Novel CRISPR-Cas12a Clades Discovery Using Large Language Model

doi:10.21203/rs.3.rs-4817511/v1

Download PDF

Article

Novel CRISPR-Cas12a Clades Discovery Using Large Language Model

https://doi.org/10.21203/rs.3.rs-4817511/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The identification and engineering of CRISPR-Cas systems revolutionized life science. Metagenome contains millions of unknown Cas proteins, which require precise prediction and characterization. Traditional protein mining mainly depends on protein sequence alignments. In this work, we harnessed the capability of the evolutionary scale language model (ESM) to learn the information beyond the sequence. After training with the CRISPR-Cas sequences and their functional annotation, the ESM model can identify the CRISPR-Cas proteins from the annotated genome sequences accurately and robustly without sequence alignment. However, due to the lack of experimental data, the feature prediction is limited by the small sample size. Integrated with machine learning on small size experimental data, the model is able to predict the trans-cleavage activity of novel Cas12a. Furthermore, we discovered 7 novel subtypes of Cas12a proteins with unique organization of CRISPR loci and protein sequences. Notably, structural alignments revealed that Cas1, Cas2, and Cas4 also exhibit 8 subtypes, with the absence of integrase proteins correlating with a reduction in spacer numbers within CRISPR loci. In addition, the Cas12a subtypes displayed distinct 3D foldings, a finding further corroborated by CryoEM analyses that unveiled unique interaction patterns with RNA. Accordingly, these proteins show distinct double-strand and single-strand DNA cleavage preferences and broad PAM recognition. Finally, we established a specific detection strategy for the oncogene SNP without traditional Cas12a PAM. This study shows the great potential of the language model in the novel Cas protein function exploration via gene cluster classification.

Biological sciences/Microbiology/CRISPR-Cas systems/CRISPR-Cas9 genome editing

Biological sciences/Systems biology/Synthetic biology

language model

CRISPR

Cas12a

gene editing

Clustered regularly interspaced short palindromic repeats (CRISPR) and CRISPR-associated (Cas) proteins constitute the adaptive immune system in prokaryotes to defend against invasive genetic elements.¹ The CRISPR-Cas systems keep evolving during natural evolution, currently, two classes have been identified and new subtypes are emerging.² Class 1 comprises types I, III, and IV, featured with multiple Cas proteins as effector modules. Class 2 comprises types II, V, and VI, mainly using single multidomain proteins, e.g. Cas9, Cas12, and Cas13, as effector modules. Due to its ease of reprogrammability, class 2 Cas proteins have been widely applied in genome editing, nucleic acid detection, imaging, and annotation, which lead to the research revolution from basic science to translational medicine.³ Novel CRISPR-Cas discovery is fundamental for the technology iteration. Metagenome provides a precious reservoir for the novel Cas exploration. Recently, the compact gene editor Φ was discovered in the genome of huge phages, which were identified from the metagenomic dataset.^4–6 Nevertheless, millions of unknown Cas proteins, scattering in the metagenomes, are still waiting to be characterized.

Traditional Cas protein mining mainly depends on the primary sequences to predict the protein function and classification.² Sequence similarity-based search could be performed by Basic Local Alignment Search Tool (BLAST) and Hidden Markov Model (HMM).^7,8 Based on the sequence similarity search, MacSyFinder and HMMCAS have been developed.^9,10 But these methods are restrained by the known Cas sequence and difficult to discover new Cas proteins. Machine learning-based methods predict novel Cas proteins in a data-driven manner, for example, CASpredict, CASboundary, CRISPRCasTyper.^11–13 However, the protein function is directly determined by the three-dimensional structure but not amino acid sequences. Learning the hidden biological information from protein sequences motivates the development of various evolutionary-scale language models.¹⁴ They can predict secondary structure and tertiary structures.^15,16 Recently, structure-based protein clustering discovered novel clades of deaminase and generates an efficient cytosine base editor, indicating the importance of structural information in the novel protein discovery.¹⁷ In addition to the three-dimensional structural information, an alternative approach could be the evolutionary-scale language model. The recently developed ESM-2 language model, scaling up to 15 billion trainable parameters, can capture the protein feature at the atomic-resolution level.¹⁸

Here we developed an Artificial Intelligence assisted CRISPR-Cas Scan (AIL-Scan) strategy based on the ESM language model. After training with the CRISPR-Cas sequences and their functional annotation, the AIL-scan can accurately distinguish the different CRISPR-Cas types from the annotated genome sequences. However, only a few of Cas proteins are experimentally evaluated. We integrated the ESM and machine learning on small sample size and developed a trans-cleavage activity predictive model with accuracy. The Cas12a family was taken as an example to explore in the metagenomic database. Different from the classical CRISPR loci of Cas12a, we discovered 8 subtypes of the Cas12a family, which were characterized by the unique organization of CRISPR loci and protein sequences. Furthermore, the integrase proteins, i.e., Cas1, Cas2, and Cas4 also have 8 subtypes, respectively, according to the structural alignments. The missing integrase proteins result in the decrease of spacer numbers in the CRISPR-loci. In addition, the novel Cas12a proteins show diverse 3D foldings between subtypes. The CryoEM analyses further discovered unique interaction patterns with RNA. Accordingly, these proteins show distinct double-strand and single-strand DNA cleavage preferences and broad PAM recognition, which enables the specific detection of the oncogene single nucleotide polymorphisms (SNP) without traditional Cas12a PAM and efficient cellular genome editing with minimal off-targets. The study provides new insights into machine learning in the discovery of novel functional Cas proteins via gene cluster classification.

Development of an Artificial Intelligence assisted CRISPR-Cas Scan (AIL-Scan) strategy based on ESM large language model

We assumed by embedding the functional feature with protein primary sequences, we could trace the natural evolution rules and identify the CRISPR-Cas proteins in the metagenomics data directly without sequence alignments. To identify the CRISPR-Cas proteins, we developed an Artificial Intelligence assisted CRISPR-Cas Scan (AIL-Scan) strategy (Fig. 1a). It includes the following steps:

CRISPR-Cas training data is created by extracting CRISPR-associated (Cas) proteins from NCBI database, classification by genes, and removing redundant sequences.

Supervised fine-tuning of ESM on the CRISPR-Cas training data based on the biological information to predict the novel Cas protein.

Feature prediction of Cas proteins, including cleavage activity, CRISPR-loci type, CRISPR loci-length, direct repeats, spacers, evolutionary analyses, MSA, and structures.

We generated our training data using reviewed NCBI gene data. We annotated the Cas1, Cas2, Cas3, Cas4, Cas5, Cas6, Cas7, Cas8, Cas9, Cas10, Cas12, and Cas13. Non-Cas proteins were extracted according to the following rules, without the annotation of Cas, and remove the proteins with sequence similarity over 40%. The Cas protein database was separated into training or validation database using the CD-HIT-2D with 40% identity threshold to remove the redundant sequences and avoid overfitting. We collected 77684 non-redundant positive sequences and 13047 non-Cas proteins. The maximal protein length is less than 1764 amino acids. To obtain the best classification, we introduced the “focal loss” in the classification to solve the unbalance of the input data. We obtained the best model during the 13th Epoch of model training and obtained 97.75% accuracy for the ESM 2 model with 650 million (650M) parameters (Extended Data Fig. 1). For the 15 billion (15B) parameters model, we achieve the best performance in the 9th Epoch with 98.22% accuracy (Extended Data Fig. 1). The accuracy and prediction speed of AIL-Scan is comparable to the CRISPRcasIdentifier, which integrates hidden Markov models and machine learning (Table 1 and Extended Data Fig. 2). CASPredict perfomed the highest speed among the four softwares although its accuracy is lower than the machine learning based softwares, i.e. AIL-Scan and CRISPRcasIdentifier. Considering the resource consumption, the 650M model is sufficient for the Cas prediction. We used ESM embeddings to reduce dimensionality with t-SNE for 77684 sequences and discovered that ESM can distinguish the differences in various Cas classifications (Fig. 1b). The receiver operating characteristic (ROC) curve (AUC) indicates that the probability that the positive sample’s decision value is greater than the negative sample’s decision value for all the Cas and Non-Cas proteins (Fig. 1c). The test loss and test accuracy also indicate that the model generalizes correctly and performs well on unseen data (Fig. 1d).

Table 1

Cas protein prediction accuracy using different models.
	650 million ESM	15 billion ESM	CasIdentifier HMM + ML	HMMCAS HMM	CASPredict SVM
cas1	98.5%	98.5%	99.5%	97.4%	70.8%
cas2	96.3%	99.4%	96.9%	94.4%	73.8%
cas3	98.1%	98.7%	98.1%	97.7%	66.3%
cas4	89.9%	94.4%	92.1%	96.6%	65.2%
cas5	97.7%	98.5%	93.1%	82.3%	61.5%
cas6	99.0%	98.7%	99.4%	98.1%	61.5%
cas7	100.0%	100.0%	100.0%	95.5%	77.3%
cas8	88.9%	100.0%	100.0%	88.9%	77.8%
cas9	100.0%	100.0%	98.7%	99.3%	91.4%
cas10	99.6%	99.6%	99.2%	99.6%	82.3%
cas12	100.0%	100.0%	97.4%	39.5%	76.3%
cas13	100.0%	100.0%	100.0%	N.A.	93.3%
Non-Cas	99.8%	99.9%	96.5%	N.A.	75.2%

We use the Global Microbial Gene Catalog (GMGC) metagenomic database for the novel Cas protein discovery.¹⁹ We selected 50 000 bins with high quality from GMGC and extracted 20 000 MAGs including CRISPR-loci to test the performance of AIL-scan. The protein sequences were predicted by Prodigal software.²⁰ We collected ca. 20 000 000 protein sequences shorter than 1500 amino acids for prediction. In comparison with the established methods, the AIL-scan predicts 1379 Cas12a sequences.

Development of trans-cleavage activity prediction model

The trans-cleavage activity of Cas12a has been used in various applications. Although many CRISPR-Cas12a proteins have been identified, few of them have been tested in the trans-cleavage experiments. Therefore, the main challenge encountered during this study lies in dealing with a small sample size coupled with high-dimensional embeddings which often leads to convergence issues when employing most models. A total of 69 labeled Cas12a proteins (included with three known Cas12a) were included in our analysis (Table S1). Their trans-cleavage activity weas assessed by the fluorophore-quencher (FQ) reporter assay. Trans-cleavage activity was defined as proteins displaying fluorescence intensity twice that of the negative control. Thirty-three proteins classified as active in trans-cleavage activity and the remaining 36 proteins categorized as inactive. To evaluate the performance of our predictive model, a test set comprising 13 randomly selected proteins (approximately 20% of the sample) was used while the remaining 56 proteins were employed for training purposes. Initially, we recorded the last embedding layers based on our fine-tuned ESM model for all labeled Cas12a protein sequences. These embeddings (1280 dimensions) were utilized as covariates to predict trans-cleavage activity.

Different forms of decision tree models are evaluated in this task. The results of our study demonstrate that Light Gradient Boosting Machine (LightGBM) achieves the highest accuracy among mainstream machine learning models, with an accuracy rate of 69.2% on the test set trained on embeddings. In order to mitigate dimensionality-related effects, principal component analysis (PCA) was applied to eliminate redundant embeddings while assessing if prediction power remained unaffected by this reduction in dimensionality range spanning from 2 to 15 principal components. Detailed results are presented in Table 2. LightGBM, CatBoost and RandomForest achieve the accuracy of 92.3% in the test set (12 out of 13 proteins are correctly labeled) with 4, 6 and 8 principal components correspondingly. We can see that comparing to train models directly with embeddings, removing redundant dimensions with PCA provides higher accuracies in predicting trans-cleavage activity.

Table 2

Cas12a protein trans-cleavage activity prediction accuracy using different strategies.
Best Model + Covariates	Accuracy in Test Set
LightGBM [1] + Embeddings	69.2%
CatBoost [2] + PCA-2	69.2%
CatBoost [2] + PCA-3	84.6%
CatBoost [2] + PCA-4	92.3%
CatBoost [2] + PCA-5	84.6%
LightGBM [1] + PCA-6	92.3%
CatBoost [2] + PCA-7	76.9%
RandomForest [3] + PCA-8	92.3%
Neural Network [4] + PCA-9	84.6%
CatBoost [2] + PCA-10	76.9%
CatBoost [2] + PCA-11	84.6%
CatBoost [2] + PCA-12	69.2%
CatBoost [2] + PCA-13	69.2%
CatBoost [2] + PCA-14	84.6%
Neural Network [4] + PCA-15	69.2%

New CRISPR-Cas12a loci predicted from the metagenomics

We did further feature analyses of Cas12a candidate proteins. Phylogenetic analysis of Cas12 proteins suggests that the novel Cas12a proteins fall into the Cas12a clade (Fig. 2a). The classical CRISPR-loci, comprising essential elements such as Cas1, Cas2, and Cas4, plays a pivotal role in type classification. To delve into these features, we employed AIL-scan to predict Cas1, Cas2, and Cas4 proteins within the same CRISPR loci adjacent to the Cas12a sequence. Subsequently, we meticulously verified 300 predicted CRISPR loci to gain deeper insights manually. Normally, the Cas12a is considered to have a unique CRISPR loci, comprising Cas1, Cas2 and Cas4. Intriguingly, the observed count of Cas1, Cas2, and Cas4 proteins was notably lower than that of Cas12a, suggesting the absence of these small Cas proteins in some Cas12a loci (Fig. 2b, c). Further stratification based on the number of integrase proteins led to the classification of CRISPR loci into eight distinct subtypes. The distribution of integrase proteins across these subtypes exhibited a sparse pattern (Fig. 2d). Notably, Subtype VIII lacked any integrase proteins, Subtype I encompassed Cas1, Cas2, and Cas4, while Subtype VI exclusively featured Cas2. This nuanced classification sheds light on the diversity within CRISPR loci and underscores the intricate variations in the composition of integrase proteins among different subtypes. Our observations may provide new perspectives on correlations among different CRISPR-Cas systems and integrase proteins. Remarkably, the analyses using the 1000 predicted CRISPR Cas12a loci without manual verification show a strikingly similar distribution pattern as the result from the 300 manually confirmed ones, indicating this distribution is a universal phenomenon (Extended Data Fig. 3). To provide further insights, we measured the length of CRISPR loci, beginning from the start of the Cas12 protein and concluding at the first spacer. Subtype VIII emerged as the shortest, spanning a mere 4200 bp, while Subtype I claimed the longest, extending over 6100 bp. Particularly noteworthy were certain Subtype I CRISPR loci exhibiting extraordinary lengths of up to 6700 bp, raising the possibility of harboring enigmatic protein elements (Fig. 2e). Aligned with the integrase variation, the numbers of spacers notably decreased in Subtypes IV, VI, and VIII, underscoring the pivotal roles of integrases in spacer capture (Fig. 2f). Despite the divergence in spacer numbers, the stem-loop region corresponding to direct repeat sequences remained conserved (Fig. 2g). This consistent conservation hints at a shared structural element, emphasizing the importance of the stem-loop region in CRISPR loci across different subtypes.

To explore the distribution of the novel proteins in the organisms, we constructed a phylogenetic tree using 300 candidate Cas12a proteins, which were manually verified, along with three known Cas12a (LbCas12a, FnCas12a and AsCas12a). 232 Cas12a proteins from the Lachnospiraceae family cluster into one clade. Within this clade, subclade 1 consisted of 62 subtype I Cas12a proteins, 81 subtype VII Cas12a proteins, and a modest representation of other subtypes. Notably, Subtype I and Subtype IV emerge as the principal constituents within Subclade 2. Furthermore, Subclade 3 is marked by the exclusive presence of 28 subtype VIII Cas12a proteins originating from the Acutalibacteraceae family. It is worth noting, 94.6% of the novel Cas12a proteins originate from enteric microorganisms (Fig. 2h), which may be due to the ease to recover high quality genome from enteric microorganisms.

New Cas integrases in CRIPSR loci

New insights into Cas integrases within CRISPR loci have emerged, reshaping our understanding of their structural diversity and functional roles.^21–25 The Cas1, Cas2, and Cas4 serve as integrases to capture foreign DNA motifs into bacterial CRISPR loci.²⁴ Therefore, they are key elements in the bacterial immune system. To our surprise, pairwise structural alignments for the integrases predicted by AlphaFold2²⁶ revealed 8 subtypes in Cas1, Cas2, and Cas4 family, respectively (Fig. 3 and Extended Data Fig. 4).

The Cas1 proteins, varying in length from 92 to 331 amino acids, were solicitously classified into eight distinct types based on both their length and protein structural features (Fig. 3a, 3b and Extended Data Fig. 4b). Predominantly, Cas1 proteins were mainly clustered within Type 8, demonstrating structural resemblance to AfCas1 (PDB: 4N06)²⁷ upon examination via the Dali server²⁸ (Extended Data Fig. 4c). Type 8 Cas1 proteins comprise an N-terminal domain (NTD) and a C-terminal domain (CTD), linked together by a connecting loop. The catalytic sites are located in α6, α8 and the loop between α7 and α8. In comparison to type 8, type 7 Cas1 proteins lack the α1 helix and five β sheets (β2-β4 and β6) of NTD, Type 6 Cas1 proteins, on the other hand, lack both the α9 and α10 helices in the CTD. Type 5 Cas1 proteins only contain the last five helix (α6-α10) of CTD, however, type 4 Cas1 proteins only contain the last four helix (α7-α10) of CTD, albeit with a lengthened α7 helix. Type 3 Cas1 proteins present a complete NTD and two CTD helices (α3 and α4). The interface formed between the NTD domain of Cas1and Cas2 is essential for the acquirement of foreign DNA, and partial deletion of Cas1 CTD domain did not abolish spacer acquirement. Therefore, the structural elements absence in the CTD domain may not impact the DNA acquisition process²⁹. However, type 1 Cas1 protein present distinct structure feature with a βαββαβα fold. This systematic classification not only highlights the diverse structural variations among Cas1 proteins but also provides a comprehensive understanding of their distinct domains and catalytic sites.

The Cas2 proteins span a range of 70 to 146 amino acids, classified into 8 types (Fig. 3c, 3d and Extended Data Fig. 4e). Among these, Type 8 Cas2 proteins are recognized for their N-terminal domain comprising three helices and a C-terminal domain adopting an αββαββ fold structure. The CTD of type 8 Cas2 proteins resembles with E.Coli Cas2 (PDB: 5DQT), however, it is unclear whether the additional three helices participate in spacer acquirement. (Extended Data Fig. 4f)³⁰. Conversely, Type 7 Cas2 proteins lack two C-terminal β sheets (β3-β4) and exhibit distinct conformational variations in their N-terminal helices, the absence β sheets are partial Cas2 dimer interface, which may disturb its function. Type 6 Cas2 proteins share the absence of two C-terminal β sheets (β3-β4) and N-terminal helices (α1 and α2). Predominantly, Cas2 proteins belong to Type 5, characterized by the retention of the C-terminal domain, which resembles with traditional Cas2 proteins Type 4 Cas2 proteins retain only the C-terminal domain but with a different 2D structural arrangement, i.e., βαββαβ, initiated by a β0 sheet replacing the original β3 sheet. Type 3 Cas2 proteins possess a partial C-terminal domain with altered orientations of the β1 and β2 sheets. Comparatively, Type 2 Cas2 proteins exhibit shorter β1 and β2 sheets. Finally, Type 1 Cas2 proteins are defined by the absence of the β3 and β4 sheets in their C-terminal domain. The varying lengths and flexible feature of the Cas2 C-termini give rise to the prospect of diverse binding modalities within Cas1–Cas2 complexes³¹.

Cas4 proteins span from 79 to 206 amino acids and can be categorized into eight distinct types (Fig. 3e, 3f and Extended Data Fig. 4h). Notably, Type 8 Cas4 proteins bear structural resemblance to I-C Cas4 protein (PDB: 8D3Q)²² (Extended Data Fig. 4i). Type 8 Cas4 proteins present the core αβα protein domain capped by a small, mostly α-helical subdomain. Compared to I-C Cas4 protein, type 8 Cas4 protein has an additional N-terminal helix, but lack a C-terminal helix, which contacts with Cas2 protein and have essential effect on protospacer cleavage²² Comparing Type 8 Cas4 proteins to their counterparts, Type 7 Cas4 proteins exhibit the absence of the α1 helix and a shorter α3 helix. Furthermore, Type 6 Cas4 proteins are characterized by the lack of β3 sheets in addition to the aforementioned features. Type 5 Cas4 proteins feature a truncated α4 helix and the absence of α1, α6, and α7 helices. Type 4 Cas4 proteins lack the C-terminal α6 and α7 helices. Building upon the features of Type 4 Cas4 proteins, Type 3 Cas4 proteins further lack the β6 sheet. Additionally, Type 2 Cas4 proteins, extending the pattern of deficiency, lack the β5 sheet. The absence of β5 and β6 sheets in Cas4 proteins disrupt spacer insertion orientation, while other missing structural elements are less likely to impact spacer integration with Cas1 or Cas2 proteins. However, Type 1 Cas4 proteins exhibit the most concise structure, comprising only four helices (α2-α4 and αX). This in-depth analysis not only reveals the universal subtypes within Cas integrases but also deepens our comprehension of the structural intricacies that define their functions within bacterial immune systems.

New Cas12a proteins have distinct cis- and trans-cleavage activities

Cas12a processes the pre-crRNA transcripts into mature crRNA by its endoribonuclease activity. Then the Cas12a-crRNA complex efficiently cis-cleaves a double-strand DNA (dsDNA) which is initiated by a PAM motif recognition. The cleaved DNA segment that remains bound then induces non-specific degradation of single-strand DNA (ssDNA) (Fig. 4a).

Therefore, we evaluated the RNA binding efficiency, DNA binding efficiency, cis- and trans-acting DNase activities of sixteen novel Cas12a proteins from eight subtypes derive from Anaeroglobus micronuciformis (AmCas12a), Eubacterium_G ventriosum (EvCas12a_1 and EvCas12a_2), Erysipelatoclostridium sp. (EspCas12a), Ruminococcus_E sp. (RspCas12a_1 and RspCas12a_2), Agathobacter rectale (ArCas12a), Lachnospira eligens (LeCas12a_1 and LeCas12a_2), UBA3388 sp. (UBACas12a), RC9 sp. (RCCas12a), CAG-127 sp. (CAGCas12a), Ruminococcus_E bromii_B (RbrCas12a_1, RbrCas12a_2, RbrCas12a_3 and RbrCas12a_4) (Fig. 4 and Extended Data Fig. 5a, 5b and Table S2). Remarkably, the direct repeat sequence of these candidate Cas12a proteins is conserved alongside their celebrated counterparts, i.e. LbCas12a (Fig. 2g). Therefore, we chose LbCas12a as the positive control in the following assays as well as its crRNA scaffold in the screening step. All the novel Cas12a proteins show RNA and DNA binding ability as expected (Fig. 4b, 4c and Extended Data Fig. 5c, 5d). However, the DNA binding ability of subtype I and subtype VIII are higher than other Cas12a proteins. According to the inherent trans DNase activity of Cas12a, as well as the 4 bp PAM length, we develop a simple and efficient PAM detection method. We constructed 6 short dsDNA target arrays by annealing 256 kinds of PAM sequence primer pairs in each well, which target EMX1 site1, DNMT1 site1, FANCF site1, MerS site1, eGFP site1 and eGFP site 3. Each dsDNA target was incubated with candidate Cas12a proteins, crRNA and FAM-BHQ reporter to detect fluorescence of each reaction system (Fig. 4d). Using this assay, we determined the PAM preference of EvCas12a_2, AmCas12a, RspCas12a_2, CAGCas12a and RbrCAS12a_1, EcCas12_2, RspCas12a_2, and CAGCas12a recognize T rich PAM, but AmCas12a prefer G-start PAM, RbrCas12a_1 recognize 5-GTV-3 PAM (Fig. 4e-4i and Extended Data Fig. 6).

To corroborate the cis-acting DNase activity of candidate Cas12a proteins, we incubated Cas12a proteins with a crRNA and a linearized plasmid dsDNA. All linearized dsDNA were degraded by candidate Cas12a proteins with comparable efficiency to LbCas12a at 37^oC, with the exception of RCCas12a (Fig. 5a and Extended Data Fig. 7a). Sanger sequencing of the cleaved DNA ends revealed that AmCas12a introduced INDELs at 18 in NTS and 23 in TS, consistent with other Cas12a orthologs. (Extended Data Fig. 7e and 7f) However, Most Cas12a variants exhibited diminished DNase activity, resulting in the production of uncleaved DNA at room temperature (RT), except for subtype VIII Cas12a proteins, which lacks integrases.(Fig. 5b, Extended Data Fig. 7b). Subtype II Cas12a variants slightly less active than LbCas12a in single-strand (ssDNA) degradation, while EspCas12a, EvCas12a_1, EvCas12a_2 and ArCas12a exhibited moderate activity. In contrast, the other Cas12a variants displayed notably lower activity (Fig. 5c and Extended Data Fig. 7c). Most of the novel Cas12a proteins represent considerable cis cleavage activity but a bit different in trans-cleavage activity compared to LbCas12a. The ion preference assay reveals that the novel Cas12a proteins can be activated by Mn²⁺, similar to the LbCas12a.³² Divalent Mg ions prove ineffective in activating the trans ssDNA cleavage activity of low activity Cas12a variants, Mn²⁺ cation emerges as the catalyst for their trans DNase activity. (Fig. 5d and Extended Data Fig. 8) To investigate the genome-editing ability of candidate Cas12a in eukaryotic cells, we select 6 target sites with canonical PAM, which can be recognized by all the novel Cas12a (Fig. 5e and Extended Data Table 5). AmCas12aexhibits an average editing efficiency of 49.6% across six sites, with remarkable peaks at sites 3 (85.4%) and 6 (84.9%). In contrast, EvCas12a_2 displays an average editing efficiency of 20.3%, with its highest performance observed at site 1 (25.8%). RspCas12a_2 and RbrCas12a_2, which lack integrase, yield modest average editing efficiencies of 14.3% and 17.8%, respectively, with notable peaks at site 3 (26.3% and 37.3%, respectively). ArCas12a shows comparable average editing efficiencies with AmCas12a (45.4%), which get notable peaks at sites 3 (75.8%). LeCas12a_1 shows an average editing efficiency of 6.2% and a maximum efficiency of 25.7% at site 2. UBACas12a exhibits nearly negligible editing efficiency with the highest activity reaching 2.1%. At sites 4, CAGCas12a and LeCas12a_2 demonstrate peak genome-editing efficacy, at 81.7% and 73.8%, respectively, with mean editing efficiencies of 28.8% and 26%, respectively. AsCpf1 attains an impressive average editing efficiency of 65.5%, with its zenith at site 6 (84.7%). Finally, LbCas12a shows an average editing efficiency of 25.6% and a maximum efficacy of 53.5% at site 6.

The AmCas12a-crRNA binary complex

The protein sequence identity of 16 candidate Cas12a proteins to AsCas12a, FnCas12a and LbCas12a are pretty low, ranging from 30%-46% (Fig. 6a and Extended Data Fig. 9a). We built the structure models of 300 novel Cas12a proteins using AlphaFold2, except for the failed construction. Within this three-dimensional landscape, it is apparent that Cas12a proteins congregating within the same subclade share a commendable degree of structural concordance, However, AmCas12a presents a subtle deviation, distinguishing itself somewhat from its subclade I Cas12a counterparts (Fig. 6d and 6f).

To understand the molecular details underlying the RNA binding behavior of AmCas12a, we achieved the cryo-EM map of crRNA binding complex, which consists of AmCas12a and a 44-nt crRNA, at 2.9 Å resolution (Fig. 6b, 6c, Extended Data Fig. 10 and Table 7). The AmCas12a-crRNA structure maintains a bilobed architecture (Fig. 6c), similar to other Cas12a structures.^33,34 Nonetheless, it is noteworthy that the AmCas12a-crRNA complex exhibits a distinct conformation when juxtaposed with its counterparts. Specifically, an observable rotational variance is discernible within the REC domain of AmCas12a when compared to the LbCas112a-crRNA and FnCas12a-crRNA complexes. Relative to LbCas12a and FnCas12a, the REC1 domain of AmCas12a presents a deviation of 7.3° and 9.4°, respectively. Simultaneously, the REC2 domain of AmCas12a manifests a rotational disparity of 4.8° and 6.2°, respectively (Extended Data Fig. 11d and 11e).

As observed in the LbCas12a and FnCas12a crRNA binary structures, the repeat-derived pseudoknot in the 5’ handle of the crRNA is ordered. However, the crRNA conformation is markedly different from that of the crRNA bound by LbCas12a or FnCas12a. Due to the flexibility of the spacer-derived part of crRNA, it’s almost unclear in the Cas12a-crRNA binary complex.^34,35 Notably, an extra RNA stem formed by A(1)-A(5) and U(18)-U(22) within the crRNA spacer region makes a part of spacer region including seed sequence well-defined in the central cavity of AmCas12a and adopt an A-form-like helical conformation, but A(-10)-G(-6) and G(6)-A(15) nucleotides of crRNA are unclear (Fig. 6b and Extended Data Fig. 9b). To accommodate the double RNA stem substrate, the REC lobe of AmCas12a rotates away from the NUC lobe. Unsurprisingly, the docked AmCas12a-AF-crRNA `complex presents a severe clash in the REC domain (Extended Data Fig. 9c). The attainment of conformational integrity within the extra RNA stem is orchestrated by intricate interplays involving the ribose and phosphate moieties of the crRNA backbone, engaging in multiple interactions with specific residues within the WED, REC1, and RuvC domains of AmCas12a (Fig. 6e). These include residues T19, H751, K522, and H861 from the WED domain, Y50 and R168 from the REC1 domain and Q1003 from the RuvC domain, all of which are conserved with Cas12a orthologs, except Q1003 which form a hydrogen bond with the phosphate of U(18) (Extended Data Fig. 12). Distinct from the FnCas12-crRNA complex, the spacer segment of crRNA major interacts with the WED domain of AmCas12a.

Compared to the LbCas12a-crRNA complex and FnCas12a-crRNA complex, the divalent Mg ions are located in the same location (Extended Data Fig. 11a, b and c). Consistent with a seed sequence-dependent mechanism of DNA targeting and in broad agreement with previous analyses of AsCas12a, LbCas12a activities in vivo and FnCas12a activities in vitro^33,36,37, cleavage of DNA substrates with single-nucleotide mismatches in the seed segment was almost completely impaired, while mismatches in the PAM-distal region of the DNA target were mostly tolerated (Fig. 6g).

Specific detection of single nucleotide mutation by AmCas12a

Cas12a is promising tool in the next generation molecule diagnosis, however it is suffered from the PAM limitation.³⁸ The oncogene SNP only has a small sequence window to probe, the traditional PAM, TTTV, could not cover all the SNPs. Therefore, we tested whether the AmCas12a can distinguish the SNPs without traditional PAM. (Fig. 7a) The oncogene mutants, KRAS c.34G > T (G12C) did not contain the available TTTV in the adjacent sequences (Fig. 7b). We designed the crRNA targeting the SNP (Fig. 7b). According to the fluorescence intensity, we selected the crRNAs inducing the strongest signals, i.e. crRNA 1 for KRAS mutant (Fig. 7c). The AmCas12a can detect 10 copies of KRAS mutant (Fig. 7d). Furthermore, We diluted the target mutant and evaluated the sensitivity of detection. the AmCas12a can even distinguish 0.1% KRAS mutant in the wild type gene background, which is more sensitive than the Sanger sequencing (Fig. 7e and 7f).

CRISPR-Cas system keeps evolving during the arm-race between bacteria and phages. The proteins gain or lose function via mutation or domain reorganization, accordingly the accessory proteins such as integrases Cas1, Cas2, and Cas4 changes in the CRISPR loci. Therefore, tracing the Cas protein sequences in the CRISPR loci provides a practical strategy to decipher the evolution of the CRISPR-Cas system. However, this process is complex due to the diversity and huge data from metagenomics. In this work, the AIL-Scan model leveraged the capability of ESM-2 in residue resolution prediction for the CRISPR-Cas identification. The 15 billion model shows superior prediction capability for all types of CRISPR-Cas in comparison with the 650 million model and other machine learning software. (Table 1) The small Cas proteins, e.g. Cas1, Cas2, Cas3, Cas4, Cas5, and Cas8 are difficult to predict, as shown by the relatively low prediction accuracy (Table 1), because the short sequences provide too limited information to accurately predict. The high accurate model can extract more precise information from short sequences, when we increase the size of the model, we observe a significant increase in the prediction accuracy for the small Cas proteins. The superior capability of the large model also contributed to the precise prediction of the non-Cas proteins (Table 1). However, considering resource consumption, the 650M model can practically handle the Cas protein classification.

Given the absence of universal cas genes and the frequent modular recombination, CRISPR-Cas classification requires multipronged parameters, including the signature cas genes, sequence similarity between the shared Cas proteins, and the organization of the genes in the CRISPR-Cas loci.^39,40 The Cas12a proteins are well-characterized Cas proteins, normally containing three integrases, i.e. Cas1, Cas2, and Cas4. Paradoxically, we discovered 7 new subtypes with distinct integrase combinations in the metagenomic data (Fig. 2). We aligned the Cas12a protein sequences and integrated with the information of subtypes and species. Remarkably, even in the same bacterial family, such as Lachnospiraccae and Acutalibacteraccae, the variants are diverse. The Cas12a proteins in the Lachnospireaccae family share high similarities, but the CRISPR-loci is mainly dominated by the subtype I (40.7%) and followed by subtype VI (35.9%). Interestingly, most CRISPR-Cas12a proteins in the Acutalibacteraccae family do not contain any integrases, only 11.8% belong to subtype 2. Cas1, Cas2, and Cas4 locates the downstream of Cas12 in a tandem pattern. Cas1-Cas2 complex is necessary for the site-selective CRISPR array expansion during the initial step of bacterial adaptive immunity.²⁴ Cas4 is an endonuclease defining the PAM and assisting the insertion of spacer into the CRISPR array unidirectionally.^23,41 Many CRISP-Cas systems lack Cas4 and some hosts use alternative exonucleases to acquire new CRISPR immune sequences.²¹ Alternatively, parts of hosts encode a solo-Cas4 outside the CRISPR-Cas loci⁴², but due to the incompleteness of the metagenomics, we could not exclude this possibility in the current study. In the evolution system, loss-of-function in certain components will drive the gain-of-function in other components to keep the robustness of the whole system. Integrase-deficient Cas12a variants can all achieved dsDNA degration, (Fig. 5ab and Extend Data Fig. 7ab). However, it remains unclear whether the absence of either ancillary integrase necessitates alternative genes in the genome, collaboration with integrases from other CRISPR-Cas systems, loss of function in immunological memory acquisition, or gain of function in Cas12a proteins. Notably, the diverse CRISPR-Cas variants within the same family highlight the intricacies of evolution, warranting further studies in the future.

The novel Cas12 variants showed distinct biochemical properties. They can bind to the crRNA and DNA but with different affinities (Fig. 4b, c). All the Cas12a variants cleave the dsDNA in a temperature dependent manner. Divalent metal ions play key roles in the cleavage and conformational rearrangements of CRISPR-Cas12a.⁴³ Despite Mg²⁺ ions, Mn²⁺ is able to activates some cas12a ^32,44. EvCas12a_2, AmCas12a, RspCas12a_2, RbrCas12a_1, and CAGCas12a prefer Mn²⁺ in the ssDNA cleavage. In addition, Co²⁺ can activate RspCas12a_2 better than Mg²⁺ (Fig. 5d). The ssDNA cleavage by Cas12a has been successfully applied in nucleic acid detection.^45,46 Nevertheless, Cas12a proteins are restrained by the PAM sequence⁴⁷, therefore, novel Cas12a with different PAM preference are required. In our work, we have systematically evaluated the PAM preference of the novel Cas12a proteins (Fig. 4). We found that AmCas12a recognizes broader PAM. We took the SNP of the oncogene KRAS as the target, which can not be detected by the traditional Cas12a due to the lack of PAM sequences near the mutation site. After optimization, the AmCas12a can specifically distinguish the KRAS G12C mutations, but LbCas12a can not. These findings extend the toolbox of the Cas12a detection.

The structure of crRNA is key for the conformational change of Cas12a. The stem-loop region and seed region are stable, but the spacer region in an apo form is largely elusive. Our cryo-EM structure discovered a novel folding of the crRNA spacer region, which forms a stem (Fig. 6b). This structure demonstrates the structural tolerance of the AmCas12a for crRNA flexibility. In the crRNA binary complex, Cas12a interacts with the pseudoknot of crRNA with conserved residues. However, due to the extra RNA stem, AmCas12a interacts with the spacer region by WED domain instead of REC1 and WED domains.

In summary, we developed an artificial intelligent language model for Cas proteins prediction. The increasing parameters in large data model can enhance the accuracy of predicting Cas proteins especially for the short protein sequences. This feature contributes to the novel CRISPR loci discovery and analyses of Cas protein evolution. Importantly, some novel Cas12 proteins have broader PAM recognition patterns and can be developed into efficient genome editors in mammalian cells or specific SNP detection kits. These findings will substantially increase the diversity of CRISPR-Cas12a systems and largely expand the programmable DNA-editing toolbox. This study shows the great potential of the language model with tremendous parameters in protein function exploration. Our study provides new insights into machine learning on the natural evolution of the CRISPR system and detailed characterization will discover more valuable gene editing tools.

Training data for language models

We generated sequences from NCBI databases to train ESM-2 650 million and 15 billion language models. We first use the keyword ‘CRISPR-associated protein’ to download all the gene IDs and then analyze the annotation 'gene' with ‘cas'. We further removed redundant sequences. We finally collected 77684 Cas protein sequences, including 11248 Cas1, 15148 Cas2, 12309 Cas3, 7708 Cas4, 8656 Cas5, 11281 Cas6, 340 Cas7, 299 Cas8, 6706 Cas9, 3525 Cas10, 334 Cas12, 130 Cas13 and 13047 non-Cas proteins. We split these sequences into two datasets, 80% as training data and 20% as validation data.

Training ESM language models

We performed fine-tuning on the open-source sequence classifier provided by ESM to adapt it to our application. The model consists of two fully connected layers, which are specifically designed for classification tasks. In the first layer of the model, the fully connected layer applies a linear transformation to the output features of the ESM model, mapping them to the same dimensional space, followed by a hyperbolic tangent (Tanh) activation function to introduce non-linearity. The second fully connected layer then projects the processed features to the dimension of target classes. This model structure effectively combines the excellent feature extraction ability of ESM and the efficient classification performance of fully connected neural networks, achieving effective classification of sequences. This model design maintains the high-dimensional sequence features while effectively learning the classification task.

To further improve training efficiency and model performance, we employ the DeepSpeed training acceleration framework, particularly its ZeRO-3 offload feature, to optimize memory utilization and accelerate the training process. AdamW is employed as the optimizer for its weight decay and momentum feature to enhance the training stability and efficiency. Meanwhile, WarmupLR is adopted as the learning rate scheduler, which gradually increases the learning rate in the early stages of training to facilitate model convergence. FocalLoss is used as the training loss function, which adjusts the weights of positive and negative samples to mitigate the class imbalance problem. The hyperparameter 'a' of FocalLoss is determined by considering the ratio of class sizes in the Cas data. The model architecture and training are implemented in PyTorch, ensuring code modernity and efficiency.

The training was conducted on the Zhejiang Lab Alkaid Intelligent Computing Operating System using NVIDIA Volta A100 GPUs.

Protein Expression and Purification

The candidate discovered from Cas12a was expressed and purified as previously described⁴⁸. In brief, the coding sequences of Cas proteins were codon-optimized and synthesized by Tsingke Biotech. (China) and then cloned into pET28a (Novagen) with a C-terminal 10×His tag. The pET28a-Cas12a plasmid was transformed into E. coli Rosetta and induced with 0.2 mM IPTG for 16 h at 18 °C before the cell harvesting. After cell pellets lysis, the Cas12a protein was purified using a Ni-NTA resin column and Heparin Sepharose column according to the manufacturer's instructions (GE Healthcare). Then the purified Cas12a protein was concentrated in storage buffer (50 mM Tris-HCl, pH 7.5, 500 mM NaCl, 10% (v/v) glycerol, 2 mM DTT), quantitated using the absorption at 280 nm, and frozen at -80 °C until use.

Nucleic acid preparation

The double strand DNA fragment of Cas12a variants were synthesized by Tsingke Biotech. (Nanjing, China) and cloned into the pUC57 vector with a T7 primer. The crRNAs were synthesized by GenScript (Nanjing, China), and sequences are listed in Table S3.

Cas12a-mediated nucleic acid detection

The detection assays were performed according to previously reported with minor modifications⁴⁸. In a 20 μL detection assay, with 200 ng Cas12a protein, 25 pM ssDNA FQ probe sensor, 50 nM crRNA and 10 ng of target dsDNA in a reaction buffer (100 mM NaCl, 50 mM Tris-HCl, 100 µg/mL BSA, pH 7.9) supplied with 10 mM MgCl₂or MnSO₄, incubate at 37 °C until detection. A PerkinElmer EnSpire reader with excitation at 485 nm and emission at 520 nm was used for fluorescence detection. For the divalent ion preference screen. The metal ion preference assay was performed as previously described³². In brief, the CRISPR-Cas12 detection assay was supplemented with 10 mM CaCl₂, CoCl₂, CuSO₄, NiSO₄, MgSO₄, MnSO₄, or ZnSO₄.

In vitro RNA and DNA binding assays

For RNA binding assays, Cas12a (100 nM) was incubated with Cy3-DNA (10 nM) at room temperature for 10 min in the reaction buffer. The reaction was quenched with glycerol loading buffer (10 mM Tris-HCl [pH 8.0], 10% glycerol). Reaction products were resolved by 12% PAGE and visualized by phosphorimaging (GE Health Care).

For DNA binding assays, Cas12a protein was first complexed with crRNA at a 1:2 ratio at room temperature for 10 min in reaction buffer. Cas12a complex (100 nM) was incubated with annealed FAM-DNA (25 nM) for 10 min at room temperature. The reaction was quenched with glycerol loading buffer (10 mM Tris-HCl [pH 8.0], 10% glycerol). Reaction products were resolved by 12% PAGE and visualized by phosphorimaging (GE Health Care).

PAM preference assay

The six short dsDNA target arrays were constructed by annealing 256 kinds of PAM sequence primer pairs in each well, which target EMX1 site1, DNMT1 site1, FANCF site1, MerS site1, eGFP site1 and eGFP site 3 (Table S4), Next, same as nucleic acid detection, Cas12a protein (200 ng), ssDNA FQ probe sensor (25 pM), crRNA (50 nM) and short target dsDNA (8.5 nM) were mixed in a reaction buffer supplied with 10 mM MgCl₂ (EvCas12a_2 and RspCas12a_2) or MnSO₄ (AmCas12a, CAGCas12a and RbrCas12a_1), and incubated at 37 °C, Viia 7 Real-Time PCR system were used for fluorescence tracing. In each detection plate, triple repeats of dsDNA with TTTG PAM as the fluorescence correction.

Phylogenetic analysis

The phylogenetic tree of Fig2a was constructed by a dataset of 87 sequences, including 30 Cas9 proteins, 43 Cas12 proteins and 14 Cas13 proteins. And the tree of Fig2f was constructed by 300 Cas12a variants as well as FnCas12a, LbCas12a and AsCas12a. The sequences were aligned with MAFFT-linsi (v7.480)⁴⁹. A phylogenetic tree was constructed by FastTree⁵⁰ with default parameters. The phylogenetic tree is annotated by iTol⁵¹.

Targeted deep sequencing

HEK293T cells were from Cell Bank/Stem Cell Bank, Chinese Academy of Sciences, and cultured in Dulbecco's modified Eagle's medium (GIBCO) supplemented with 10% fetal calf serum (v/v) (Gemini) and 1% penicillin–streptomycin at 37 °C with 5% CO₂. For plasmid transfection, cells were in 24-well plates in three biological replicates and transfected with 1.2 μg plasmids (including 900 ng editor and 300 ng sgRNA) per well, when cells reached an approximate 70-90% confluency. Transfections were carried out with the aid of EZ Trans (Life-iLab; Cat. No.: AC04L091) reagent and according to the manufacturer′s protocols. Three days after transfection, cells were harvested for deep sequencing. Target sites were amplified from extracted genomic DNA using Phanta® Max Super-Fidelity DNA Polymerase (Vazyme). PCR products with different barcodes were pooled together for deep sequencing on the Illumina HiSeq X Ten platform (2 × 150 PE) by Annoroad Gene Technology (Beijing, China). Different experimental conditions were differentiated by bar codes and experimental repetitions were included in different pools. Sequencing reads were demultiplexed using AdapterRemoval (version 2.2.2), and the pair-end reads with 11 bp or more alignments were combined into a single consensus read. All processed reads were then mapped to the target sequences using the BWA-MEM algorithm (BWA v0.7.16). Indel frequency was calculated as: number of indel-containing reads/total mapped reads. The targets and primers used in this study are provided in Supplementary Table 5 and Supplementary Table 6.

Reconstruction of AmCas12a-crRNA complex

AmCas12a was expressed and purified as described above, but further purified by size exclusion column (Superdex 200 Increase 10/300, GE Healthcare) in SEC buffer1 (10 mM Tris-HCl, pH 7.5, 500 mM NaCl) for complex preparation. The sgRNA was diluted to 100 μM in refolding buffer (50 mM KCl, 5 mM MgCl₂) and refolded at 72 ℃ for 5 min. The AmCas12a-crRNA binary was reconstituted by incubating 25 μM AmCas12a and 30 μM crRNA for 30 min at room temperature in a total volume of 450 μL assembly buffer (10 mM Tris-HCl, pH 7.5, 500 mM NaCl, 10 mM MgCl₂). Subsequently, the mixture was purified by size exclusion column in SEC buffer2 (10 mM Tris-HCl, pH 7.5, 500 mM NaCl, 1 mM MgCl₂). The purified aliquots were concentrated at 2 mg/mL, flash frozen and stocked at -80 ℃.

Cryo-EM sample preparation and data collection

Sample vitrification was performed using a Vitrobot Mark IV (Thermo Fisher) operating at 4 ℃ and 100% humidity. A 4 μL sample was applied to holey amorphous nickel–titanium alloy foil (ANTA foil 1.2/1.3) that had been glow-discharged for 30 s. The grids were blotted for 4 s at a ‘blot force’ -2 by standard Vitrobot filter paper (Ted Pella) and were then plunge-frozen in liquid ethane. Cryo-EM data were collected on a Titan Krios electron microscope operated at 300 kV equipped and Falcon4 direct electron detector with Quantum energy filter using EPU. Micrographs were recorded in counting mode at a nominal magnification of 165000x, resulting in a physical pixel size of 0.74 Å per pixel. The defocus was set between -0.6 μm and -1.8 μm. The total exposure time of each movie stack led to a total accumulated dose of 46.73 electrons per Å² which fractionated into 32 frames. More parameters for data collection are shown in Supplementary information, Table S7.

Image processing and 3D reconstruction

The raw dose-fractionated image stacks were 2× Fourier binned, aligned, dose-weighted, and summed using MotionCor2.⁵² CTF-estimation, blob particle picking, 2D reference-free classification, initial model generation, final 3D refinement and local resolution estimation were performed in cryoSPARC.⁵³ The details of data processing were summarized in Supplementary information, Fig. S6 and Table S7.

Model building and refinement

The initial protein model was generated using AlphaFold2 and manually revised in UCSF-Chimera and Coot. ^26,54,55 The crRNA was manually built in Coot based on the cryo-EM density. The complete model was refined against the EM map by PHENIX in real space with secondary structure and geometry restraints.⁵⁶ The final model was validated in PHENIX software package. The structural validation details for the final model are summarized in Supplementary information, Table S7.

RPA and fluorescence detection

The RPA assay was performed with a GenDx ERA Kit (Suzhou GenDx Biotech, China). According to the instructions in the manual, the 50 µl RPA system contains 2 µl DNA template, 2.5 µl forward primer (10 µM), 2.5 µl reverse primer (10 µM), 10 µl ERA basic buffer, 20 µl reaction buffer, 2 µl activator, and supplementary ddH2O.Three microliters of the RPA reaction product were transferred to the Cas12a reaction. In a 20 μL Cas12a reaction, additional with 100 ng Cas12a protein, 25 pM ssDNA FQ probe sensor, 50 nM crRNA in a reaction buffer (100 mM NaCl, 50 mM Tris-HCl, 100 µg/mL BSA, pH 7.9) supplied with 2.5 mM MnSO₄, incubate at 37 °C until detection. A PerkinElmer EnSpire reader with excitation at 485 nm and emission at 520 nm was used for fluorescence detection.

Quantification and statistical analysis

All values in the text and figures are presented as mean ± SEM of independent experiments with given n sizes. For image analysis, images were collected from at least three independent experiments. Graphs were compiled and statistical analyses were performed with Prism software (GraphPad) and Excel. Statistical significance was evaluated with the two-tailed unpaired t-test when comparing two groups. Differences between more than two samples were calculated using a one-way analysis of variance (ANOVA). Statistical details, including sample sizes (n), are indicated in the figures and legends.

Data availability

Materials can be obtained from the corresponding authors upon request. The deep sequencing data was deposited in NCBI with the accession number (PRJNA1043844). The structural data was deposited in the RCSB Protein Data Bank with the accession number (PDB: 8KGF) and Electron Microscopy Data Bank with the accession number (EMDB: EMD-37219). The code was deposited in the GitHub (https://github.com/LUCA-BioTech/cas_classification).

Acknowledgments

We thank the Molecular and Cell Biology Core Facility (MCBCF) and the Molecular Imaging Core Facility (MICF) at the School of Life Science and Technology, ShanghaiTech University for providing technical support, Shuimu Biosciences for providing technical support on cryo-EM sample evaluation and screening, Dr. Yong-Xiang Gao at the Cryo-EM Center, University of Science and Technology of China for technical support on cryo-EM data collection of AmCas12a, the Research Center for Intelligent Computing Platforms at the Zhejiang lab for providing the computing resources. The study is supported by the National Science Foundation of China (22177073 P.M., 32161133022 X.Z.), the Shanghai Science and Technology Committee (23ZR1437600 P.M. ), Shanghai Municipal Science and Technology Major Project (23HC1400700 Y.Q. )，National Key R&D Program of China (2022YFB4501500，2022YFB4501504 Y.Y.), and Shanghai Frontiers Science Center of Degeneration and Regeneration in Skeletal System.

Conflict of Interest

The authors declare no competing interests.

Author Contributions

P.M., X.H., J.T., and X.Z. conceived and designed this project. Y.F., J.S. performed biochemical and cellular experiments. Z.L., Y.L., J.Y., Y.Y., Q.L., and J.T., preformed the computational experiments. J.Z., Y.Q., J.Z., and W.H., contributed in biochemical experiments and analysis. Y.F and X.Z. performed the structural analyses. J.Z and S.H. performed the bioinformatics analysis. Y.F., P.M., X.H., Q.L., L.W., and C.H. wrote and revised the manuscript with input from the other authors.

Hille F et al (2018) The Biology of CRISPR-Cas: Backward and Forward. Cell 172:1239–1259
Makarova KS et al (2020) Evolutionary classification of CRISPR–Cas systems: a burst of class 2 and derived variants. Nat Rev Microbiol 18:67–83
Pickar-Oliver A, Gersbach CA (2019) The next generation of CRISPR–Cas technologies and applications. Nat Rev Mol Cell Biol 20:490–507
Al-Shayeb B et al (2020) Clades of huge phages from across Earth’s ecosystems. Nature 578:425–431
Devoto AE et al (2019) Megaphages infect Prevotella and variants are widespread in gut microbiomes. Nat Microbiol 4:693–700
Pausch P et al (2020) CRISPR-CasΦ from huge phages is a hypercompact genome editor. Science 369:333–337
Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local alignment search tool. J Mol Biol 215:403–410
Eddy SR (1998) Profile hidden Markov models. Bioinformatics 14:755–763
Chai G, Yu M, Jiang L, Duan Y, Huang JHMMCAS (2019) A Web Tool for the Identification and Domain Annotations of CAS Proteins. IEEE/ACM Trans Comput Biol Bioinform 16:1313–1315
Abby SS, Néron B, Ménager H, Touchon M, Rocha EP (2014) MacSyFinder: a program to mine genomes for molecular systems with an application to CRISPR-Cas systems. PLoS ONE 9:e110726
Padilha VA et al (2021) Casboundary: automated definition of integral Cas cassettes. Bioinformatics 37:1352–1359
Yang S, Huang J, He B (2021) CASPredict: a web service for identifying Cas proteins. PeerJ 9:e11887
Russel J, Pinilla-Redondo R, Mayo-Muñoz D, Shah SA, Sørensen SJ, CRISPRCasTyper (2020) Automated Identification, Annotation, and Classification of CRISPR-Cas Loci. Crispr j 3:462–469
Rives A et al (2021) Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences. Proc Natl Acad Sci U S A 118
Chowdhury R et al (2022) Single-sequence protein structure prediction using a language model and deep learning. Nat Biotechnol 40:1617–1623
Elnaggar A et al (2022) ProtTrans: Toward Understanding the Language of Life Through Self-Supervised Learning. IEEE Trans Pattern Anal Mach Intell 44:7112–7127
Huang J et al (2023) Discovery of deaminase functions by structure-based protein clustering. Cell 186:3182–3195e3114
Lin Z et al (2023) Evolutionary-scale prediction of atomic-level protein structure with a language model. Science 379:1123–1130
Coelho LP et al (2022) Towards the biogeography of prokaryotic genes. Nature 601:252–256
Hyatt D et al (2010) Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinformatics 11:119
Wang JY et al (2023) Genome expansion by a CRISPR trimmer-integrase. Nature 618:855–861
Dhingra Y, Suresh SK, Juneja P, Sashital DG (2022) PAM binding ensures orientational integration during Cas4-Cas1-Cas2-mediated CRISPR adaptation. Mol Cell 82:4353–4367e4356
Hu C et al (2021) Mechanism for Cas4-assisted directional spacer acquisition in CRISPR-Cas. Nature 598:515–520
Wright AV et al (2017) Structures of the CRISPR genome integration complex. Science 357:1113–1118
Xiao Y, Ng S, Nam KH, Ke A (2017) How type II CRISPR-Cas establish immunity through Cas1-Cas2-mediated spacer integration. Nature 550:137–141
Jumper J et al (2021) Highly accurate protein structure prediction with AlphaFold. Nature 596:583–589
Kim TY, Shin M, Thi Yen H, L., Kim JS (2013) Crystal structure of Cas1 from Archaeoglobus fulgidus and characterization of its nucleolytic activity. Biochem Biophys Res Commun 441:720–725
Holm L (2022) Dali server: structural unification of protein families. Nucleic Acids Res 50:W210–w215
Nuñez JK et al (2014) Cas1-Cas2 complex formation mediates spacer acquisition during CRISPR-Cas adaptive immunity. Nat Struct Mol Biol 21:528–534
Wang J et al (2015) Structural and Mechanistic Basis of PAM-Dependent Spacer Acquisition in CRISPR-Cas Systems. Cell 163:840–853
Tang D et al (2021) A distinct structure of Cas1-Cas2 complex provides insights into the mechanism for the longer spacer acquisition in Pyrococcus furiosus. Int J Biol Macromol 183:379–386
Ma P, Meng Q, Sun B, Zhao B, Dang L, Zhong M, Liu S, Xu H, Mei H, Liu J, Chi T, Yang G, Liu M, Huang X, Wang X (2020) MeCas12a, a highly sensitive and specific system for COVID-19 detection. Adv Sci (Weinh), advs.202001300R202001302
Swarts DC, van der Oost J, Jinek M (2017) Structural Basis for Guide RNA Processing and Seed-Dependent DNA Targeting by CRISPR-Cas12a. Mol Cell 66:221–233
Dong D et al (2016) The crystal structure of Cpf1 in complex with CRISPR RNA. Nature 532:522–
Swarts DC, van der Oost J, Jinek M (2017) Structural Basis for Guide RNA Processing and Seed-Dependent DNA Targeting by CRISPR-Cas12a. Mol Cell 66:221–
Kim D et al (2016) Genome-wide analysis reveals specificities of Cpf1 endonucleases in human cells. Nat Biotechnol 34:863–868
Kleinstiver BP et al (2016) Genome-wide specificities of CRISPR-Cas Cpf1 nucleases in human cells. Nat Biotechnol 34:869–874
Kim HK et al (2017) In vivo high-throughput profiling of CRISPR-Cpf1 activity. Nat Methods 14:153–159
Makarova KS et al (2015) An updated evolutionary classification of CRISPR-Cas systems. Nat Rev Microbiol 13:722–736
Koonin EV, Makarova KS, Zhang F (2017) Diversity, classification and evolution of CRISPR-Cas systems. Curr Opin Microbiol 37:67–78
Shiimori M, Garrett SC, Graveley BR, Terns MP (2018) Cas4 Nucleases Define the PAM, Length, and Orientation of DNA Fragments Integrated at CRISPR Loci. Mol Cell 70:814–824e816
Hudaiberdiev S et al (2017) Phylogenomics of Cas4 family nucleases. BMC Evol Biol 17:232
Son H et al (2021) Mg(2+)-dependent conformational rearrangements of CRISPR-Cas12a R-loop complex are mandatory for complete double-stranded DNA cleavage. Proc Natl Acad Sci U S A 118
Sundaresan R, Parameshwaran HP, Yogesha SD, Keilbarth MW, Rajan R (2017) RNA-Independent DNA Cleavage Activities of Cas9 and Cas12a. Cell Rep 21:3728–3739
Li SY et al (2018) CRISPR-Cas12a-assisted nucleic acid detection. Cell Discov 4:20
Chen JS et al (2018) CRISPR-Cas12a target binding unleashes indiscriminate single-stranded DNase activity. Science 360:436–439
Zetsche B et al (2015) Cpf1 is a single RNA-guided endonuclease of a class 2 CRISPR-Cas system. Cell 163:759–771
Wang X et al (2020) CRISPR/Cas12a technology combined with immunochromatographic strips for portable detection of African swine fever virus. Commun Biol 3:62
Katoh K, Misawa K, Kuma K, Miyata T (2002) MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res 30:3059–3066
Price MN, Dehal PS, Arkin AP (2010) FastTree 2–approximately maximum-likelihood trees for large alignments. PLoS ONE 5:e9490
Letunic I, Bork P (2021) Interactive Tree Of Life (iTOL) v5: an online tool for phylogenetic tree display and annotation. Nucleic Acids Res 49:W293–w296
Zheng SQ et al (2017) MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nat Methods 14:331–332
Punjani A, Rubinstein JL, Fleet DJ, Brubaker MA (2017) cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nat Methods 14:290–296
Emsley P, Cowtan K (2004) Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60:2126–2132
Pettersen EF et al (2004) UCSF Chimera–a visualization system for exploratory research and analysis. J Comput Chem 25:1605–1612
Adams PD et al (2010) PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr D Biol Crystallogr 66:213–22

There is NO Competing Interest.

FengSI.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Novel CRISPR-Cas12a Clades Discovery Using Large Language Model

Status:

Version 1

Abstract

Figures

Introduction

Results

Development of trans-cleavage activity prediction model

New CRISPR-Cas12a loci predicted from the metagenomics

New Cas integrases in CRIPSR loci

The AmCas12a-crRNA binary complex

Specific detection of single nucleotide mutation by AmCas12a

Discussion

Methods

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1