And based on
[00:06:20.520]all the outputs, the discovered potential
[00:06:24.150]anti-CRISPR operands will be provided with corresponding
[00:06:27.240]confidence levels as the tools output.
[00:06:30.530]So we've also made a, a web server for the tool.
[00:06:34.580]You can upload the genome of interest
[00:06:36.170]and then screen for potential anti-CRISPRs.
[00:06:39.350]And I've attached the link here.
[00:06:41.100]You are more than welcome to check it out
[00:06:42.890]if you are interested.
[00:06:44.510]So based on ,
[00:06:46.370]this, this AcrFinder tool that we, that we developed
[00:06:50.040]we built a database called AcrDB or anti-CRISPR DB.
[00:06:55.229]We also collaborated with other groups
[00:06:57.660]when building the database,
[00:06:58.660]we collaborated with the AcRanker team
[00:07:00.750]which is the first machine learning based tool
[00:07:02.970]to discover anti-CRISPR.
[00:07:04.850]And the PaCRISPR team.
[00:07:06.030]which is, which is another machine learning tool
[00:07:07.950]but also company the most accurate
[00:07:10.150]of machine learning tool to anti-CRISPR discovery.
[00:07:14.470]And we've built a web server for
[00:07:21.490]Here, a couple of screenshots, you can browse
[00:07:25.784]by different options.
[00:07:26.670]And then result sections
[00:07:28.970]and B here has tables contain a variety
[00:07:31.530]of information on the anti-CRISPR.
[00:07:33.840]There's even a circular representation
[00:07:35.890]of the genome of which the anti-CRISPR in
[00:07:38.360]and also give the user a better understanding
[00:07:42.280]of their genetic, a type sporogenic opera.
[00:07:47.330]So right now I'm developing a sequence processing
[00:07:50.630]and machine learning hybrid anti-CRISPR discovery tool
[00:07:53.670]and using this tool combining with others
[00:07:56.070]I'm hoping to screen the human microbiome
[00:07:58.780]and observe for novel biological phenomenon.
[00:08:02.250]And with that, I'd like to conclude my presentation.
[00:08:05.870]Thank you very much for listening.
Log in to post comments