1. Introduction

Clustered regularly interspaced short palindromic repeats (CRISPR) and their associated cas genes (CRISPR-Cas) are a family of DNA sequences, as an adaptive immune system, which protects microbial cells against infection by foreign nucleotide elements including plasmids and phages [1]. CRISPR are widespread in prokaryotes, and found in approximately 50% bacterial genomes including Salmonella belonging to the family of Enterobacteriaceae [2], which is a primary cause of bacterial foodborne illness worldwide.

Through a computational analysis of CRISPR-Cas systems, a classification system was determined based on the gene or genes encoding the effector molecules [3]. This analysis determined that CRISPR-Cas systems can fall into two classes; class 1 systems (types I, III and IV) which use a complex of multiple proteins to degrade foreign nucleic acids, and class 2 systems (types II, V, and VI) which only require a single large Cas protein (Figure 1) [4]. The six types of systems are further divided into 36 subtypes (Figure 1) [3, 5, 6, 7, 8, 9, 10]. Fully functional CRISPR-Cas systems consist of CRISPR array, Cas proteins and AT-rich leader sequences. The phylogeny of CRISPR and associated cas genes could reflect different evolutionary histories [11, 12]. The CRISPR array is composed of a succession of highly conserved direct repeats (DR) of 24–47 bp separated by similar sized unique sequences (spacers) [13]. The cas genes are usually located near the CRISPR locus but can also be located elsewhere on the genomes. Cas proteins perform many functions, for instance, destroying foreign genomes, mediating foreign sequences acquisition into CRISPR array, and assisting the mature CRISPR RNAs (crRNAs) production [14, 15, 16, 17]. CRISPR-Cas systems adapt by acquiring new spacers at the leader proximal end [1]. The units (DR+spacer) may target an invading piece of DNA and result in its degradation via a proposed mechanism similar to RNA interference. The distribution of CRISPR-Cas loci in different Enterobacteriaceae families showed that Eshcerichia and Salmonella are the top two genera containing type I-E subtypes [2]. CRISPR are reported in two pairs of loci in Escherichia, and one single pair in Salmonella, with each pair loci showing similar repeat sequences and putative linkage to common cas genes [11].

Figure 1.
General classification of CRISPR-Cas systems. Two classes—indicated by the red and blue colouring—cover six types. A total of 36 subtypes are further divided under the six types with the known signature proteins listed.

It has been shown that CRISPR spacer DNA sequences are molecular signatures used for pathogen subtyping [18] and CRISPR content correlates with the pathogenic potential of bacteria as CRISPR-Cas limits acquisition of foreign nucleotides in bacteria [19]. It has been demonstrated that there is a negative correlation between the amount of CRISPR units and pathogenicity traits, i.e. a higher number of virulence factors with lower CRIPSR repeats [19]. Based on the specific spacers, CRISPR array based quantitative PCR can be used to detect the presence of different serotypes in both Escherichia and Salmonella, with prominent sensitivity and specificity [20, 21]. Hence, the CRISPRs represent a promising genetic marker and diagnostic tool for comparative and evolutionary analysis of closely related bacterial strains [2]. Furthermore, the recognition of CRISPR-Cas9 by the Nobel Prize in Chemistry in 2020 [22] reflected its outstanding impact in genome editing field. Originated from bacteria, the CRISPR technology has already been broadly applied to fungus, yeast, insects, plants, and animals [23]. This technology has also demonstrated to functionally inactivate genes in human cell lines and cells. For instance, in 2019, CRISPR was used to treat a 34-year-old patient with sickle cell disease which is a blood genetic disorder disease [24]; and in 2020, CIRSPR-modified virus was injected into a patient’s eye to treat Leber congenital amaurosis [25]. In this chapter, we will mainly focus on a foodborne pathogen Salmonella which is a primary cause of bacterial gastroenteritis worldwide.

Salmonella enterica is a tremendously diverse species comprising six subspecies and over 2600 serovars. S. enterica subsp. enterica accounts for the majority of clinical cases of salmonellosis and the majority of serovar diversity. Serovars of Enteritidis, Typhimurium, and Heidelberg are three main ones causing human illness. This book chapter will mainly focus on CRISPR-Cas in the immune response system of Salmonella, as well as its application in genome editing, pathogen typing, diagnosis and surveillance.

2. CRISPR-Cas systems of Salmonella in comparison with other bacteria

2.1 Immune function of CRISPR-Cas systems

In prokaryotes, bacterial CRISPR-Cas systems are unique in providing adaptive immunity against exogenous nucleotides elements, by utilizing sequence-specific RNA-guided nucleases to defend against bacteriophage infection. Bacteriophages (phages) are viruses infecting bacteria, and they are the most abundant life forms on earth. Generally, three major steps are involved in the CRISPR immune functional process: (1) new spacer acquisition—Cas proteins integrate short sequences of invading DNA into the CRISPR array; (2) CRISPR expression—CRISPR arrays are transcribed and processed to produce small crRNA; (3) CRISPR interference—crRNA along with Cas nucleases target the spacer sequence, resulting in degradation of the invader’s nucleotides (DNA or RNA) [26, 27].

2.2 Characterization of CRISPR loci and cas genes

Like many other bacteria, the Salmonella genome also contains CRISPR loci. It usually contains two CRISPR loci, CRISPR1 and CRISPR2, both found on the minus strand. These two loci are separated by ~16 kb and share the same consensus direct repeat sequence (29 nt). Each CRISPR loci is fairly conserved in Salmonella, with the CRISPR1 locus being more conserved than CRISPR2. There are eight cas genes—cas3, cse1, cse2, cas7, cas5, cas6, cas1 and cas2, which are located upstream of CRISPR1. Among these eight genes, cas1 and cas2 are universal and both are present in all CRISPR-Cas systems; cas3 is a signature gene in the type I system; the remaining cas genes are type I-E dependent [28]. Furthermore, the cse2, cas5, cas6e, cas1, cas2 and cas3 genes are crucial for the expression of a master porin regulator named OmpR which is a two-component system regulator inducing the synthesis of OmpC, MmpF, and OmpS2 portins [29]. By taking advantages of whole genome sequencing (WGS), in 2014, researchers have demonstrated two distinct cas gene profiles and a high diversity of length for both CRISPR arrays, among the analysis of 64 Salmonella serovars [30].

2.3 CRISPR and anti-CRISPR

To combat bacterial CRISPR-Cas system, numerous phages are well known to produce proteins which can block the function of CRISPR-Cas systems, i.e. anti-CRISPR function [31]. For class 1 CRISPR system, the first discovered phage-encoded anti-CRISPR protein (Acr) was from type I-F and I-E CRISPR-Cas systems in Pseudomonas aeruginosa; these anti-proteins encode distinct, small proteins (50–150 aa) with different sequences and structures [32, 33]. Furthermore, these anti-CRISPRs are produced from prophages (phage sequences that have integrated into bacterial genomes) and inactivate the host (bacterial) CRISPR-Cas systems [32]. For class 2 CRISPR, four unique type II CRISPR-Cas9 inhibitor proteins have been discovered from the prophage sequences integrated into another foodborne pathogen Listeria monocytogenes genomes, which have type II-A CRISPR-Cas systems and their spacers have been identified by various virulent, temperate phages [34, 35]. Given more than half of L. monocytogenes strains with cas9 contain at least one prophage-encoded inhibitor, this suggests the possibility of widespread CRIPSR-Cas9 inactivation. Two of the discovered inhibitors in L. monocytogenes are also able to block the Streptococcus pyogenes Cas9 when analyzed in Escherichia coli and human cells. Similarly, in Streptococcus thermophiles, AcrIIA6 acts as an allosteric inhibitor and induced Cas9 dimerization [36]. Thus, the concept of natural Cas9-specific “anti-CRISPRs” presents a tool which can be used to regulate the genome engineering activities of CRISPR-Cas9 [31]. Similar to L. monocytogenes, in different Salmonella serovars, they all contain various types of prophage sequences [37]. To date there is no reported anti-CRISPR proteins in Salmonella, though this could change as more studies are carried out.

3. Identification and characterization of CRISPR arrays

Next generation sequencing (NGS) and especially WGS has emerged in recent years and has made it possible to sequence bacterial genomes within hours, a notable accomplishment that is revolutionizing the field of microbiology [38]. With the advent of microbial WGS, new light is shed on the nature of pathogens, for instance CRISPRs, and our understanding of the biology of Salmonella is steadily increasing as Salmonella genomes are generated at a rapid rate and are deposited in public database such as National Center for Biotechnology Information (NCBI). Based on the availability of genome sequences, various genomic and bioinformatics tools have been developed for identifying the potential CRISPR arrays in Salmonella genomes.

3.1 In silico genomics based CRISPRs identification tools

An example of how the field of CRISPRs has evolved can be seen in the work done with in silico analysis. In silico identification and analyses of CRISPRs started in 1995 [39], and several CRISPR software tools have been developed since then. In April 2007, the first specific stand-alone developed tool was CRISPRFinder, which was a web tool in identifying CRISPRs [40]. CRISPRFinder was able to define DRs and extract spacers; to get the flanking sequences and to determine the leader sequences; and then BLAST the identified spacers to check if the identified DR was present in other genomes [40]. Two months later in June 2007, in order to dissect and understand CRISPR structure and flanking sequences evolution, the same group created a public database named CRISPRdb, for which CRISPRFinder was used to analyze all the available prokaryotic genomes [41]. In the same month June 2007, a tool named CRISPR Recognition Tool (CRT) for automatic detection of the CRISPR arrays was also released [42]. CRT was demonstrated to be very reliable, with significant improvements in regards to performance in measures of precision, recall and quality, as compared to the previous existed detection tools Patscan and Pilercr [42]. In April 2008, a website based tool CRISPRcompar was created to compare CRISPRs that present a useful genetic marker for comparative analysis of closely related bacterial strains; this facilitated the development of CRISPR based pathogen typing processes [43]. More CRISPR-Cas related online tools can be found in CRISPR-Cas++, which are available at https://crisprcas.i2bc.paris-saclay.fr/. In 2018, an improved CRISPRs identification tool CRISPRCasViewer was released, which can predict CRISPR orientation, possess the latest classification scheme, and facilitate expert validation based on a rating system [44]. Alternatively, the public available “standalone” Unix/Linux version of CRISPRCasViewer can also be downloaded and installed in high-performance computing cluster bioinformatics infrastructures (https://github.com/dcouvin/CRISPRCasViewer). Thus, with the availability of all the genomic tools, the CRISPRs and cas genes present in each Salmonella isolate are able to be detected. Subsequently, comparative and evolutionary analysis can also be carried out to identify potential genetic markers, which will be useful for diagnosis and surveillance tools development in food safety.

Typically, the identified CRISPR arrays are represented by colored shapes based on nucleotide sequence identity. For facilitating and easy handling this process, Dion et al. [45] have introduced CRISPRStudio which is a user-friendly command-line tool to accelerate CRISPR analysis and standardize CRISPR array figures preparation. CRISPRStudio is able to compare nucleotide spacer sequences and then cluster them based on sequence similarity to assign a representative color; it also supports automatic sorting of CRISPR loci and highlighting shard spacers [45].

3.2 CRISPR target

In bacterial and archaeal adaptive immune systems, CRISPR-Cas targets specific protospacer nucleotide sequences in invading organisms, which requires nucleotide base pairing between processed crRNAs and target protospacer. Biswas et al. [46] have developed a flexible, interactive tool CRISPRTarget for the discovery of the target of crRNAs in diverse database. CRISPRTarget is available at http://crispr.otago.ac.nz/CRISPRTarget/crispr_analysis.html, it can be used to discover targets from both genomic and metagenomics dataset in many pathogens, including the foodborne pathogen Salmonella.

3.3 Conservation and diversity of Salmonella CRISPR arrays

Similar to other genetic components, CRISPR sequences can be conserved throughout a pathogen family. Through genomic sequence analyses of four clinically relevant Salmonella serovars; Enteritidis, Typhimurium, Newport and Heidelberg, it was determined that both cas operons and leaders are conserved among these four serovars [28]. Furthermore, Salmonella seems to be lacking in spacer acquisition, and the majority of CRISPR allelic polymorphisms usually arise from deletion or duplication of direct repeat-spacer units [47, 48, 49].

With the development of NGS technology, more and more Salmonella isolates have complete genome sequences available. In order to eliminate the potential bias caused by incomplete genome sequences, a collection of 115 representative Salmonella isolates with complete genomes (size range 4,482,117–5,395,280 bp) were analyzed in this chapter (Table 1). Those selected isolates come from four different subspecies, with the subspecies enterica as the dominant one. For the isolates within these four subspecies, a total of 90 different S. enterica serovars were included in this analysis, the details of each isolate can be found in Table 1.

ID	Subsp.	Serovar	CRISPR	Strain	Size(bp)
CP054422.1	diarizonae	61:k:1,5,(7)	1	14-SA00836-0	4,832,352
CP006602.1	enterica	4,[5],12:i:-	1,2	08-1736	4,822,189
CP034711.1	enterica	43:a:1,7	1,2	RSE20	4,665,063
CP007532.1	enterica	Abaetetuba	1,2	ATCC35640	4,547,600
CP007534.1	enterica	Abony	1	str.0014	4,737,447
CP049880.1	enterica	Adjame	1,2	381330	4,678,052
CP001138.1	enterica	Agona	1,2	SL483	4,798,660
CP019177.1	enterica	Albany	1,2	ATCC51960	4,805,448
CP007531.1	enterica	Anatum	1,2	ATCCBAA-1592	4,706,101
CP019116.1	enterica	Antsalova	1,2	S01-0511	4,648,086
CP019403.1	enterica	Apapa	1,2	SA20060561	4,801,658
CP019405.1	enterica	Bergen	1,2	ST350	4,801,835
CP030005.1	enterica	Berta	1,2	SA20141895	4,725,468
CP019406.1	enterica	Blegdam	1,2	S-1824	4,693,979
CP019407.1	enterica	Borreze	1,2	SA20041063	4,777,558
HF969015.2	enterica	Bovismorbificans	1,2	3114	4,680,283
CP022490.1	enterica	Braenderup	1,2	SA20026289	4,734,880
CP030002.1	enterica	Brandenburg	1	SA20064858	4,677,648
CP007533.1	enterica	Bredeney	1,2	CFSAN001080	4,603,849
CP012833.1	enterica	Cerro	1,2	CFSAN001588	4,651,400
AE017220.1	enterica	Choleraesuis	1,2	SC-B67	4,755,700
CP027677.1	enterica	Corvallisain	1,2	12-01738	4,887,378
CP019408.1	enterica	Crossness	1,2	1422-74	4,847,468
CP006055.1	enterica	Cubana	2	CFSAN002050	4,977,480
CP022494.1	enterica	Derby	1,2	SA20035215	4,850,334
CP019409.1	enterica	Djakarta	1,2	S-1087	4,668,861
CP001144.1	enterica	Dublin	1,2	CT_02021853	4,842,908
NC_011294.1	enterica	Enteritidis	1,2	P125109	4,685,848
CP032444.1	enterica	Fresnoain	1,2	USMARC-69835	4,732,430
NC_011274.1	enterica	Gallinarum	1,2	287/91	4,658,697
*CP024165.1	enterica	Gaminara	1,2,3	CFSAN070644	4,801,841
CP017719.1	enterica	Hayindogo	1,2	CFSAN050752	4,765,719
CP001120.1	enterica	Heidelberg	1,2	SL476	4,888,768
CP019410.1	enterica	Hillingdon	1,2	N1529-D3	4,618,056
CP022503.1	enterica	Hvittingfoss	1	SA20014981	4,940,239
CP022015.1	enterica	India	1	SA20085604	5,395,280
CP022450.1	enterica	Indiana	1,2	D90	4,779,514
LN649235.1	enterica	Infantis	1,2	SINFA	4,710,675
CP019181.1	enterica	Inverness	1,2	ATCC10720	4,865,682
LT571437.1	enterica	Javaain	1	NCTC5706	4,756,780
CP004027.1	enterica	Javiana	1,2	CFSAN001992	4,634,161
CP019411.1	enterica	Johannesburg	1,2	ST203	4,651,794
CP034709.1	enterica	Karamoja	1,2	RSE21	4,764,896
CP022500.1	enterica	Kentucky	1,2	SA20030505	4,782,363
CP019412.1	enterica	Koessen	1	S-1501	4,566,169
CP019413.1	enterica	Krefeld	1,2	SA20030536	4,942,273
CP032817.1	enterica	Lubbock	1,2	11TTU1590	4,985,874
CP032814.1	enterica	Lubbock	1,2	10TTU468x	4,985,863
CP022117.1	enterica	Macclesfield	1,2	S-1643	4,822,139
CP019414.1	enterica	Manchester	1,2	ST278	4,532,753
CP022497.1	enterica	Manhattan	1	SA20084699	4,732,484
**CP019183.1	enterica	Mbandaka	1,2	ATCC51958	4,905,181
CP022489.1	enterica	Mbandaka	1,2	SA20026234	4,796,292
CP034713.1	enterica	Mikawasima	1,2	RSE15	4,650,494
CP030175.1	enterica	Milwaukee	1,2	SA19950795	4,822,474
CP019184.1	enterica	Minnesota	1,2	ATCC49284	4,592,393
CP034705.1	enterica	Moeroain	1,2	RSE29	4,582,521
CP007530.1	enterica	Montevideo	1,2	507440-20	4,694,375
CP019415.1	enterica	Moscow	1,2	S-1843	4,690,402
CP045056.1	enterica	Muenchenain	1	LG24	4,930,424
CP019201.1	enterica	Muenster	1,2	CFSAN001301	4,756,014
CP022663.1	enterica	NA	1,2	RM11065	4,991,140
CP022658.1	enterica	NA	1	RM11060	4,892,239
CP033348.2	enterica	NA	1,2	CFSA1096	4,696,663
NC_011080.1	enterica	Newport	1,2	SL254	4,827,641
CP019416.1	enterica	Nitra	1,2	S-1687	4,691,807
CP022034.1	enterica	Onderstepoort	1,2	SA20060086	4,774,926
CP033344.1	enterica	Oranienburg	2	CFSAN076211	4,651,134
CP022116.1	enterica	Ouakam	1,2	SA20034636	4,874,915
CP012346.1	enterica	Panama	1	ATCC7378	4,555,576
CP000026.1	enterica	Paratyphi A	1,2	ATCC9150	4,585,229
CP000886.1	enterica	Paratyphi B	1	SPB7	4,858,887
CP000857.1	enterica	Paratyphi C	1,2	RKS4594	4,833,080
CP019186.1	enterica	Pomona	1	ATCC10729	4,482,117
CP019189.1	enterica	Poona	1,2	ATCCBAA-1673	4,876,720
CP012347.1	enterica	Pullorum	1,2	ATCC9120	4,694,842
CP022491.1	enterica	Saintpaul	1,2	SA20031783	4,775,303
CP001127.1	enterica	Schwarzengrund	2	CVM19633	4,709,075
CP029038.1	enterica	Senftenberg	1,2	CFSAN045763	4,766,139
CP012349.1	enterica	Sloterdijk	1,2	ATCC15791	4,817,791
CP017723.1	enterica	Stanleyville	1,2	CFSAN000624	4,888,463
CP007505.1	enterica	Tennessee	1,2	TXSC_TXSC08-19	4,864,410
CP006717.1	enterica	Thompson	1,2	RM6836	4,707,648
NC_003198.1	enterica	Typhi	1	CT18	4,809,037
HF937208.1	enterica	Typhimurium	2	DT104	4,933,631
NC_003197.2	enterica	Typhimurium	1,2	LT2	4,857,450
CP006048.1	enterica	Typhimurium var. 5	1,2	CFSAN001921	4,859,931
CP019417.1	enterica	Wandsworth	1,2	SA20092095	4,916,040
CP022138.1	enterica	Waycross	1,2	SA20041608	4,812,886
LN890520.1	enterica	Weltevreden,	1,2	C2346	5,129,845
CP029041.1	enterica	Worthington	1,2	CFSAN051295	4,914,635
CP019418.1	enterica	Yovokome	1	S-1850	4,640,929
CM001471.1	houtenae	NA	1	ATCCBAA-1581	4,672,567
CP030181.1	NA	NA	1,2	SA20030575	4,772,343
CP030185.1	NA	NA	1,2	SA20094620	4,854,398
CP030190.1	NA	NA	1,2	SA20104250	4,813,547
CP030196.1	NA	NA	1,2	SA20051401	4,869,528
CP030202.1	NA	NA	1,2	SA20052327	4,763,586
CP030203.1	NA	NA	1,2	SA20083530	5,062,813
CP030207.1	NA	NA	1,2	SA19992307	4,844,554
CP030209.1	NA	NA	1,2	SA20044414	4,805,225
CP030211.1	NA	NA	1,2	SA20051528	4,719,399
CP030214.1	NA	NA	1,2	SA20025921	4,882,461
CP030217.1	NA	NA	1,2	SA20075157	4,716,530
CP030223.1	NA	NA	1	SA20083039	4,688,830
CP030225.1	NA	NA	1,2	SA20041605	4,739,617
CP030231.1	NA	NA	1,2	SA20043041	4,603,878
CP030233.1	NA	NA	2	SA20101045	4,729,786
CP030235.1	NA	NA	1,2	SA20031245	4,522,338
CP030236.1	NA	NA	1,2	SA20041606	4,524,637
CP030238.1	NA	NA	1	SA20055162	4,640,729
CP022139.1	salamae	55:k:z39	1,2	1315K	4,859,044
*CP029992.1	salamae	56:z10:e,n,x	1,2,3	SA20011914	4,807,680
*CP029995.1	salamae	56:b:[1, 5]	1,2,3	SA20053897	4,920,300
*CP034717.1	salamae	42:r:-	1,2,3	RSE09	4,860,626

Table 1.

Representative 115 Salmonella enterica isolates with complete genomes containing four known subspecies covering 90 serovars used for CRISPR arrays analysis in this study.

*

These four Salmoenlla enterica isolates contain three CRISPR arrays, which is different from the common ones containing two.

**

This isolate has the longest CRISPR2 array.

Briefly, all available Salmonella complete genomes were downloaded from the NCBI database using Bioinformatics Tools (bit) (https://github.com/AstrobioMike/bit#bioinformatics-tools-bit) from GitHub and NCBI EDirect tools (https://astrobiomike.github.io/unix/ncbi_eutils). By applying common NCBI BLAST keywords, the used commands for downloading those complete genomes were:

“esearch -db assembly -query ' ("Salmonella"[Organism] OR Salmonella[All Fields]) AND (latest[filter] AND "complete genome"[filter] AND all[filter] NOT anomalous[filter])' | esummary | xtract -pattern DocumentSummary -element AssemblyAccession > Salmonella_complete_genome.txt”;

“bit-dl-ncbi-assemblies -w Salmonella_complete_genome.txt -f fasta -j 12”.

Then a total of 115 representative genomes were manually selected and compiled by including as many serovars as possible.

To identify CRISPR arrays in those 115 representative Salmonella isolates, two main used software were CRISPRDetect_2.2 (https://github.com/ambarishbiswas/CRISPRDetect_2.2) [50] and CRISPR_Studio (https://github.com/moineaulab/CRISPRStudio) [45]. Detailed procedures were described in the above two github links. Briefly, a specific python3 conda environment was created for this project, the used command for CRISPRDetect was: “perl ../bin/CRISPRDetect_2.2/CRISPRDetect.pl -f interested.fasta -o output_file -array_quality_score_cutoff 3 -T 0”. Subsequently, the CRISPRDetect produced “output_file” containing detected CRISPR arrays was fed into and visualized using CRISPR_Studio, and the used command was: “python CRISPR_Studio_1.0.py -i ../CRISPRDetect/output_file”, with Figure 2 presented in this chapter as final outputs.

Figure 2.
Graphic representation of spacer profiles in three arrays of CRISPR1, CRISPR2 and CRIPSR3, detected from 115 Salmonella enterica isolates with complete genomes consisting of four known subspecies covering 90 serovars. The figure was created by CRISPRStudio. Each spacer is represented by a colored square and a geometric symbol. The earliest acquired spacer is shown on the right hand side and the newly acquired space is on the left hand side. Specifically, the four isolates containing CRISPR3 are: three S. enterica subsp. salamae isolates, and one S. enterica subsp. enterica serovar Gaminara, which are indicated as “*” in the Table 1. The identical CRISPR spacer profiles are grouped and indicated by red, blue, green, and orange dots.

Among the analyzed 115 Salmonella isolates, prominent diversity was observed in the detected CRIPSR array profiles. Unlike commonly reported knowledge that Salmonella usually contains two CRISPR loci [28], there were four isolates containing the 3^rd loci, CRISPR3. Three of these isolates belonged to S. enterica subsp. salamae and the last one belonged to S. enterica subsp. enterica serovar Gaminara (Table 1 and Figure 2). Additionally, there were five isolates that only contained CRISPR2 and 18 isolates that only had CRISPR1 (Table 1 and Figure 2). Although prominent diversity was observed among isolates, respective identical CRISPR spacer profiles were observed for four groups. The first group (indicated by red dots) included a total of two serovars from the subspecies enterica, namely Nitra (CP019416) and Enteritidis (NC_011294.1); these two serovars showed high similarities and are known to be difficult to distinguish in nature using different microbiological methods. In the second group (indicated by blue dots), one was S. enterica subsp. enterica serovar Dublin (CP001144.1), and the other one is unknown serovar from the same subspecies. The third group (indicated by green dots) consisted of three serovars of S. enterica subsp. enterica, namely India (CP022015.1), Panama (CP012346.1), and Koessen (CP019412.1). The last group (indicated by orange dots) has two serovars of Yovokome (CP019418.1) and Manhattan (CP022497.1) belonging to S. enterica subsp. enterica (Figure 2). The discovered CRIPSR arrays with certain similarities or dissimilarities might shed light on the phylogeny and evolutionary analysis of Salmonella isolates in the future.

4. Major CRISPR applications of genome editing, AMR patterns and typing tools in Salmonella and other microorganisms

CRISPR-Cas originates from bacteria and has also been broadly applied back in functionally studying bacteria. Here, CRISPR/Cas9 genome editing used in Salmonella host-pathogen interaction will be described, followed by the CRISPR-Cas diversity and its strong correlation with antimicrobial resistance (AMR) pattern studies will be introduced. The emerging application of CRISPR typing/subtyping will be explained. Finally, more advanced CRISPR-Cas related diagnosis and surveillance methods related to other microorganisms will also be demonstrated, as similar methods could be used as potential alternative methods for studying foodborne pathogens including Salmonella.

4.1 CRISPR/Cas9 genome editing in Salmonella host-pathogen interaction studies

The discovery of Cas9 has allowed for impressive advances in the field of genome editing. This protein can be utilized to modify the genome of interest, based on the segments in the CRISPR array. Cas9 endonuclease activity needs crRNAs to guarantee precise targeting, and an immediate downstream protospacer adjacent motif (PAM). With the aim of editing bacterial genomes, a vector encoding Cas9 and its guide RNAs, as well as recombination template containing required mutation are required [51]. For preventing the re-cleavage of Cas9 of the target genome, the spacer of PAM sequences will need to be modified [52, 53]. Using such approach, mutations have been introduced into the sdiA gene to study its effect on S. enterica pathogenesis. The introduced mutations affected S. enterica biofilm formation, cell adhesion and invasion [54]. CRISPR/Cas9 has also been used in generating macrophage knockout mice cell lines, which facilitates S. Typhimurium infection studies by determining the contribution of background contaminations in the phenotypes of primary macrophages from congenic mice [55]. It has also demonstrated that the CRISPR-Cas system is involved in the resistance to bile salts and biofilm formation in S. Typhi [29]. This demonstrated CRISPR/Cas9 genome editing based methods contribute significantly in carrying out functional studies of Salmonella.

4.2 CRISPR/Cas diversity and its strong correlation with AMR pattern

AMR is a global concern for human health and a World health organization global priority (https://www.who.int/news-room/fact-sheets/detail/antimicrobial-resistance). WGS is replacing traditional phenotypic method such as disk diffusion method for routine testing of foodborne pathogens AMR. The tools of ResFinder [56, 57] or the Comprehensive Antibiotic Resistance Database (CARD) [58] detect the presence of AMR genes in an isolate by comparing its sequence against known genes cataloged in a reference database of known AMR determinants. Although knowledge of the CRISPR-Cas systems has been applied in many research areas, there are not many studies in applying it to the analysis of antibiotic resistance in Salmonella. Recently, by using large-scale bioinformatics investigation of the 1059 isolates of S. Typhi CRISPR-Cas systems, 47 unique spacers and 15 unique DRs were identified, as well as unique conservation and clonality of the S. Typhi type I-E CRISPR-Cas system was observed [59]. The identified spacers and repeats showed specific patterns which demonstrated significant associations with AMR status, genotype, and demographic characteristics. This suggests they have the potential to be used as biomarkers to develop rapid and inexpensive diagnostics tests [59]. Similarly, on Chinese poultry farms, analysis of 75 Salmonella isolates consisting of 11 serovars, found that there were close correlations between CRISPR loci and AMRs, however, there was no close correlations between CRISPR loci and antibiotics [60].

4.3 CRISPR typing and subtyping as improved laboratory diagnostic tool in Salmonella

Various molecular and phenotypic typing techniques have been developed to track bacterial origins, for instance, pulse-field gel electrophoresis (PFGE), phage typing, multi-locus sequence typing (MLST), multi-locus variable number tandem repeat (MLVA) and single nucleotide polymorphism (SNP) pipelines [61]. The above mentioned typing methods are limited in both speed and precision. In recent years, improved and innovative surveillance tools of CRISPR typing have been developed, which are used to gain knowledge in better understanding a variety of bacteria, such as Salmonella [47]. Serving as a complementary tool for the high-resolution core genome single nucleotide variant (cgSNV) method, CRISPR typing was useful for determining source attribution in foodborne S. Heidelberg outbreaks [62, 63]. CRISPR typing was also shown to facilitate further studies in understanding the virulence and global distribution of the S. Virchow serovar [64]. Furthermore, the combination of both MLVA and CRISPR (CRISPR-MLVA method) gave better genotyping results than using each one alone, when testing 171 Salmonella strains from nine serovars [48]. There are limitations to this method of typing, particularly in very closely related isolates. In these instances, it has been shown that using CRISPR typing in conjunction with a SNP analysis allows for better resolution, indicating the use of CRISPR typing still exhibits clear benefits [65]. A few CRISPR based typing tools are illustrated in details as below.

4.3.1 Conventional CRISPR typing

In conventional CRISPR typing (CCT), all spacer sequences in the two loci of CRISPR1 and CRISPR2 are extracted [48]. Then CRISPR1 and CRISPR2 spacer sequences profiles are analyzed and visualized using CRIPSRviz [66]. There are three main procedures: (1) CRISPR arrays are obtained by either directed whole-genome sequencing or PCR amplification of CRISPR loci using conservative sequences following by sequencing; (2) the identification and characterization of potential CRISPR arrays based on the previous sequencing results; (3) finally, clustering of analyzed isolates based on the absence or presence of analyzed CRISPR arrays.

4.3.2 CRISPR locus spacer pair typing

CCT can be labor intensive to carry out. To increase the easy of typing, CRISPR locus spacer pair typing (CLSPT) was developed [67]. Instead of using all the obtained spacer sequences, only one spacer sequence in both CRISPR1 and CRISPR2 loci will be used for typing. This spacer sequence is the first one, found closest to the leader sequences. In this method, the first spacer sequence of the CRISPR1 leader sequences is combined with the first spacer sequence of the CRISPR2 leader sequences. Then these two spacer sequences were used as the total sequences for Salmonella strain typing.

4.3.3 CRISPR locus three spacer sequences typing

Usually, during the evolution of bacterial strains which contain CRISPR arrays, the first captured exogenous nucleotide sequence could display strain origins, and the spacers from the same serotype possess certain conservation. Thus, Li et al. have developed a Salmonella typing method, called CRISPR locus three spacer sequences typing (CLTSST) method, which could be used to distinguish different serotype clusters. They used three spacer sequences including the initial two spacer sequences (the first acquisitions or the ones with the furthest distance to the leader sequence) and latest spacer sequence close to the leader sequence are combined and used as the total analyzed sequences for strain typing [60].

4.3.4 Conserved CRISPR arrays serving as quantitative PCR targets

In addition to the sequences analyses of CRISPR loci typing in Salmonella, the conserved CRISPR arrays can also be used as targets for qPCR primers and probes design. It has been demonstrated that a S. Infantis-specific qPCR assay is able to detect the Infantis serovar from mixed cultures of Salmonella down to 0.1% of the population, and with the detection sensitivity of 10 colony forming units [21]. For the utility of this CRIPSR based qPCR molecular approach in improved surveillance system, two main parameters need to be met in regards to the CRISPR spacer sequences that are to be used for designing primers and probes: (1) the used spacers need to be specific for the tested serovar; (2) the selected spacers need to be conserved and present in all strains of that specific serovar.

4.3.5 Other related applications of CRISPR-Cas

It is well-known that efficient delivery of a CRISPR/Cas9 plasmid is critical for effective therapy in clinical settings. Other than a receipt of a plasmid carrying CRISPR/Cas, it has also been found that Salmonella can be used as a CRISPR/Cas9 plasmid carrier for in vivo therapy against virus-induced cancer [68]. It has been demonstrated that the usage of Salmonella in CRISPR system provides a simpler and more effective platform for in vivo therapy [68].

The CRISPR-Cas system can also be used for diagnostics by utilizing the properties of the proteins themselves. For example, the Cas9 protein has been used in detection assays to help increase the percentage of the genomic regions of interest that is present. During library preparation in NGS projects, a method known as Depletion of Abundant Sequences by Hybridization uses recombinant Cas9 protein complexed with a library of guide RNAs to target and cleave unwanted DNA, leading to the increased yield of sequences of interest [69].

Additionally, the method Finding Low Abundance Sequences by Hybridization (FLASH) uses Cas9 and guide RNAs that allow sequences of interest to be cleaved into an ideal size for NGS sequencing, increasing the presence of reads that can be captured from the sequence of interest [70]. This can be used for example to detect antimicrobial resistant populations that may be present in low levels compared to the wild type. Recently, a modified Cas9 variant has been developed (SpCas9 named SpRY), that allows for the digestion of specific regions without the requirement of the PAM sequence [71].

Other diagnostic tools have also been developed with the CRISPR-Cas system. Although the most reported cas protein in Salmonella is related to Cas9, the other two most popular cas proteins related to diagnosis tests are cas12a and cas13. Different from Cas9 which cuts double stranded DNA relying on a precise location “T rich” PAM, both cas12a [72] and cas13 [73] remain bound to the target and then cleave other DNA/RNA non-discriminately. This feature is recognized as collateral cleavage of trans-cleavage activity, which has been broadly applied in the development of various diagnostic technologies [74]. DNA endonuclease-targeted CRISPR trans reporter is a method developed using the Cas12a protein and its ability to degrade single-stranded DNA (ssDNA). Using this property, along with a ssDNA reporter, this system can be used to detect if specific pathogen types are present in a sample [72]. Another example can be seen from the global pandemic associated with betacoronavirus severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). For the detection of SARS-CoV-2 in clinical validations, specific high-sensitivity enzymatic reporter unlocking (SHERLOCK), has been shown to be a promising alternative method to qPCR with regards to its visualization speed and experimental settings with limited resources [75, 76, 77]. These are just a few ways the CRISPR-Cas system can be alternatively used for diagnostic purposes. These diagnostic or detection methods could also be adapted and used as alternative methods for surveillance or typing in Salmonella.

5. Conclusions

The CRISPR-Cas system in Salmonella has been shown to be useful in differentiating between different strains. According to the WGS based genome analysis of 115 isolates, three S. enterica subsp. salamae isolates, and one S. enterica subsp. enterica serovar Gaminara possess three CRISPR loci, namely CRISPR1, CRISPR2 and CRISPR3, which differs from the commonly reported two CRIPR loci CRISPR1 and CRISPR2 in Salmonella. On the contrary, 18 isolates only had CRISPR1 and five isolates only had CRISPR2. With the emerging applications of CRISPR arrays in Salmonella genome editing, AMR studies, typing and subtyping in diagnosis and surveillance, a thorough investigation of the uses of CRISPR-Cas will facilitate better understanding its host-pathogen interaction, immune response and its usages in improving laboratory tests. Adapting the many advanced CRIPSR-based diagnostic tools such as SHERLOCK, and FLASH, will allow for faster detection and/or the ability for more detailed analyses to be carried out. This will allow for improved laboratory diagnosis and surveillance endeavors in food safety, as well as offer better tools for any future outbreak responses.

Acknowledgments

This work was supported by the Public Health Agency of Canada and Canadian Food Inspection Agency.

Conflict of interest

The authors declares no conflict of interest.

Acronyms and abbreviations

Acr	anti-CRISPR protein
AMR	antimicrobial resistance
CARD	comprehensive Antibiotic Resistance Database
CRISPR	clustered regularly interspaced short palindromic repeats
CCT	conventional CRISPR typing
CLSPT	CRISPR locus spacer Pair Typing
CLTSST	CRISPR Locus Three Spacer Sequences Typing
crRNAs	CRISPR RNAs
DR	direct repeats
FLASH	Finding Low Abundance Sequences by Hybridization
MLVA	multi-locus variable number tandem repeat
MLST	multi-locus sequence typing
NCBI	National Center for Biotechnology Information
NTS	non-typhoidal Salmonella
NGS	next generation sequencing
PAM	protospacer adjacent motif
PFGE	pulsed-field gel electrophoresis
SARS-CoV-2	severe acute respiratory syndrome coronavirus 2
SHERLOCK	specific high-sensitivity enzymatic reporter unlocking
ssDNA	single-stranded DNA
SNP	single nucleotide polymorphism
WGS	whole genome sequencing

Stella Elkabes

Chapters authored

Related collaborators

Tomislav Benjak

Stephen Bustin

Gorka Vuletic

Ponzio

Mili Mandal

Pan Zhang

Junichi Sadoshima

Sayantani Basak

Peiyong Zhai

Robert Donnelly