Bioinformatic and literature assessment of toxicity and allergenicity of a CRISPR-Cas9 engineered gene drive to control Anopheles gambiae the mosquito vector of human malaria
Malaria Journal volume 22, Article number: 234 (2023)
Population suppression gene drive is currently being evaluated, including via environmental risk assessment (ERA), for malaria vector control. One such gene drive involves the dsxFCRISPRh transgene encoding (i) hCas9 endonuclease, (ii) T1 guide RNA (gRNA) targeting the doublesex locus, and (iii) DsRed fluorescent marker protein, in genetically-modified mosquitoes (GMMs). Problem formulation, the first stage of ERA, for environmental releases of dsxFCRISPRh previously identified nine potential harms to the environment or health that could occur, should expressed products of the transgene cause allergenicity or toxicity.
Amino acid sequences of hCas9 and DsRed were interrogated against those of toxins or allergens from NCBI, UniProt, COMPARE and AllergenOnline bioinformatic databases and the gRNA was compared with microRNAs from the miRBase database for potential impacts on gene expression associated with toxicity or allergenicity. PubMed was also searched for any evidence of toxicity or allergenicity of Cas9 or DsRed, or of the donor organisms from which these products were originally derived.
While Cas9 nuclease activity can be toxic to some cell types in vitro and hCas9 was found to share homology with the prokaryotic toxin VapC, there was no evidence from previous studies of a risk of toxicity to humans and other animals from hCas9. Although hCas9 did contain an 8-mer epitope found in the latex allergen Hev b 9, the full amino acid sequence of hCas9 was not homologous to any known allergens. Combined with a lack of evidence in the literature of Cas9 allergenicity, this indicated negligible risk to humans of allergenicity from hCas9. No matches were found between the gRNA and microRNAs from either Anopheles or humans. Moreover, potential exposure to dsxFCRISPRh transgenic proteins from environmental releases was assessed as negligible.
Bioinformatic and literature assessments found no convincing evidence to suggest that transgenic products expressed from dsxFCRISPRh were allergens or toxins, indicating that environmental releases of this population suppression gene drive for malaria vector control should not result in any increased allergenicity or toxicity in humans or animals. These results should also inform evaluations of other GMMs being developed for vector control and in vivo clinical applications of CRISPR-Cas9.
The World Health Organization (WHO) has estimated that in 2020 there were 241 million cases of malaria worldwide, resulting in 627,000 deaths, with countries from Africa shouldering circa 95% of this burden . In recent years, the progress that had previously been observed over the last decade in reducing malaria transmission has stalled, principally due to the emergence of resistance to both insecticide-based vector control against Anopheles mosquitoes and pharmaceuticals targeting Plasmodium pathogens . This has created the need for additional, complementary approaches to tackle malaria transmission, including novel vector control tools such as engineered gene drives [2, 3].
Typically, engineered gene drives are based on the CRISPR-Cas9 genome editing system and effectuated by a process known as homing [4,5,6]. The engineered gene drive involves a transgene encoding both the CRISPR-Cas9 endonuclease and guide RNAs (gRNAs) targeting a specific genomic location. The transgene is introduced via germline transformation into its specific genomic target location on one of a pair of homologous chromosomes. Next, in germline cells the Cas9/gRNA ribonucleoprotein (RNP) complex produces a double-stranded break specifically in the same genomic target location of the non-transgenic homologous chromosome. Via homology directed repair (HDR), the transgene and its flanking sequences are pasted into the genomic target location of the homologous chromosome so that both of the homologous chromosomes in that germline cell now contain the transgene. Thus, homing increases the proportion of parental germ cells that are transgenic to above the circa 50% associated with Mendelian inheritance, leading to a sustained increase in the frequency, and, ultimately potential fixation, of the engineered gene drive in the population. This property can be used to propagate transgenes into populations of mosquito vectors of malaria, which can either be used to disrupt reproductive capacity, in the case of population suppression gene drive, or reduce vector competence for Plasmodium transmission, in the case of population replacement gene drive .
The population suppression gene drive dsxFCRISPRh is currently being investigated as a potential tool for vector control against malaria . dsxFCRISPRh encodes (i) a human codon-optimized version of CRISPR-Cas9 endonuclease (hCas9) derived from Cas9 of Streptococcus pyogenes (SpCas9), whose expression is controlled by a germline promoter and (ii) the T1 gRNA, which is expressed ubiquitously and acts in concert with hCas9 to target the intron 4-exon 5 boundary of the doublesex gene, leading to disruption of the sex-specific transcript dsxF, and thus sterility, in females; along with (iii) the DsRed fluorescent marker that was originally sourced from Discosoma species and which is expressed in the nervous and ocular systems to facilitate visual detection of the transgene in transgenics .
Mosquito vectors of malaria containing such engineered gene drives are considered to be genetically-modified organisms (GMOs), or more specifically genetically-modified mosquitoes (GMMs) [8, 9]. Before any GMM or GMO can be proposed for release in the environment, a rigorous environmental risk assessment (ERA) must be performed [8,9,10,11]. As part of this ERA, any potential impacts on human health must be examined, including any potential allergenic or toxic effects from the transgenic proteins [8, 9]. In cases where the GMO is a food crop, allergenicity and toxicity assessments are based on the food to be consumed from that GMO [12, 13]. These evaluations are performed on a case-by-case basis, depending on the nature of the genetic modification, and are based on a weight of evidence approach, as no single investigation can definitively determine allergenicity and toxicity. Initial assessments of the transgenic proteins are tiered to consider amino acid sequence homology comparisons with known allergens or toxins, their sensitivity to pepsin digestion, and evaluation of any new substances or metabolites that they may produce. For example, where a transgenic protein shows less than 35% identity within a segment less than 80 amino acids with a known allergen, it is considered neither to be a known allergen nor likely to be cross-reactive to known allergens [12, 13]. Where a transgenic protein shows greater than 35% identity in a segment of 80 or more amino acids with known allergens, then it should be assessed using serum from individuals sensitized to the identified allergenic source. In contrast to this “35%/80aa cutoff” in allergenicity bioinformatic assessments, there are no specified criteria for homology searches between transgenic proteins and toxins [12,13,14]. Instead, a weight of evidence approach applied to a tiered evaluation of toxicity has been proposed based on (i) mode of action (ii) amino acid sequence similarity of the transgenic protein to a known toxin, (iii) history of safe use of the protein or source organism, and (iii) probability and level of exposure [12,13,14]. Alignments of primary amino acid sequences of known toxins with transgenic proteins therefore form an integral part of initial assessment of their potential for toxicity . This tiered approach of substantial equivalence in toxicity assessment has been suggested by the WHO, FAO and OECD .
Previously, the first stage of an ERA, problem formulation, was conducted to identify potential harms to protection goals for the environment, as well as human and animal health, from simulated releases of the dsxFCRISPRh population suppression gene drive in Anopheles gambiae sensu lato (s.l.) in West Africa . That analysis identified 46 plausible pathways to potential harm, nine of which could be evaluated by testing the risk hypothesis that dsxFCRISPRh transgenic products would not cause increased allergenicity or toxicity. In the present study, amino acid sequences of DsRed and hCas9 were interrogated against those of known toxins or allergens from four different bioinformatic databases. The sequence of the gRNA expressed in dsxFCRISPRh was also compared with those of Anopheles and human microRNAs to examine the potential for disruption to host gene expression that could be associated with increased allergenicity, toxicity or pathogenicity. Finally, literature was examined for any evidence of toxicity or allergenicity of the transgenic proteins themselves, or of the donor organisms from which they were originally derived.
Amino acid sequences of the transgenic proteins DsRed and hCas9, and nucleotide sequences of the T1 gRNA, all of which are expressed from the engineered gene drive dsxFCRISPRh, are shown in Table 1. Sequence searches were conducted between 5 and 15th November 2022 and exploited four bioinformatic repositories:
NCBI, the National Centre for Biotechnology Information of the National Library of Medicine of the National Institutes of Health of the USA (https://www.ncbi.nlm.nih.gov/). The 30th October 2022 update of Expressed Sequence Tags (EST) databases of NCBI contained 77,393,133 cDNA sequences. The 5th November 2022 update of the Transcript Reference Sequences (TRS) databases of the NCBI contained 39,247,328 cDNA sequences.
UniProt; maintained by the European Molecular Biology Laboratory’s European Bioinformatics Institute (EMBL-EBI), Swiss Institute of Bioinformatics (SIB) and Protein Information Resource at Georgetown University, USA (https://www.uniprot.org/). The UniProtKB database is a comprehensive, high-quality and freely accessible protein database made up of both the Swiss-Prot protein database that, as of 12th October 2022, contained 568,363 manually annotated and reviewed entries and the TrEMBL translated nucleotide database containing 229,928,140 automatically annotated entries.
COMPARE database: Comprehensive Protein Allergen Resource (COMPARE) database, 2022 update (http://www.comparedatabase.org/), consists of a repository of 2463 peer-reviewed protein sequences of known or putative allergens, and is maintained as a collaborative effort of the Health and Environmental Sciences Institute and Protein Allergenicity Technical Committee with programmatic support from the Joint Institute for Food Safety and Nutrition at the University of Maryland.
AllergenOnline database maintained by the Food Allergy Research and Resource Program of the University of Nebraska: version 21 (http://www.allergenonline.org/) with 2,233 allergen sequences, updated on 14th February 2021.
hCas9 or DsRed homology searches
Using TBLASTN bioinformatics program version 2.13.0 + , the amino acid sequence of hCas9 or DsRed transgenic protein was used to search NCBI TRS and EST databases. Searches were limited by Entrez query “toxin” or “allergen”. Only search returns with Expected values (“E values”) of 10 or below were returned. Algorithm parameters were based on program default settings of Matrix—BLOSUM 62, Gap Costs—Existence: 11 Extension: 2; Compositional adjustments—Conditional compositional score matrix adjustment. Any resultant protein sequence alignments were saved on the NCBI server.
UniProtKB version 2022_04 was searched using the keyword “toxin” to identify 1,280,940 entries, which were downloaded from the UniProt website as a compressed FASTA file (“UniProt-Toxin”). Separately, the UniProtKB database was searched using the keyword “allergen” to identify 53,290 entries, which were downloaded from the UniProt website as compressed FASTA file (“UniProt-Allergen”). Using standalone BLAST + software for Windows , UniProt-Toxin and UniProt-Allergen were converted to BLAST protein databases. Using BLASTP 2.13.0 + , sequence alignments with the full-length amino acid sequences of both Cas9 and DsRed were performed on the amino acid sequences from both the BLAST UniProt-Toxin and BLAST UniProt-Allergen protein databases.
For both NCBI and UniProt searches, alignments for allergens were considered significant where the protein contained a stretch of 80 or greater amino acids that showed 35% or more identity . Any alignments which were returned with an E value less than 10 from bioinformatic toxicity assessments were examined further manually for significant sequence identity over a substantive length of protein. Considerations for significance during these manual examinations were informed by a number of observations from the literature. Firstly, in comparative modelling of proteins, primary amino acid sequences are used to infer three-dimensional protein secondary structures, typically yielding high levels of accuracy above 50% sequence identity but highly inconsistent outcomes when sequence identity fails below 30% [19,20,21]. As articulated by Negi et al. , sequences with 35% identity are generally considered to have similar three-dimensional structures [23,24,25]. Indeed, their systematic analysis of over 10,000 manually curated toxin sequences led to the proposal that 35% sequence identity level could be used for functional grouping of toxins . Secondly, the 80aa cut-off for allergenicity testing adopted by Codex Alimentarius was reportedly based on the length of stretches of conserved amino acids corresponding to function domains in proteins . This assertion is consistent with studies of the free energy of unfolding of protein domain structures , or the distribution of sizes of protein domain boundaries from a three-dimensional database of 1882 sequence-dissimilar protein domains , which indicate a narrow range of around 100 amino acids for optimal sizes of protein domains. Adoption of an 80 amino acid as a cut-off for domain sizes of toxins would thus represent a conservative estimate. Therefore, for evaluation of alignments between transgenic proteins and toxins, alignments with 35% or more amino acid identity over a segment of 80 or more amino acids were considered to be significant.
The COMPASS (COMPare Analysis of Sequences with Software) program available from the COMPARE allergen database was used to search for stretches of 35% or greater identity over 80 or greater amino acids with DsRed and hCas9. Algorithm input parameters were based on program default settings of E value 10, Word Size 2, Gap Open and Gap Extension Penalties 10/2. On the recommendation of Codex Alimentarius , COMPARE was further searched by 80 amino acid segments of transgenic protein segments in order to identify structural motifs, much shorter than the intact protein, which might contain a conformational IgE binding epitope or could help to identify potentially cross-reactive proteins that are not true homologues of an allergen that have significant local identities that might provide an immunological target for IgE antibodies in those with allergies to the matched allergen. A match of 35% or more over 80 or greater amino acids with a known allergen suggested further testing for possible cross-reactivity. In addition, based on hypothetical epitope sizes, windows of 8 amino acids of the transgenic query sequence were also compared with known allergens in the COMPARE database.
Amino acid sequences of transgenic proteins DsRed and hCas9 were also used to search the AllergenOnline database using FASTA36 with scoring matrix BLOSUM 50 to identify potential allergenic proteins. Search returns were limited to E value of 10 or below. A match of 35% or more over 80 or greater amino acids with a known allergen suggested further testing for possible cross-reactivity. AllergenOnline was further searched by 80 amino acid segments of transgenic protein segments in order to identify structural motifs, much shorter than the intact protein, which might contain a conformational IgE binding epitope. A match of 35% or more over 80 or more amino acids with a known allergen indicated further testing for possible cross-reactivity. In addition, windows of 8 contiguous amino acids of the transgenic query sequence were also compared with known allergens in the AllergenOnline database.
gRNA nucleotide search strategy
The nucleotide sequence of the T1 gRNA is shown in Table 1. This is expected to be constitutively and ubiquitously expressed from the U6 promoter in dsxFCRISPRh  and thus was assumed to be present in the saliva of dsxFCRISPRh GMMs. The nucleotide sequence of the gRNA was examined for similarities to known microRNAs to establish whether they might exert allergenic or toxic effects via the potential to alter endogenous gene expression in either GMMs or humans that might be bitten by the GMMs.
The miRBase is the primary public repository and online searchable database for microRNA (miRNA) sequences and annotation, containing 38,589 entries (http://www.mirbase.org/). It was established in the UK by the Wellcome Sanger Institute for genome research and is hosted by the University of Manchester. Searches using both gRNA nucleotide sequences from dsxFCRISPRh were conducted on the miRBase22.1 data release from October 2018 using the BLASTN search method to identify any similar mature miRNA sequences from all organisms in the database, including An. gambiae and Homo sapiens. Search results were considered significant where E values were less than 0.01 .
The PubMed biomedical literature database (https://pubmed.ncbi.nlm.nih.gov/), was used as the primary data source for scientific literature on allergy and toxicity. The main objective of the searches was to identify evidence of toxicity or allergenicity of (i) ‘the transgenic proteins’ hCas9 and DsRed; and (ii) ‘the source organisms’ of the transgenic proteins, Streptococcus pyogenes and Discosoma spp., respectively. Combinations of different search strings were used, involving the terms ‘allergen’ or ‘allergenic’, ‘toxin’ or ‘toxicity’, ‘DsRed’; ‘Cas9’; ‘Discosoma’ or ‘Streptococcus pyogenes’. Each search string was conducted to obtain results from inception of search engine until the 31st October 2022. Results were downloaded to the Endnote reference manager software (Endnote V. X9.2) for storage and review. Abstracts of identified publications were examined against inclusion and exclusion criteria. The full text of surviving articles was subsequently examined against selection criteria. Based on bioinformatic and initial literature results, further specific literature searches were used to identify additional literature relevant to specific putative toxins or allergens of interest.
Research articles and published laboratory studies were included for review when they:
Were structured with introduction, materials and methods, results, and discussion sections or equivalent
Were authored in English
Were published between inception of the PubMed search engine until date of completions of searches in this study, 31st October 2022
Provided in vitro, in vivo or ex vivo evidence of acute or chronic toxicity or allergenicity (or lack thereof) caused directly by (a) Cas9; (b) DsRed; (c) any of the components or breakdown products of (a) and (b), or (d) Discosoma spp., or S. pyogenes
Were pathological studies involving source organism/s which established a mechanistic cause of disease effects
Provided evidence of allergenicity through the route of IgE antibody elicitation
Were available as full-text items through either open access, or through Imperial College London library and online literature facilities.
Defining toxicity and allergenicity
The WHO has defined a “toxic agent” as “anything that can produce an adverse biological effect” and “toxicity” as “the capacity of a substance to cause injury to a living organism. A highly toxic substance will cause damage in small quantities, while a substance of low toxicity will need large quantities to produce an effect. Toxicity is also dependent on the portal of entry, the time frame of exposure and the latent period” . With reference to genetically modified foods, the Society of Toxicology has defined potential toxicity to include “inherent toxicity of the transgenes and their products, and unintended (pleiotropic or mutagenic) effects resulting from the insertion of new genetic material into the host genome” . Therefore, for this analysis, toxicity was assumed in any study that demonstrated harm or impairment to function of a cell, or component of an organism, or cause cell death in cells or an organism, including via mutagenic effects.
Allergenicity is the potential of a substance to cause an allergy and can be caused by a variety of immunological mechanisms, however IgE-mediated allergy represents the main form of allergy, that causes the most severe reactions and the only form causing life-threatening reactions . An allergen, in the context of an IgE food allergy, exerts its effects in two stages: first, (i) sensitisation where no symptoms occur while the capacity of the immune system to react increases dramatically, and later (ii) elicitation (provocation) with clinical manifestations . This form of IgE-mediated allergy has been the focus of risk assessments of allergenicity of GMOs and novel foods [12, 32,33,34,35,36,37].
An allergen is further defined by AllergenOnline as a protein that has been demonstrated to “specifically bind IgE using sera with individuals with clear allergies to the source of the gene/protein and further that the protein causes basophil activation or histamine release, skin test reactivity or challenge test reactivity using subjects allergic to the source.” Therefore, for this analysis, allergenicity was defined as any substance that has evidence of eliciting an IgE response, in vivo or in vitro, both to the protein source or to protein itself.
Bioinformatic assessment of hCas9 toxicity
NCBI TRS and EST database searches yielded four entries with an E-value below 10, but none had 35% or greater identity over 80 or more amino acids to known or putative toxins. Searches of BLAST UniProt-Toxin protein databases with hCas9 identified six proteins with E-values below 10 (see Table 2). Only one of these sequence alignments was significant: entry H1D476 encoding a Type II toxin-antitoxin system VapC family toxin from Fusobacterium necrophorum subsp. funduliforme, which showed 43% identity to hCas9 over an alignment length of 119 amino acids (shown in blue in Table 2).
In terms of mode of action, the VapC toxin encodes a sequence specific endoribonuclease that cleaves RNA species, such as the initiator tRNAfMet or 23S rRNA, which are specifically components of bacterial mRNA translation [38,39,40,41]. In the hCas9 endonuclease, the HNH domain cleaves the DNA strand that is complementary to the gRNA target sequence, while the RuvC domain cleaves the non-target DNA strand, so that both domains must act in concert to produce a double-stranded break (DSB) in a specific DNA sequence. In F. necrophorum, the VapC gene is predicted to encode a protein of 117 amino acids , 33 and 73 of which align within the HNH and RuvC-III domains found between amino acids 805-964 and 965-1141 of hCas9, respectively  (see Fig. 1). While this alignment contains neither of the amino acids serving as catalytic residues in the HNH domain, the HHAHDAYN motif that is common to both the VapC protein and the RuvC-III domain of hCas9 contains two of the four catalytic residues of the RuvC active site (Fig. 1) and is a highly conserved motif in Cas9 homologues from multiple species of bacteria . Amino acid sequence alignment between VapC and hCas9 is thus based on a similar nucleic acid cleavage capability between both proteins.
In terms of history of safe use of the protein or source organism, when the UniProt database was searched with the predicted amino acid sequence of this VapC protein, 341 hits from a range of prokaryotic species were identified with an E-value below 10 (see Additional file 1: File S1). Of these entries, 239 (70%) included the term “Cas9”. Thus, VapC shows sequence identity to a wide range of Cas9 homologues in heterogenous bacterial species that are widely distributed throughout nature . Similarly, humans and other animals are regularly exposed to F. necrophorum, as it is commonly found within the alimentary tract; in studies based in London accounted for 10% and 21% of acute and recurrent sore throats, respectively [45,46,47,48]. However, the action of the VapC toxin does not itself directly cause symptoms in patients; rather bacterial toxin-antitoxin systems such as VapC consist of both a protein toxin that can interfere in essential bacterial physiological processes to inhibit cell growth and an antitoxin to counteract the bactericidal or bacteriostatic activity of the toxin . There are three broad functions ascribed to such toxin-antitoxin systems, each of which may confer a survival advantage to their bacterial hosts . Firstly, in “post-segregational killing” a plasmid encoding both the toxin and a less stable antitoxin leads to the death of plasmid-free offspring that are incapable of producing antitoxin . Secondly, in “abortive infection” bacteriophage multiplication in a population of bacteria is inhibited by activation of the toxin-antitoxin system when cells infected with bacteriophage undergo “altruistic” cell death [51, 52]. Thirdly, in “persister formation” a subset of bacterial cells in a population develops tolerance to antibiotics or other environmental stressors as a result of the toxin inhibiting biochemical functions normally targeted by those stressors, leading to dormancy rather than death in those cells . Therefore, any toxicity of VapC is restricted to prokaryotes via its inhibitory effects on mRNA translation, and thus cell growth, specifically in bacteria.
Literature assessment of hCas9 toxicity
Biting is a likely route of exposure to material from dsxFCRISPRh mosquitoes while they are seeking bloodmeals from human or animal hosts. However, dsxFCRISPRh Cas9 is under the control of the zero population growth (zpg) promoter which drives expression in the germline of both males and females from late pupal developmental stages. Cas9 is therefore not expected to be expressed in the saliva of dsxFCRISPRh transgenics . In the case of accidental, incidental ingestion of dsxFCRISPRh mosquitoes by humans or animals, Cas9 would also be subject to pepsin digestion  so that exposure via this route is expected to be negligible. In general, the levels of Cas9 transgenic protein in any given individual dsxFCRISPRh mosquito would be insignificant in terms of biomass within the environment .
Nonetheless, Cas9 is a nuclease capable of inducing DSBs in DNA and, therefore, has the potential to be mutagenic. Recognition of its genomic target sequence is facilitated via the particular sequence of its gRNA so that its endonuclease activity can be designed to be highly target-specific . By altering the circa 20 nucleosides target recognition sequence of the gRNA, Cas9 can be directed towards specific nuclease activity against almost any DNA target of interest. While Cas9-induced DSBs other than at the intended target sequence, known as “off-target effects”, have been observed in some studies [57, 58], many other in vivo studies have shown low (< 1%) or no off-target effects through therapeutic proof-of-concept use of Cas9 [59,60,61,62]. Cas9 has been used to investigate carcinogenic factors in vivo  as well as other clinical conditions with no detectable off-target effects [64,65,66,67,68]. In some in vitro studies, however, indels ranging from small to large, as well as chromosomal translocations, chromothripsis and chromosomal aberrations, have been observed [69,70,71,72,73,74,75,76]. For example, Kosicki et al. detected “on-target mutagenesis”, such as large deletions and insertions, at target sites in mouse and human cell lines . To understand the origins of these deletions, the authors generated a library of mouse embryonic stem cells deficient in 32 DNA repair genes based on a single clone constitutively expressing Cas9, and found that the frequency of large deletions increased when genes essential for non-homologous end-joining (NHEJ) were impaired, and decreased when genes required for microhomology-mediated end joining (MMEJ) were disrupted . Depending on the specific cell type or stage of cell cycle, some DNA repair mechanisms, whether HDR, NHEJ or MMEJ, may be favoured over others, which thus may determine the propensity for, and characteristics of, any such on-target effects [78,79,80]. For example, prolonged in vitro intracellular expression of Cas9 in human pluripotent stem cells caused DSBs in DNA that were toxic via P53 inhibition of HDR, a mechanism which appeared to be cell-specific and related to the early developmental status of the cell types [81, 82]. Kosicki et al. urged caution when using high concentrations of nucleases and proposed further investigations using inducible Cas9 or gRNAs of different target site binding affinities . Indeed, judicious choice of target sites and design of gRNAs has been widely demonstrated to improve the target specificity of Cas9 in a wide range of systems [57,58,59,60,61,62,63,64,65,66,67,68, 83,84,85,86,87,88,89,90,91,92,93,94,95,96], including in in vivo human studies .
As can be the case for ectopic expression of other heterologous proteins, constitutive intracellular expression of Cas9 can also be toxic due to stresses on protein homeostasis. For example, constitutive intracellular expression, but not transient expression, of Cas9 is toxic in the single-cell alga Chlamydomonas reinhardtii , rice blast fungus Magnaporthe oryzae  and the protozoan parasite Toxoplasma gondii . Ferreira et al.  found that strong in vivo expression of Cas9 in zebrafish muscle fibres induced cell toxicity but attributed this to disruptions to cellular protein homeostasis leading to apoptosis. The method of delivery used for Cas9, rather than Cas9 activity by itself, may also cause toxicity. In a study by Li et al. , two different delivery methods for Cas9 were compared under the same experimental settings. Authors found that an adenovirus vector in vitro resulted in attenuated toxicity seen in CD34 + haematopoietic stem cells when compared with the electroporation method of delivery. Similar results were found in vivo in mice muscle myofiber degeneration and repair, when adenovirus vector showed no significant muscle cell damage, compared with electroporation methods . Chromatin environment of the targeted locus has also been found to affect DSBs associated toxicity as a result of CRISPR/Cas9 editing , though this has been disputed by Friskes et al. who attribute major determinants of toxicity to be related to cutting efficiency and off-target DSBs, as opposed to chromatin features . Holmgaard et al.  created mutations in the vascular endothelial growth factor A (VEGFA) gene in retinal pigment epithelium in mice by performing subretinal injections of lipoplexes containing a range of concentrations of guide RNAs complexed with Cas9 as RNPs. Toxicity was only induced at high, but not lower, concentrations of RNP. Therefore, the authors concluded that RNP-based delivery in vivo was a potential strategy for treatment of retinal diseases in humans. Consistent with these observations, Garrood et al.  reported that Cas9 off-target mutagenic effects could not be detected in the genomes of dsxFCRISPRh GMMs, despite such events being observed in other gene drive strains that expressed hCas9 from the vasa2 promoter and gRNAs targeting disparate genomic targets. The specificity of hCas9 DNA endonuclease activity in dsxFCRISPRh GMMs appears to be a consequence of both the choice of sequence of the T1 gRNA specifying the genomic target and more restricted spatiotemporal expression of hCas9 from the zpg promoter, compared to vasa2. Lower levels of hCas9 expression from the zpg promoter thus appeared to favour the absence of detectable off-target nuclease activity. Overall, therefore, any toxicity of Cas9 is associated with high cellular levels of its expression or concentration.
There are numerous animal studies where Cas9 has been expressed in vivo, without any toxic effects observed. Qiu et al.  showed expression of Cas9 in vivo in mice resulted in no off-target mutagenesis or evidence of toxicity. Reisman et al.  used Cas9 in vivo in a murine model to successfully suppress leukaemia progression, with minimal toxicity observed. Xiong et al.  showed that nanoparticle delivery of Cas9 resulted in low toxicity and high biological safety both in vitro and in vivo in mice. Han et al.  demonstrated safe and effective use of Cas9 in vivo, in murine models for haemophilia therapy. Wei et al.  showed effective therapeutic in vivo transplantation of Cas9 gene-corrected autologous hepatocytes into mice with Wilson’s disease with no toxicity seen. Many other in vivo studies involving expression of Cas9 demonstrated no, negligible or low toxic effects [94, 111,112,113,114,115,116,117,118].
Human exposure to Cas9 can take the form of either ex vivo or in vivo therapeutic applications [61, 76, 119, 120]. In ex vivo therapies, autologous cells are removed from the patient and modified in vitro with a vector encoding Cas9 or using recombinant RNPs. Once the genomic editing has been performed, the edited cells are infused back into the patient. Any potential toxicity of Cas9 from ex vivo therapeutic applications would likely arise from indirect effects, such as the introduction of off- or on-target mutations in, or toxic effects on, the edited cells . In the case of in vivo therapeutic applications, Cas9 is introduced systemically into the patient via viral or lipid nanoparticle gene delivery systems. In vivo therapies could thus expose patients to direct effects of potential toxicity from Cas9 .
For ex vivo therapies, Cas9 has been exploited as a gene editing tool for cancer treatment, with minimal toxicity observed [72, 122]. For example, Foy et al.  reported the ex vivo non-viral use of CRISPR-Cas9 to treat 16 patients with refractory solid tumours. In T cells obtained from patients, two T cell receptor genes were knocked out and replaced with mutant neoantigen-specific T cell receptors that been previously identified from circulating T cells in patients. After chemotherapy to deplete pre-existing lymphocytes, the modified autologous T cells were infused into patients. There were only two adverse events associated with the T cell therapy, one involving mild cytokine release and neutropenia and another involving grade 3 encephalitis which was resolved with corticosteroid treatment, demonstrating the safety of this approach [123, 124]. The immunotherapeutic use of Cas9 to edit T-cells through disruption of the PD-1 gene has also now entered Phase I clinical trials [125, 126]. A further Phase I clinical trial has used CRISPR-Cas9 engineered CAR19 universal T cells to treat children with refractory B cell leukaemia, with all primary safety objectives met, demonstrating feasibility and therapeutic potential of this immunotherapy . Frangoul et al.  also reported that two patients, one with transfusion-deficient beta-thalassaemia (TDT) and one with sickle cell disease were administered with a Cas9 to target the BCL11A transcription factor. More than a year later, both patients had high levels of allelic editing with no off-target effects observed. While the first patient exhibited two serious side effects, one was attributed to a case history of viral hepatitis, while the other, pneumonia, resolved within 28 days. The second patient had three serious adverse events, which were all attributed to the administered Cas9 being enriched with CD34 + cells, which may have contributed to the delay in lymphocyte recovery. A further clinical trial in Phase III of development is underway, utilizing ‘Exa-cel’, an investigational, autologous, ex-vivo CRISPR/Cas9 gene-edited therapy being evaluated for patients with sickle-cell disease or transfusion-dependent beta-thalassaemia . Having provided robust evidence of potential as a one-time functional cure for these diseases, in previous clinical trials, two patients out of seventy-five reported serious adverse events, both of which were resolved .
With respect to in vivo therapeutic applications of Cas9, Gillmore et al.  reported delivery of intravenous injections of mRNA encoding Cas9 and a gRNA targeting the transthyretin gene encapsulated in lipid nanoparticles in six patients with life-threatening transthyretin amyloidosis. Safety assessments of those patients during the first 28 days of the clinical trial revealed few adverse events, all of which were mild in grade. A second clinical trial uses CRISPR-Cas9 to disrupt the kallikrein protein to counteract altered levels of the peptide hormone bradykinin in patients with hereditary angioedema, which results in severe swelling of the limbs and abdomen . The intervention involves Cas9 encoding mRNA, along with its gRNA, delivered in lipid nanoparticles and injected directly into the patients’ bloodstream which then travels to, and gets expressed in, the liver . Swelling attacks went from one to three a month in two patients, and from up to seven a month to none in another, so that all three were able to stop taking their anti-swelling pharmaceuticals . A third clinical trial is underway involving adeno-associated virus in vivo delivery of Cas9, for gene editing of photoreceptor cells to treat blindness, with preliminary results revealing a ‘favourable safety profile’ , with two patients seeing improvements in their vision . Another ongoing in vivo therapeutic study is directed at the treatment of cervical intraepithelial neoplasia and cervical cancer by directly injecting plasmid DNA encoding Cas9 that targets the human papilloma virus 
Bioinformatic assessment of hCas9 allergenicity
NCBI TRS and EST database searches yielded no entries with an E-value below 10. Searches of BLAST UniProt-Allergen protein databases with hCas9 identified two proteins with E-values below 10 (see Table 2). However, neither sequence alignment was significant. Using the COMPASS to search the COMPARE database, no entries were identified with E values below 10 (see Table 2). Sliding 80-mer searches identified no positive hits. Searches of the AllergenOnline database also identified no entries with E values below 10 and no sliding 80-mer positive hits.
In searches of both COMPARE and AllergenOnline, of the 1416 stretches of 8 amino acids identified in hCas9, one 8-MER sequence match, ‘EHIANLAG’, was also found in the protein allergen Hev b 9 (enolase isoform 1 from Hevea brasiliensis; NCBI accession number CAC005320; highlighted in blue in Table 1.) Epitopes on different proteins, as a rule evolutionarily related, may be identical or similar enough to bind to the same IgE molecules. Cross-reactivity, defined as the sensitization of one allergen causing the immune system to respond to another allergen as a result of shared epitopes , could then potentially occur. There are two levels of cross reactivity; (i) limited to IgE-binding without observed adverse effects (cross sensitization) or (ii) IgE-binding which also confers clinical reactivity (cross-allergenicity) . Cross-reactivity on the level of IgE-binding is more common and widespread than clinical cross-reactivity . The in vitro demonstration of cross-reactivity between two allergens in terms of IgE-binding is no proof of clinical cross-reactivity, and only means that one important pre-requisite for clinical reactivity is present [32, 134]. Furthermore, even the presence of IgE to a specific allergen may result in negative clinical challenges, as shown in a study exploring soybean allergenicity .
Hev b 9 is one of 15 allergens officially listed by the WHO as being identified in causing Natural Rubber Latex Allergy (NRLA) . Some of these identified proteins such as Hev b 5 and Hev b 6.01, are also found in other “lactiferous” or latex-excreting plants, and are associated with asymptomatic sensitization, as well as IgE-mediated hypersensitivity . Of importance, Hev b 9, whilst sharing one epitope with hCas9, shares 65% sequence identity and 65 epitope matches with Human enolase isoform 1, found in mammalian tissue and functions as a glycolytic enzyme . Human enolase 1 or alpha enolase, is one of three enolase isoenzymes found in mammals; and was found to have significant sequence identity to Hey b 9 but was not found to contain the ‘EHIANLAG’ epitope found in Cas9. Fagerberg et al. 2014 analysed expression of Human enolase 1 in tissue samples from 95 human individuals to determine tissue specificity of various protein-coding genes, and found ubiquitous expression in kidney, oesophagus and 25 other tissues .
The identical peptide match method using a peptide length of six amino acids has attracted criticism, since it generates many false positives in testing of potential allergenicity [138,139,140,141]. Moreover, even in linear B-cell epitopes some amino acids can be replaced without loss of antibody binding and, as a consequence, Codex Alimentarius does not include the criterion of six identical amino acids in their guidelines , but rather proposed instead that the scanning peptide size should be based on a scientifically-justified rationale. European Food Safety Authority (EFSA) guidance on the assessment of the allergenicity of GM foods aligns with the recommendations of the Codex Alimentarius for in silico testing for prediction of potential allergenicity is recommended . However, EFSA recommend that peptide match of complete identity over six contiguous amino acids to known allergens is—on its own—associated with very poor specificity and its relevance is doubtful .
Furthermore, AllergenOnline states that searches can be performed where it was not possible to find examples of an isolated identity match of six or eight amino acids with cross-reactive proteins unless there was at least a 35% identity match over 80 amino acids. In further consideration of the importance of epitopes in implicating allergy, a study by Hileman et al.  found that high numbers of non-allergens have matching sequences of seven amino-acids with known allergens, hence such an event cannot be interpreted as an indication of an allergenic potential . This notion has also been undermined as it can produce false positives in other studies [139, 140, 143,144,145]. Moreover, it has been reported that for any cross-reactivity to occur, in addition to the shared epitope, a high degree of similarity (between > 50–70%) is needed over a segment of 80 or greater amino acids [33, 146, 147]. As no such significant sequence identity match was found between hCas9 and Hev b 9, the epitope match identified between hCas9 and Hev b 9 can be considered as unlikely to pose an allergenic risk.
Literature assessment of hCas9 allergenicity
Cas9 has commonly been used in studies to evaluate allergens and immune-related diseases such as egg allergies , or to remove allergens from products such as soybean through site-directed mutagenesis  or target genes involved in allergic rhinitis ; or remove and functionally evaluate other allergens [151,152,153,154,155,156], with no literature studies having implicated Cas9 proteins themselves in allergenicity. Importantly, Nakajima et al.  reported that Cas9 was subject to rapid digestion in vitro and unlikely to cause food allergy in the case of accidental ingestion.
Human exposure to Cas9 is widespread, with acquired immune responses to Cas9 having been well documented [157,158,159,160,161]. Charlesworth et al.  detected antibodies and T cells against SpCas9 in 58% and 67% of 125 antibody donors, respectively. Wagner et al.  detected antibody titers against S. pyogenes antigens and SpCas9-reactive regulatory T cells in 85% and 96% of 48 donor individuals, respectively. Ferdosi et al.  detected anti-SpCas9 antibodies in more than 5% of 143 healthy individuals, while Simhadri et al.  found anti-SpCas9 antibodies in 2.5% of 248 donors. Charlesworth et al.  showed that donors positive for cellular activity against Cas9 were also positive for antibody activity, demonstrating a high concordance between adaptive and humoral immunity.
Moreover, a study by El-Mounadi et al.  outlines Cas9 sequence similarity of 23–58% to Cas9 protein present in Streptococcus thermophilus and Lactobacillus plantarum, two species of bacteria widely used as a probiotic in the production of cheese and yoghurt. Indeed, L. plantarum is frequently encountered as a natural inhabitant of the human gastrointestinal tract, as a result of it being a transient guest, introduced through diet ; and has been shown to contribute to inhibition of pathogenic bacteria . El-Mounadi et al.  further demonstrated more than 80% amino acid sequence similarity detected between Cas9 from S. pyogenes and that from human commensal and pathogenic bacteria such as Streptococcus dysgalactiae subsp. equisimilis, Staphylococcus aureus, Klebsiella pneumonia and Streptococcus canis.
Bioinformatic assessment of DsRed toxicity
NCBI TRS and EST database searches yielded no entries with an E-value below 10. Searches of BLAST UniProt-Toxin protein databases with DsRed identified 21 entries with E-values below 10 (see Table 3). BLAST UniProt-Toxin database searches yielded three entries with significant sequence similarity of greater than 35% over more 80 or more amino acids, but all were fluorescent proteins encoded by cloning vectors, and none were toxins. The remaining identified entries showed no significant sequence alignments to DsRed.
Literature assessment of DsRed toxicity
The fluorescent protein DsRed was originally isolated from Dicosoma spp. and is part of a broader family of intrinsically fluorescent proteins represented by Green Fluorescent Protein (GFP) . GFP is a stable and soluble monomer, which is brightly fluorescent green . DsRed is a tetramer and vibrantly fluorescent red with an emission wavelength of 583 nm . Its synthesis occurs via an extended form of the GFP synthesis pathway, and a single amino acid replacement can convert GFP chromoproteins into a DsRed‐like fluorescent proteins . Based on the homology between DsRed and GFP proteins, GFP has been used to assess the toxicity and allergenicity of the wider GFP family of proteins, including DsRed. In that regard, Richards et al.  found that ingestion of GFP in rodents did not affect growth, food intake, or relative weight of organs. Moreover, GFP was rapidly digested in simulated gastric fluid and found to possess insignificant levels of homology with known allergens.
In whole cell labelling experiments, moderate to high levels of intracellular expression of some variants of DsRed can be cytotoxic under in vitro conditions. For example, Strack et al.  found that retroviral expression of some variants of DsRed can compromise the viability of transfected HeLa cells in vitro. Retroviral-mediated expression of DsRed-Express fluorescent protein can also interfere with the viability or self-renewal of in vitro with cultures of Hematopoietic Stem Cells [170, 171]. The DsRed-Express variant has more rapidly maturing fluorescence, enhanced solubility, reduced green emission, and faster maturation rate than the original DsRed. By contrast, Strack et al.  also found that another variant, DsRed-Express2, had no such effect. Matsushima et al.  developed an in vivo imaging system using DsRed in mice to observe interleukin production in the skin, with no toxicity reported.
Nordin et al.  found that oral ingestion of transgenic Aedes aegypti expressing DsRed had no effects on life table parameters of all life stages of Toxorhynchites splendens and Tx. amboinensis. Following ERAs, including assessments of toxicity and allergenicity, environmental releases of these transgenic mosquitoes were approved by regulatory authorities in Brazil and Malaysia and the USA [174,175,176,177,178]. Likewise, the Dutch GMO Office considered there were negligible risks to human health and the environment from releases of these transgenic mosquitoes expressing DsRed . In addition, DsRed is genetic element reference record in the Biosafety Clearing House of the Cartagena Protocol on Biosafety of the Convention on Biological Diversity (https://bch.cbd.int/en/database/BCH-GENE-SCBD-101476-6), supporting a history of its safe use in GMO applications.
Bioinformatic assessment of DsRed allergenicity
NCBI TRS and EST database searches yielded no entries with an E-value below 10. Searches of BLAST Uniprot-Allergen protein databases yielded two entries with E-values below 10 but neither had significant sequence alignments (see Table 3). Using the COMPASS to search the COMPARE database, 64 entries were identified with E values below 10, but none had 35% or more sequence identity with DsRed over 80 or more amino acids. Using FASTA36 to search the AllegenOnline database, nine entries were identified with E values below 10, but none had 35% or more sequence identity with DsRed over 80 or more amino acids. For 80 sliding and 8-mer searches in COMPARE and AllergenOnline, no significant sequence alignments were identified for DsRed.
Literature assessment of DsRed allergenicity
No studies were found in which DsRed was implicated in allergenicity. In a single study by Matsuoka et al. 2012, where transgenic Anopheles stephensi mosquitoes were created to express DsRed in saliva, authors demonstrated repeated feedings inducing an anti-DsRed antibodies in mice . When having “immunized” mice with injected DsRed, upon exposure to bites from the transgenic DsRed positive mosquito, the mice produced a higher titre of antibody to DsRed, suggesting that the transgenic mosquito bites acted as a “booster” effect.
Discosoma spp. toxicity
Literature assessment of Discosoma spp. toxicity
Literature searches of Discosoma found no evidence from previous studies of potential toxicity. Long et al.  reported that high levels of expression of DsRed from Discosoma did not affect cell morphology, developmental potential or viability and fertility of mice. Barbier and Damron  reported no changes in viability or growth fitness in bacteria expressing DsRed.
Discosoma spp. allergenicity
Literature assessment of Discosoma spp. allergenicity
For the Discosoma species, no evidence from previous studies was found indicating a potential for allergenicity.
Streptococcus pyogenes toxicity
Literature assessment of S. pyogenes toxicity
There are numerous toxic effects associated with various components of S. pyogenes such as excretory toxins, cell wall components or superantigens, many of which work in concordance with one another and none of which have been directly attributed to Cas9 expression or activity. These components are outlined in more detail in Additional file 2: Tables S1, S2 and S3.
Streptococcus pyogenes allergenicity
Literature assessment of S. pyogenes allergenicity
Streptococcus pyogenes components can elicit immune responses including acute reactivation or induction of arthritis [183,184,185]; hyperinflammatory responses [186, 187] as well as one study showing evidence of anaphylactoid response limited to rabbits only, although this effect was not observed in mice . However, none of these have been directly attributed to Cas9.
T1 gRNA allergenicity and toxicity
Bioinformatic assessment of T1 gRNA
No significant similarities with an E value below 0.01 were found between the T1 gRNA and miRNAs from human, Anopheles, or any other species in the miRbase database.
Literature assessment of RNA toxicity and allergenicity
Kim et al.  showed that gRNAs can trigger RNA-sensing, type 1 interferon immune responses in cell cultures leading to cytotoxicity. Moreover, it is known that there are numerous short RNAs present in wild type mosquito saliva, which likely get transferred into human tissues during biting , which could, therefore, theoretically lead to immune responses. However, human intradermal and intramuscular exposure to RNA from therapeutic interventions such as mRNA-based vaccinations against COVID-19 have a proven safety profile for use in humans [191,192,193]. Moreover, RNA is subject to fast breakdown in vivo and does not integrate into the genome  and is only transiently expressed if at all . Indeed, several therapies using in vitro transcribed mRNA are currently being actively investigated in clinical trials and have demonstrated excellent tolerability and safety indicating that mRNA therapies as a class of intervention have no inherent risks [196, 197].
Hazard characterization of toxicity and allergenicity from dsxF CRISPRh
In bioinformatic assessments of transgenic products expressed from the dsxFCRISPRh transgene, there was no significant amino acid similarity between DsRed and known allergens or toxins. hCas9 showed no evidence of identity with proteins known to be toxic to eukaryotes. While bioinformatic searches did identify homology between hCas9 and the VapC protein, a weight of evidence approach was applied in a tiered evaluation of hCas9 toxicity in humans and livestock [12,13,14]. Firstly, in terms of mode of action, VapC is restricted to prokaryotes via its inhibitory effects on mRNA translation, and thus cell growth, specifically in bacteria. Secondly, in terms of amino acid sequence similarity, of the 117 amino acids of VapC , 33 and 73 align within the HNH and RuvC-III domains found between amino acids 805-964 and 965-1141 of hCas9, respectively  (see Fig. 1). However, VapC does not show any sequence similarity to RuvC-I, alpha-helical lobe, RuvC-II, topo-homology and C-terminal domains found in Cas9 . The two proteins therefore only share homology to a minority of the total Cas9 protein. Thirdly, in terms of history of safe use of the protein or source organism, VapC shows sequence identity to a wide range of Cas9 homologues in heterogenous bacterial species that are widely distributed throughout nature . Fourthly, in terms of probability and level of exposure to any potential toxicity, the expression of Cas9 in dsxFCRISPRh mosquitoes is controlled by the germline promoter zpg and is therefore not expected to be expressed in the saliva of dsxFCRISPRh transgenics  so that any exposure via this route would be negligible. Moreover, the levels of Cas9 transgenic protein in any given individual dsxFCRISPRh mosquito would be insignificant in terms of biomass within the environment . Hence, applying a weight of evidence approach, it was concluded that the sequence similarity identified between VapC and Cas9 required no further investigation to support the conclusion that the risk of toxicity from Cas9 expression in dsxFCRISPRh transgenics is negligible.
Bioinformatic analysis of the T1 gRNA expressed in dsxFCRISPRh GMMs indicated no sequence similarity to known microRNAs from other species, including humans and Anopheles. Coupled with literature assessment demonstrating low risk from intradermal exposure to RNA, there was no evidence to suggest that the presence of the gRNA could be toxic or allergenic, for example, by interfering host gene expression.
Most of the literature identified from searches of Cas9 and DsRed described in vitro investigations of gain or loss of function of heterologous proteins, development of fluorescent reporter systems, assay development or method validation, typically providing no direct evidence indicating toxicity or allergenicity of the transgenic proteins themselves. In assessments of toxicity and allergenicity of source organisms, while there was no evidence from previous studies of allergenicity or toxicity associated with Discosoma spp., S. pyogenes showed some evidence of toxicity, as well as allergenicity, but this was not associated directly with Cas9 activity.
The strength of evidence supporting conclusions on allergenicity or toxicity of transgenic products in humans from in vitro, in vivo and clinical studies was an important consideration in literature evaluations. Results from in vitro studies may not necessarily reliably translate into the in vivo biological responses of organisms. For example, for DsRed, expression of some of its variants can result in cytotoxicity in some cell types in vitro. However, in several in vivo transgenic animal models, expression of the DsRed variant does not result in toxicity [172, 198, 199]. For Cas9, intracellular, cell-specific in vitro toxicity has been observed, as well as toxicity when Cas9 is expressed at high levels in vivo [81, 82, 105]. Importantly, Garrood et al.  reported that Cas9 off-target mutagenic effects could not be detected in the genome of dsxFCRISPRh GMM. This is consistent with finding that Cas9 constitutive intracellular expression can be toxic to algae, fungi and parasitic organisms [98,99,100], but its transient expression in these organisms does not produce toxicity. In one in vivo case, zebrafish muscle fibre toxicity concomitant with expression of Cas9 was not directly attributed to Cas9 . In another study looking at nucleotide insertions and indels in vivo, Cas9 knock-in mice exhibited only low levels of indels . However, there were also numerous in vitro as well as in vivo studies of Cas9 providing no evidence of toxicity. Moreover, recent clinical trials involving Cas9 that target disparate organs of the body indicate a lack of any significant associated toxicity [122, 125, 126, 128, 132].
Moreover, toxicity was commonly absent from Cas9 in vivo studies and, where it was detected, this was typically associated with high levels of Cas9 expression or concentrations in vivo and absent at lower concentrations [68, 105, 108,109,110, 117, 200]. Furthermore, Cas9 has been administered in vivo to patients in a number of clinical trials without significant safety concerns [97, 128, 132] This evidence of safe human therapeutic exposure to Cas9, coupled with bioinformatic analyses finding no significant sequence identity matches between hCas9 and any toxic proteins, indicates that exposure to hCas9 from dsxFCRISPRh GMMs is unlikely to cause toxicity in vivo in humans or other mammals.
In bioinformatic searches, hCas9 was also found to share a single eight amino acid epitope within the latex allergen Hev b 9, from Hevea brasiliensis [201, 202]. This contrasts with the results of Nakajima et al. . Although they found one epitope to the food allergen serine carboxypeptidase 2 in Triticum aestivum in a soybean codon-optimized Cas9, they did not identify the Hev b 9 epitope in their searches with human codon-optimized Cas9. They used the AllergenOnline database, as was the case in this study, and the amino acid sequence of hCas9 that was used in this study was also used by Nakajima et al. . It may be that this epitope has only been identified or deposited in databases since the searches of Nakajima et al. , which would have occurred at least five years before our searches. In any event, the present study identified the single eight amino acid epitope within the latex allergen Hev b 9 from searches of both COMPARE and AllergenOnline databases with hCas9 amino acid sequence. However, it has been reported that for any cross-reactivity to occur, in addition to the shared epitope, a high degree of similarity (between > 50 and 70%) is needed over a segment of 80 or greater amino acids [33, 146, 147], and no such significant sequence identity match was found between hCas9 and Hev b 9. The single epitope match identified between hCas9 and Hev b 9 can therefore be considered as unlikely to pose an allergenic risk. This conclusion is reinforced by the evidence of widespread exposure of humans and other animals to bacteria expressing Cas9 variants. Acquired immune responses to Cas9 are widespread in humans and have been well documented, with antibodies and T cells against SpCas9 detected in 58% and 67% of donors, respectively [157,158,159,160].
Exposure characterization of toxicity and allergenicity from dsxF CRISPRh
Key considerations in the ERA for toxicity and allergenicity of transgenic proteins involve routes and levels of exposure. Although GMMs would not be actively consumed by humans, accidental incidental ingestion or inhalation of mosquitoes, or their body parts, may occur but is still likely to be a negligible source of human exposure to transgenic proteins. Because of the restricted temporal and spatial expression of the promoter driving expression of Cas9, its expression should be restricted to the germline in males and females from late pupal stages; and the levels of Cas9 transgenic protein in any given individual mosquito would be negligible in terms of biomass . Even if accidental ingestion of minimal amounts of material were to occur, Cas9 is subject to pepsin digestion  and, therefore, exposure in the body to Cas9 through accidental ingestion of a GMM, would be negligible.
However, female mosquitoes require mammalian, in some cases preferentially human, blood to produce viable eggs and offspring. Therefore, the substantive route of human exposure to mosquito proteins is intradermal via the blood-feeding of female mosquitoes and local allergic reactions to saliva from mosquito bites, which have been observed in humans [203,204,205]. In dsxFCRISPRh GMMs, homozygous females are unable to bite, limiting the risk of exposure to transgenic products via saliva from the bites from heterozygous females. In addition, Cas9 is expressed under the control of the germline promoter zpg and DsRed under the control of an ocular and nervous system promoter 3X3P . Therefore, neither protein should be expressed in the salivary glands, nor be present in the saliva, of dsxFCRISPRh GMMs. By contrast, in the dsxFCRISPRh transgene, the T1 gRNA is under the control of the ubiquitous and constitutive U6 promoter. It is, therefore, expected that the T1 gRNA transgenic product would be present in the saliva of these transgenic mosquitoes.
The nuclease expressed from the dsxFCRISPRh transgene is a human codon-optimized version of Cas9 that originates from S. pyogenes¸ a dominant member of Group A Streptococci (GAS) [7, 206]. Streptococcus pyogenes is commonly found throughout the world in the throat, skin and anogenital tract of humans, including infants and healthy individuals. Approximately 20% of children are chronic carriers of GAS . Shaikh et al.  reported that prevalence of Group A Streptococci in children presenting with pharyngitis was 37%, compared with 12% in those with no symptoms. Another study reported oropharyngeal colonization by S. pyogenes in 9.6% of young adults . Carapetis et al.  have reported a global disease burden of over 111 million cases of GAS pyoderma, and over 616 million incident cases per year of GAS pharyngitis. Human exposure to S. pyogenes Cas9 (SpCas9) is thus common and frequent, with development of acquired immune responses [157,158,159,160]. These studies provide evidence of widespread pre-existing human exposure to the Cas9 protein. Moreover, in food toxicity evaluations, if a modified protein retains biological function, and this function is found in related proteins with a history of safe use in food, and the exposure level is similar to functionally related proteins, then the modified protein could also be considered to be ‘as-safe-as’ those that have the history of safe use . It has been proposed that preexisting immunity to Cas9 may decrease its efficacy in vivo in a therapeutic context [212, 213] however a population’s prior exposure to Cas9 orthologues from pathogenic bacteria such as S. thermophilus may prompt effective recognition, targeting by cytotoxic T cells and ultimately clearance of any cells expressing Cas9 .
These results indicate generation of an immune response to SpCas9, wherein subsequent exposure to the protein would result in antibodies binding to and rendering the protein inactive, or formation of immune complexes that are actively cleared by the circulation , which would be favourable in case of GMMs expressing Cas9 protein, or should the protein itself be exposed to the immune system. The conclusion that a single epitope match between hCas9 and Hev b 9 is unlikely to pose an allergenic risk is therefore further reinforced by the evidence of widespread exposure of humans and other animals to bacteria expressing Cas9 variants.
Evaluation of potential toxicity or allergenicity from dsxF CRISPRh mosquitoes to organisms in the environment other than humans
Although bioinformatic and literature analyses of potential toxicity and allergenicity of dsxFCRISPRh transgenic products focused primarily on human health protection goals, they are nonetheless also relevant to organisms in the environment other than humans. Five of the 46 identified plausible pathways to potential harm from field releases of mosquitoes in West Africa could result in negative impacts on livestock, animals or other non-target organisms in the environment as a result of toxicological effects . As with pathways involving toxicological or allergenic effects on humans, a tiered approach based on weight-of-evidence has been applied to evaluation of potential toxicity or allergenicity from dsxFCRISPRh mosquitoes to organisms in the environment other than humans [12, 13, 16, 214]. This allows the procurement of evidence only when it might directly reduce uncertainties in the ERA to prevent unwarranted, wasteful or uninformative investigations .
In terms of mode of action, many of the previous studies identified here draw on toxicological impacts Cas9 and DsRed transgenic proteins in animals. For example, expression of Cas9 in vivo in mice resulted in no off-target mutagenesis or evidence of toxicity  and oral ingestion of transgenic Ae. aegypti expressing DsRed had no effects on life table parameters of all life stages of Tx. splendens and Tx. amboinensis . Hence, hazard assessments of toxicological effects and mode of action from dsxFCRISPRh on humans should have relevance to impacts on other animals in the environment. Importantly, VapC toxicity is restricted to prokaryotes via specific inhibition of bacterial mRNA translation. Considering amino acid sequence similarity, although VapC aligns within the HNH and RuvC-III domains of hCas9, it does not show any sequence similarity to RuvC-I, alpha-helical lobe, RuvC-II, topo-homology and C-terminal domains  so that the two proteins therefore only share homology to a minority of the total Cas9 protein.
In terms of probability and level of exposure to any potential toxicity, as the dsxFCRISPRh transgene is expected to introgress into sibling species of the An. gambiae complex that are zoophilic [17, 215], it is possible that some non-human animal could be exposed to transgenic products via biting as dsxFCRISPRh mosquitoes attempt to obtain blood meals. However, as is the case for human exposure, neither Cas9 nor DsRed is expected to be present in the saliva of dsxFCRISPRh mosquitoes; non-human animal exposure via this route would therefore be negligible. However, even though levels of Cas9 transgenic protein in any given individual dsxFCRISPRh mosquito would be insignificant in terms of biomass within the environment , ingestion of dsxFCRISPRh mosquitoes could be a more substantive route of exposure for predator or scavenger non-target organisms in the environment than for humans. Foraging theory predicts that predation of adult mosquitoes would be energetically unfavourable for large vertebrates, such as birds or bats [55, 216]. Species of the An. gambiae complex also tend to favour oviposition in aquatic habitats such as temporary and shallow puddles, irrigation channels of rice paddies or slow-moving water by riverbanks that are less likely to support colonization by many vertebrate predators . Consumption of aquatic stages of these mosquitoes tends to be opportunistic and by generalist predators. Consequently, there is no evidence that any species preys exclusively on mosquitoes of the Anopheles genus . Hence, exposure of generalist predators to dsxFCRISPRh transgenic products is expected to be low, but not negligible. Nonetheless, considering the “history of safe use of the protein or source organism”, as is the case for humans, any organisms in the environment would also be exposed throughout their lifetimes, including via consumption, to a wide range of Cas9 homologues in heterogenous bacterial species that are widely distributed throughout nature .
Applying a weigh-of-evidence approach to the above evidence, therefore, any additional contribution from transgenic Cas9 to overall exposure to from Cas9 transgenic proteins should be considered as negligible, as would be the toxicological or allergenic hazard. Hence, overall risk of toxicity or allergenicity from releases of dsxFCRISPRh mosquitoes to organisms in the environment other than humans would be negligible.
Previously, the first stage of an ERA, problem formulation, was conducted to identify potential harms from simulated releases of dsxFCRISPRh in West Africa as a malaria vector control intervention . That analysis identified 46 plausible pathways to potential harm, nine of which could be evaluated by testing the risk hypothesis that dsxFCRISPRh transgenic products would not cause increased allergenicity or toxicity. The present study formed part of the next stages of that ERA where those risk hypotheses were tested. Amino acid sequences of DsRed and hCas9 were interrogated against those of known toxins or allergens from four different bioinformatic databases and literature was examined for any evidence of toxicity or allergenicity of the transgenic products themselves, or of the donor organisms from which they were originally derived. Potential exposure to dsxFCRISPRh transgenic proteins from environmental releases was assessed as negligible. Nonetheless, Cas9 nuclease activity can be toxic to some cell types in vitro and hCas9 was found to share homology with the prokaryotic toxin VapC. However, there was no evidence from previous studies indicating a risk of toxicity to humans and other animals. hCas9 was also found to contain an 8-mer epitope shared with the latex allergen Hev b 9. However, the full amino acid sequence of hCas9 was not homologous to any known allergens. Combined with a lack of evidence in the literature of Cas9 allergenicity, this indicated negligible risk of allergenicity from hCas9. No matches were found between the T1 gRNA and microRNAs from Anopheles or humans. Therefore, taking into account the weight of evidence on transgenic products (i) modes of actions (ii) sequence similarity to known or putative toxins or allergens, (iii) history of safe use, or the safe use of source organisms, and (iv) probability and level of exposure, there was no convincing evidence from previous studies to suggest that transgenic products expressed from dsxFCRISPRh would be allergenic or toxic. These data supported acceptance of the risk hypothesis that dsxFCRISPRh transgenic products would not cause increased allergenicity or toxicity, allowing the rejection of nine of a total of 46 plausible pathways to potential harm, namely pathways 1, 9, 11, 12, 16, 17, 28, 31 and 44 . Hence, environmental releases in West Africa of the dsxFCRISPRh population suppression gene drive for malaria vector control should not result in any increased allergenicity or toxicity in humans and animals. These results should also inform evaluations of the field use of other GMM strains for control of vectors of malaria and other infectious diseases, as well as pre-clinical risk assessments of in vivo clinical applications of CRISPR-Cas9.
Availability of data and materials
Comprehensive Protein Allergen Resource Database
COMPare Analysis of Sequences with Software Program
European Food Safety Authority
European Molecular Biology Laboratory’s European Bioinformatics Institute
Environmental risk assessment
Expressed sequence tags
Group A Streptococci
Genetically modified organism
Genetically modified mosquito
Human codon-optimized version of CRISPR-Cas9 endonuclease, derived from SpCas9
Homology directed repair
Microhomology-mediated end joining
National Centre for Biotechnology
Non-homologous end joining
Natural rubber latex allergy
Organization of Economic Cooperation and Development
Swiss Institute of Bioinformatics and Protein Information Resource
Cas9 of S. pyogenes
Transcript reference sequences
World Health Organization
WHO. World malaria report. Geneva: World Health Organization; 2021.
African Union Development Agency—New Partnership for Africa's Development. Gene Drives for Malaria Control and Elimination in Africa; 2018.
WHO. Evaluation of genetically modified mosquitoes for the control of vector-borne diseases. Geneva: World Health Organization; 2020.
Burt A. Site-specific selfish genes as tools for the control and genetic engineering of natural populations. Proc Biol Sci. 2003;270:921–8.
Burt A, Coulibaly M, Crisanti A, Diabate A, Kayondo JK. Gene drive to reduce malaria transmission in sub-Saharan Africa. J Responsible Innov. 2018;5:S66–80.
Hammond AM, Galizi R. Gene drives to fight malaria: current state and future directions. Pathog Glob Health. 2018;111:412–3.
Kyrou K, Hammond AM, Galizi R, Kranjc N, Burt A, Beaghton AK, et al. A CRISPR-Cas9 gene drive targeting doublesex causes complete population suppression in caged Anopheles gambiae mosquitoes. Nat Biotechnol. 2018;36:1062–6.
European Food Safety Authority, Naegeli H, Bresson JL, Dalmay T, Dewhurst IC, Epstein MM, et al. Adequacy and sufficiency evaluation of existing EFSA guidelines for the molecular characterisation, environmental risk assessment and post-market environmental monitoring of genetically modified insects containing engineered gene drives. EFSA J. 2020;18:e06297.
WHO. Guidance framework for testing genetically modified mosquitoes. 2nd ed. Geneva: World Health Organization; 2021.
National Academies of Sciences, Engineering, and Medicine. Gene drives on the horizon: advancing science, navigating uncertainty, and aligning research with public values. Washington, DC: The National Academies Press; 2016.
Connolly JB, Mumford JD, Glandorf DCM, Hartley S, Lewis OT, Evans SW, et al. Recommendations for environmental risk assessment of gene drive applications for malaria vector control. Malar J. 2022;21:152.
Codex Alimentarius Commission. Foods derived from modern biotechnology, second edition. Principles for the risk analysis of foods derived from modern biotechnology (CAG/GL 44-3003) and Guidelines for the conduct of food safety assessment of foods derived from recombinant-DNA animals (CAG/GL 68-2008). World Health Organization, Food and Agriculture Organization of the United Nations, Rome 2009.
European Food Safety Authority. Scientific Opinion on the Guidance on the risk assessment of food and feed from genetically modified animals and animal health and welfare aspects. EFSA J. 2012;10:2501.
Bauman PA, Doxey AC, Eberini I, Islamovic E, Jungo F, Kessenich C, et al. “From Protein Toxins to Applied Toxicological Testing” virtual workshop identifies the need for a bioinformatic framework to assess novel food protein safety. Regul Toxicol Pharmacol. 2022;131:105146.
Kessenich C, Silvanovich A. Challenges of automation and scale: bioinformatics and the evaluation of proteins to support genetically modified product safety assessments. J Invertebr Pathol. 2021;186:107587.
Hardy A, Benford D, Halldorsson T, Jeger MJ, Knutsen HK, More S, et al. Guidance on the use of the weight of evidence approach in scientific assessments. EFSA J. 2017;15:e04971.
Connolly JB, Mumford JD, Fuchs S, Turner G, Beech C, North AR, et al. Systematic identification of plausible pathways to potential harm via problem formulation for investigational releases of a population suppression gene drive to control the human malaria vector Anopheles gambiae in West Africa. Malar J. 2021;20:170.
Tao T. Standalone BLAST Setup for Windows PC. 2010 May 31 [Updated 2020 Aug 31]. In: BLAST® Help [Internet]. Bethesda (MD): National Center for Biotechnology Information (US); 2008. https://www.ncbi.nlm.nih.gov/books/NBK52637/. Accessed 16 June 2023
Rost B. Twilight zone of protein sequence alignments. Protein Eng. 1999;12:85–94.
Baker D, Sali A. Protein structure prediction and structural genomics. Science. 2001;294:93–6.
Blake JD, Cohen FE. Pairwise sequence alignment below the twilight zone. J Mol Biol. 2001;307:721–35.
Negi SS, Schein CH, Ladics GS, Mirsky H, Chang P, Rascle JB, et al. Functional classification of protein toxins as a basis for bioinformatic screening. Sci Rep. 2017;7:13940.
Chothia C, Lesk AM. The relation between the divergence of sequence and structure in proteins. EMBO J. 1986;5:823–6.
Abagyan RA, Batalov S. Do aligned sequences share the same fold? J Mol Biol. 1997;273:355–68.
Sillitoe I, Lewis TE, Cuff A, Das S, Ashford P, Dawson NL, et al. CATH: comprehensive structural and functional annotations for genome sequences. Nucleic Acids Res. 2015;43:D376-381.
Herman RA, Song P, Thirumalaiswamysekhar A. Value of eight-amino-acid matches in predicting the allergenicity status of proteins: an empirical bioinformatic investigation. Clin Mol Allergy. 2009;7:9.
Xu D, Nussinov R. Favorable domain size in proteins. Fold Des. 1998;3:11–7.
Wheelan SJ, Marchler-Bauer A, Bryant SH. Domain size distributions can predict domain boundaries. Bioinformatics. 2000;16:613–8.
Griffiths-Jones S. miRBase: the microRNA sequence database. Methods Mol Biol. 2006;342:129–38.
WHO. Chemical safety and health unit, emergencies preparedness. Manual for the public health management of chemical incidents. Geneva: World Health Organization; 2009.
Hollingworth RM, Bjeldanes LF, Bolger M, Kimber I, Meade BJ, Taylor SL, et al. The safety of genetically modified foods produced through biotechnology. Toxicol Sci. 2003;71:2–8.
European Food Safety Authority. Scientific opinion on the assessment of allergenicity of GM plants and microorganisms and derived food and feed. EFSA J. 2010;8:1700.
Goodman RE. Performing IgE serum testing due to bioinformatics matches in the allergenicity assessment of GM crops. Food Chem Toxicol. 2008;46(Suppl 10):S24-34.
European Food Safety Authority. Safety and nutritional assessment of GM plants and derived food and feed: the role of animal feeding trials. Food Chem Toxicol. 2008;46(Suppl 1):S2-70.
Devos Y, Aguilera J, Diveki Z, Gomes A, Liu Y, Paoletti C, et al. EFSA’s scientific activities and achievements on the risk assessment of genetically modified organisms (GMOs) during its first decade of existence: looking back and ahead. Transgenic Res. 2014;23:1–25.
European Food Safety Authority. Administrative guidance for the preparation of applications on genetically modified plants. EFSA Supp Publ. 2021;18:6473E.
Kearns PWE, Kleter GA, Bergmans HEN, Kuiper HA. Biotechnology and biosafety policy at OECD: future trends. Trends Biotechnol. 2021;39:965–9.
Winther KS, Brodersen DE, Brown AK, Gerdes K. VapC20 of Mycobacterium tuberculosis cleaves the sarcin-ricin loop of 23S rRNA. Nat Commun. 2013;4:2796.
Winther K, Tree JJ, Tollervey D, Gerdes K. VapCs of Mycobacterium tuberculosis cleave RNAs essential for translation. Nucleic Acids Res. 2016;44:9860–71.
Dienemann C, Bøggild A, Winther KS, Gerdes K, Brodersen DE. Crystal structure of the VapBC toxin–antitoxin complex from Shigella flexneri reveals a hetero-octameric DNA-binding assembly. J Mol Biol. 2011;414:713–22.
Winther KS, Gerdes K. Enteric virulence associated protein VapC inhibits translation by cleavage of initiator tRNA. Proc Natl Acad Sci USA. 2011;108:7403–7.
Umana A, Lemkul JA, Slade DJ. Complete genome sequence of Fusobacterium necrophorum subsp. necrophorum ATCC 25286. Microbiol Resour Announc. 2019;8:e00025-e119.
Jinek M, Jiang F, Taylor DW, Sternberg SH, Kaya E, Ma E, et al. Structures of Cas9 endonucleases reveal RNA-mediated conformational activation. Science. 2014;343:1247997.
Bhaya D, Davison M, Barrangou R. CRISPR-Cas systems in bacteria and archaea: versatile small RNAs for adaptive defense and regulation. Annu Rev Genet. 2011;45:273–97.
Tan ZL, Nagaraja TG, Chengappa MM. Fusobacterium necrophorum infections: virulence factors, pathogenic mechanism and control measures. Vet Res Commun. 1996;20:113–40.
Aliyu SH, Marriott RK, Curran MD, Parmar S, Bentley N, Brown NM, et al. Real-time PCR investigation into the importance of Fusobacterium necrophorum as a cause of acute pharyngitis in general practice. J Med Microbiol. 2004;53:1029–35.
Batty A, Wren MW. Prevalence of Fusobacterium necrophorum and other upper respiratory tract pathogens isolated from throat swabs. Br J Biomed Sci. 2005;62:66–70.
Batty A, Wren MW, Gal M. Fusobacterium necrophorum as the cause of recurrent sore throat: comparison of isolates from persistent sore throat syndrome and Lemierre’s disease. J Infect. 2005;51:299–306.
Harms A, Brodersen DE, Mitarai N, Gerdes K. Toxins, targets, and triggers: an overview of toxin-antitoxin biology. Mol Cell. 2018;70:768–84.
Gerdes K, Rasmussen PB, Molin S. Unique type of plasmid maintenance function: postsegregational killing of plasmid-free cells. Proc Natl Acad Sci USA. 1986;83:3116–20.
Dy RL, Przybilski R, Semeijn K, Salmond GP, Fineran PC. A widespread bacteriophage abortive infection system functions through a Type IV toxin-antitoxin mechanism. Nucleic Acids Res. 2014;42:4590–605.
Dy RL, Richter C, Salmond GP, Fineran PC. Remarkable mechanisms in microbes to resist phage infections. Annu Rev Virol. 2014;1:307–31.
Harms A, Maisonneuve E, Gerdes K. Mechanisms of bacterial persistence during stress and antibiotic exposure. Science. 2016;354:aa4268.
Nakajima O, Nishimaki-Mogami T, Kondo K. Cas9 in genetically modified food is unlikely to cause food allergy. Biol Pharm Bull. 2016;39:1876–80.
Collins CM, Bonds JAS, Quinlan MM, Mumford JD. Effects of the removal or reduction in density of the malaria mosquito, Anopheles gambiae s.l., on interacting predators and competitors in local ecosystems. Med Vet Entomol. 2019;33:1–15.
Gasiunas G, Barrangou R, Horvath P, Siksnys V. Cas9-crRNA ribonucleoprotein complex mediates specific DNA cleavage for adaptive immunity in bacteria. Proc Natl Acad Sci USA. 2012;109:E2579–86.
Doench JG, Fusi N, Sullender M, Hegde M, Vaimberg EW, Donovan KF, et al. Optimized sgRNA design to maximize activity and minimize off-target effects of CRISPR-Cas9. Nat Biotechnol. 2016;34:184–91.
Kleinstiver BP, Pattanayak V, Prew MS, Tsai SQ, Nguyen NT, Zheng Z, et al. High-fidelity CRISPR-Cas9 nucleases with no detectable genome-wide off-target effects. Nature. 2016;529:490–5.
Dong Z, Wu S, Zhu C, Wang X, Li Y, Chen X, et al. Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)/Cas9-mediated kif15 mutations accelerate axonal outgrowth during neuronal development and regeneration in zebrafish. Traffic. 2019;20:71–81.
Zhang JP, Cheng XX, Zhao M, Li GH, Xu J, Zhang F, et al. Curing hemophilia A by NHEJ-mediated ectopic F8 insertion in the mouse. Genome Biol. 2019;20:276.
Musunuru K. CRISPR hits home in a first-in-human study. CRISPR J. 2021;4:460–1.
Musunuru K, Chadwick AC, Mizoguchi T, Garcia SP, DeNizio JE, Reiss CW, et al. In vivo CRISPR base editing of PCSK9 durably lowers cholesterol in primates. Nature. 2021;593:429–34.
An HW, Kim SY, Kwon JW, Seok SH, Woo SH, Kim DY, et al. In vivo CRISPR-Cas9 knockout screening using quantitative PCR identifies thymosin beta-4 X-linked that promotes diffuse-type gastric cancer metastasis. Mol Carcinog. 2021;60:597–606.
Wang M, Zuris JA, Meng F, Rees H, Sun S, Deng P, et al. Efficient delivery of genome-editing proteins using bioreducible lipid nanoparticles. Proc Natl Acad Sci USA. 2016;113:2868–73.
Engelholm LH, Riaz A, Serra D, Dagnaes-Hansen F, Johansen JV, Santoni-Rugiu E, et al. CRISPR/Cas9 engineering of adult mouse liver demonstrates that the Dnajb1-Prkaca gene fusion is sufficient to induce tumors resembling fibrolamellar hepatocellular carcinoma. Gastroenterology. 2017;153:1662–73.
Xu L, Lau YS, Gao Y, Li H, Han R. Life-long AAV-mediated CRISPR genome editing in dystrophic heart improves cardiomyopathy without causing serious lesions in mdx mice. Mol Ther. 2019;27:1407–14.
Zhang L, Wang L, Xie Y, Wang P, Deng S, Qin A, et al. Triple-targeting delivery of CRISPR/Cas9 to reduce the risk of cardiovascular diseases. Angew Chem Int Ed Engl. 2019;58:12404–8.
Qiu M, Glass Z, Chen J, Haas M, Jin X, Zhao X, et al. Lipid nanoparticle-mediated codelivery of Cas9 mRNA and single-guide RNA achieves liver-specific in vivo genome editing of Angptl3. Proc Natl Acad Sci USA. 2021;118:e2020401118.
Giannoukos G, Ciulla DM, Marco E, Abdulkerim HS, Barrera LA, Bothmer A, et al. UDiTaS, a genome editing detection method for indels and genome rearrangements. BMC Genomics. 2018;19:212.
Kosicki M, Tomberg K, Bradley A. Repair of double-strand breaks induced by CRISPR-Cas9 leads to large deletions and complex rearrangements. Nat Biotechnol. 2018;36:765–71.
Song Y, Liu Z, Zhang Y, Chen M, Sui T, Lai L, et al. Large-fragment deletions induced by Cas9 cleavage while not in the BEs system. Mol Ther Nucleic Acids. 2020;21:523–6.
Stadtmauer EA, Fraietta JA, Davis MM, Cohen AD, Weber KL, Lancaster E, et al. CRISPR-engineered T cells in patients with refractory cancer. Science. 2020;367:eaba7365.
Alanis-Lobato G, Zohren J, McCarthy A, Fogarty NME, Kubikova N, Hardman E, et al. Frequent loss of heterozygosity in CRISPR-Cas9-edited early human embryos. Proc Natl Acad Sci USA. 2021;118:e2004832117.
Leibowitz ML, Papathanasiou S, Doerfler PA, Blaine LJ, Sun L, Yao Y, et al. Chromothripsis as an on-target consequence of CRISPR-Cas9 genome editing. Nat Genet. 2021;53:895–905.
Turchiano G, Andrieux G, Klermund J, Blattner G, Pennucci V, El Gaz M, et al. Quantitative evaluation of chromosomal rearrangements in gene-edited human stem cells by CAST-Seq. Cell Stem Cell. 2021;28:1136–47.
Raguram A, Banskota S, Liu DR. Therapeutic in vivo delivery of gene editing agents. Cell. 2022;185:2806–27.
Kosicki M, Allen F, Steward F, Tomberg K, Pan Y, Bradley A. Cas9-induced large deletions and small indels are controlled in a convergent fashion. Nat Commun. 2022;13:3422.
Lans H, Vermeulen W. Tissue specific response to DNA damage: C. elegans as role model. DNA Repair (Amst). 2015;32:141–8.
Hustedt N, Durocher D. The control of DNA repair by the cell cycle. Nat Cell Biol. 2016;19:1–9.
Ferreira da Silva J, Meyenberg M, Loizou JI. Tissue specificity of DNA repair: the CRISPR compass. Trends Genet. 2021;37:958–62.
Ihry RJ, Worringer KA, Salick MR, Frias E, Ho D, Theriault K, et al. p53 inhibits CRISPR-Cas9 engineering in human pluripotent stem cells. Nat Med. 2018;24:939–46.
Li C, Psatha N, Gil S, Wang H, Papayannopoulou T, Lieber A. HDAd5/35(++) adenovirus vector expressing anti-CRISPR peptides decreases CRISPR/Cas9 toxicity in human hematopoietic stem cells. Mol Ther Methods Clin Dev. 2018;9:390–401.
Brunetti-Pierri N, Ng T, Iannitti D, Cioffi W, Stapleton G, Law M, et al. Transgene expression up to 7 years in nonhuman primates following hepatic transduction with helper-dependent adenoviral vectors. Hum Gene Ther. 2013;24:761–5.
Hsu PD, Scott DA, Weinstein JA, Ran FA, Konermann S, Agarwala V, et al. DNA targeting specificity of RNA-guided Cas9 nucleases. Nat Biotechnol. 2013;31:827–32.
Mali P, Aach J, Stranges PB, Esvelt KM, Moosburner M, Kosuri S, et al. CAS9 transcriptional activators for target specificity screening and paired nickases for cooperative genome engineering. Nat Biotechnol. 2013;31:833–8.
Fu Y, Sander JD, Reyon D, Cascio VM, Joung JK. Improving CRISPR-Cas nuclease specificity using truncated guide RNAs. Nat Biotechnol. 2014;32:279–84.
Schiml S, Fauser F, Puchta H. The CRISPR/Cas system can be used as nuclease for in planta gene targeting and as paired nickases for directed mutagenesis in Arabidopsis resulting in heritable progeny. Plant J. 2014;80:1139–50.
Wang Y, Cheng X, Shan Q, Zhang Y, Liu J, Gao C, et al. Simultaneous editing of three homoeoalleles in hexaploid bread wheat confers heritable resistance to powdery mildew. Nat Biotechnol. 2014;32:947–51.
Slaymaker IM, Gao L, Zetsche B, Scott DA, Yan WX, Zhang F. Rationally engineered Cas9 nucleases with improved specificity. Science. 2016;351:84–8.
Chen JS, Dagdas YS, Kleinstiver BP, Welch MM, Sousa AA, Harrington LB, et al. Enhanced proofreading governs CRISPR-Cas9 targeting accuracy. Nature. 2017;550:407–10.
Sugano SS, Nishihama R, Shirakawa M, Takagi J, Matsuda Y, Ishida S, et al. Efficient CRISPR/Cas9-based genome editing and its application to conditional genetic analysis in Marchantia polymorpha. PLoS ONE. 2018;13:e0205117.
Dilliard SA, Siegwart DJ. Disrupting off-target Cas9 activity in the liver. Nat Biomed Eng. 2022;6:106–7.
Piao X, Meng D, Zhang X, Song Q, Lv H, Jia Y. Dual-gRNA approach with limited off-target effect corrects C9ORF72 repeat expansion in vivo. Sci Rep. 2022;12:5672.
Yoon HH, Ye S, Lim S, Jo A, Lee H, Hong F, et al. CRISPR-Cas9 gene editing protects from the A53T-SNCA overexpression-induced pathology of Parkinson’s disease in vivo. CRISPR J. 2022;5:95–108.
Zhou J, Liu Y, Wei Y, Zheng S, Gou S, Chen T, et al. Eliminating predictable DNA off-target effects of cytosine base editor by using dual guiders including sgRNA and TALE. Mol Ther. 2022;30:2443–51.
Sago CD, Lokugamage MP, Loughrey D, Lindsay KE, Hincapie R, Krupczak BR, et al. Augmented lipid-nanoparticle-mediated in vivo genome editing in the lungs and spleen by disrupting Cas9 activity in the liver. Nat Biomed Eng. 2022;6:157–67.
Gillmore JD, Gane E, Taubel J, Kao J, Fontana M, Maitland ML, et al. CRISPR-Cas9 in vivo gene editing for transthyretin amyloidosis. N Engl J Med. 2021;385:493–502.
Jiang W, Brueggeman AJ, Horken KM, Plucinak TM, Weeks DP. Successful transient expression of Cas9 and single guide RNA genes in Chlamydomonas reinhardtii. Eukaryot Cell. 2014;13:1465–9.
Foster AJ, Martin-Urdiroz M, Yan X, Wright HS, Soanes DM, Talbot NJ. CRISPR-Cas9 ribonucleoprotein-mediated co-editing and counterselection in the rice blast fungus. Sci Rep. 2018;8:14355.
Markus BM, Bell GW, Lorenzi HA, Lourido S. Optimizing systems for Cas9 expression in Toxoplasma gondii. mSphere. 2019;4:e00386-e419.
Ferreira FJ, Carvalho L, Logarinho E, Bessa J. foxm1 modulates cell non-autonomous response in Zebrafish skeletal muscle homeostasis. Cells. 2021;10:1241.
Chew WL, Tabebordbar M, Cheng JK, Mali P, Wu EY, Ng AH, et al. A multifunctional AAV-CRISPR-Cas9 and its host response. Nat Methods. 2016;13:868–74.
Alvarez MM, Biayna J, Supek F. TP53-dependent toxicity of CRISPR/Cas9 cuts is differential across genomic loci and can confound genetic screening. Nat Commun. 2022;13:4520.
Friskes A, Koob L, Krenning L, Severson TM, Koeleman ES, Vergara X, et al. Double-strand break toxicity is chromatin context independent. Nucleic Acids Res. 2022;50:9930–47.
Holmgaard AB, Askou AL, Jensen EG, Alsing S, Bak RO, Mikkelsen JG, et al. Targeted knockout of the Vegfa gene in the retina by subretinal injection of RNP complexes containing Cas9 protein and modified sgRNAs. Mol Ther. 2021;29:191–207.
Garrood WT, Kranjc N, Petri K, Kim DY, Guo JA, Hammond AM, et al. Analysis of off-target effects in CRISPR-based gene drives in the human malaria mosquito. Proc Natl Acad Sci USA. 2021;118:e2004838117.
Reisman BJ, Guo H, Ramsey HE, Wright MT, Reinfeld BI, Ferrell PB, et al. Apoptolidin family glycomacrolides target leukemia through inhibition of ATP synthase. Nat Chem Biol. 2022;18:360–7.
Xiong J, Tan S, Yu L, Shen H, Qu S, Zhang C, et al. E7-targeted nanotherapeutics for key HPV afflicted cervical lesions by employing CRISPR/Cas9 and poly(beta-amino ester). Int J Nanomedicine. 2021;16:7609–22.
Han JP, Kim M, Choi BS, Lee JH, Lee GS, Jeong M, et al. In vivo delivery of CRISPR-Cas9 using lipid nanoparticles enables antithrombin gene editing for sustainable hemophilia A and B therapy. Sci Adv. 2022;8:eabj6901.
Wei R, Yang J, Cheng CW, Ho WI, Li N, Hu Y, et al. CRISPR-targeted genome editing of human induced pluripotent stem cell-derived hepatocytes for the treatment of Wilson’s disease. JHEP Rep. 2022;4:100389.
Wang X, Tang Y, Lu J, Shao Y, Qin X, Li Y, et al. Characterization of novel cytochrome P450 2E1 knockout rat model generated by CRISPR/Cas9. Biochem Pharmacol. 2016;105:80–90.
Li Q, Louden E, Zhou J, Drewlo S, Dai J, Puscheck EE, et al. Stress forces first lineage differentiation of mouse embryonic stem cells; validation of a high-throughput screen for toxicant stress. Stem Cells Dev. 2019;28:101–13.
Rodrigues M, McBride SW, Hullahalli K, Palmer KL, Duerkop BA. Conjugative delivery of CRISPR-Cas9 for the selective depletion of antibiotic-resistant enterococci. Antimicrob Agents Chemother. 2019;63:e01454-e1519.
Wang Y, Yin X, Zhou Z, Hu S, Li S, Liu M, et al. Cas9 regulated gene expression and pathogenicity in Riemerella anatipestifer. Microb Pathog. 2019;136:103706.
Huang D, Tran JT, Olson A, Vollbrecht T, Tenuta M, Guryleva MV, et al. Vaccine elicitation of HIV broadly neutralizing antibodies from engineered B cells. Nat Commun. 2020;11:5850.
Nahmad AD, Raviv Y, Horovitz-Fried M, Sofer I, Akriv T, Nataf D, et al. Engineered B cells expressing an anti-HIV antibody enable memory retention, isotype switching and clonal expansion. Nat Commun. 2020;11:5851.
Lucas CG, Redel BK, Chen PR, Spate LD, Prather RS, Wells KD. Effects of RAD51-stimulatory compound 1 (RS-1) and its vehicle, DMSO, on pig embryo culture. Reprod Toxicol. 2021;105:44–52.
Martinez-Turrillas R, Martin-Mallo A, Rodriguez-Diaz S, Zapata-Linares N, Rodriguez-Marquez P, San Martin-Uriz P, et al. In vivo CRISPR-Cas9 inhibition of hepatic LDH as treatment of primary hyperoxaluria. Mol Ther Methods Clin Dev. 2022;25:137–46.
Wang D, Zhang F, Gao G. CRISPR-based therapeutic genome editing: strategies and in vivo delivery by AAV vectors. Cell. 2020;181:136–50.
Ferrari S, Vavassori V, Canarutto D, Jacob A, Castiello MC, Javed AO, et al. Hopes and hurdles toward clinical translation. Front Genome Ed. 2021;3:618378.
European Food Safety Authority. Guidance on the environmental risk assessment of genetically modified animals. EFSA J. 2013;11:3200.
Ghaffari S, Khalili N, Rezaei N. CRISPR/Cas9 revitalizes adoptive T-cell therapy for cancer immunotherapy. J Exp Clin Cancer Res. 2021;40:269.
Foy SP, Jacoby K, Bota DA, Hunter T, Pan Z, Stawiski E, et al. Non-viral precision T cell receptor replacement for personalized cell therapy. Nature. 2023;615:687–96.
Ledford H. CRISPR cancer trial success paves the way for personalized treatments. Nature. 2022;611:433–4.
Liu Q. World-first Phase I clinical trial for CRISPR-Cas9 PD-1-edited t-cells in advanced nonsmall cell lung cancer. Glob Med Genet. 2020;7:73–4.
Bernard BE, Landmann E, Jeker LT, Schumann K. CRISPR/Cas-based human T cell engineering: basic research and clinical application. Immunol Lett. 2022;245:18–28.
Ottaviano G, Georgiadis C, Gkazi SA, Syed F, Zhan H, Etuk A, et al. Phase 1 clinical trial of CRISPR-engineered CAR19 universal T cells for treatment of children with refractory B cell leukemia. Sci Transl Med. 2022;14:eabq3010.
Frangoul H, Ho TW, Corbacioglu S. CRISPR-Cas9 gene editing for sickle cell disease and beta-thalassemia. Reply N Engl J Med. 2021;384:e91.
Vertex Pharmaceuticals. Vertex and CRISPR therapeutics present new data on more patients with longer follow-up treated with exagamglogene autotemcel (exa-cel) at the 2022 European Hematology Association (EHA) Congress. https://news.vrtx.com/news-releases/news-release-details/vertex-and-crispr-therapeutics-present-new-data-more-patients. Accessed 16 June 2023.
Locatelli FF, Corbacioglu S, de la Fuente J, Wall D, Domenica Capellini M, de Montalembert M, et al. Efficacy and safety of a single dose of CTX001 for transfusion-dependent beta-thalassemia and severe sickle cell disease. https://library.ehaweb.org/eha/2022/eha2022-congress/366210/franco.locatelli.efficacy.and.safety.of.a.single.dose.of.ctx001.for.html. Accessed 16 June 2023.
Kaiser J. CRISPR infusion eases symptoms in genetic disease. Science. 2022;377:1367.
Editas Medicine. Clinical data from editas medicine’s ongoing phase 1/2 BRILLIANCE clinical trial of EDIT-101 For LCA10 to be presented at the European Society of Gene and Cell Therapy Annual Congress. https://ir.editasmedicine.com/news-releases/news-release-details/clinical-data-editas-medicines-ongoing-phase-12-brilliance. Accessed 16 June 2023.
Kaiser J. Groundbreaking CRISPR treatment for blindness only works for subset of patients. https://www.science.org/content/article/groundbreaking-crispr-treatment-blindness-only-works-subset-patients. Accessed 16 June 2023.
Aalberse RC, Crameri R. IgE-binding epitopes: a reappraisal. Allergy. 2011;66:1261–74.
Sten E, Skov PS, Andersen SB, Torp AM, Olesen A, Bindslev-Jensen U, et al. A comparative study of the allergenic potency of wild-type and glyphosate-tolerant gene-modified soybean cultivars. APMIS. 2004;112:21–8.
Parisi CAS, Kelly KJ, Ansotegui IJ, Gonzalez-Diaz SN, Bilo MB, Cardona V, et al. Update on latex allergy: new insights into an old problem. World Allergy Organ J. 2021;14:100569.
Fagerberg L, Hallstrom BM, Oksvold P, Kampf C, Djureinovic D, Odeberg J, et al. Analysis of the human tissue-specific expression by genome-wide integration of transcriptomics and antibody-based proteomics. Mol Cell Proteomics. 2014;13:397–406.
Hileman RE, Silvanovich A, Goodman RE, Rice EA, Holleschak G, Astwood JD, et al. Bioinformatic methods for allergenicity assessment using a comprehensive allergen database. Int Arch Allergy Immunol. 2002;128:280–91.
Stadler MB, Stadler BM. Allergenicity prediction by protein sequence. FASEB J. 2003;17:1141–3.
Silvanovich A, Nemeth MA, Song P, Herman R, Tagliani L, Bannon GA. The value of short amino acid sequence matches for prediction of protein allergenicity. Toxicol Sci. 2006;90:252–8.
Goodman RE, Vieths S, Sampson HA, Hill D, Ebisawa M, Taylor SL, et al. Allergenicity assessment of genetically modified crops—what makes sense? Nat Biotechnol. 2008;26:73–81.
European Food Safety Authority. Guidance document for the risk assessment of genetically modified microorganisms and their derived products intended for food and feed use by the Scientific Panel on Genetically Modified Organisms (GMO). EFSA J. 2006;4:374.
Thomas K, Bannon G, Hefle S, Herouet C, Holsapple M, Ladics G, et al. In silico methods for evaluating human allergenicity to novel proteins: International Bioinformatics Workshop Meeting Report, 23–24 February 2005. Toxicol Sci. 2005;88:307–10.
Thomas WR, Hales BJ, Smith WA. Structural biology of allergens. Curr Allergy Asthma Rep. 2005;5:388–93.
Ladics GS, Bardina L, Cressman RF, Mattsson JL, Sampson HA. Lack of cross-reactivity between the Bacillus thuringiensis derived protein Cry1F in maize grain and dust mite Der p7 protein with human sera positive for Der p7-IgE. Regul Toxicol Pharmacol. 2006;44:136–43.
Aalberse RC, Akkerdaas J, van Ree R. Cross-reactivity of IgE antibodies to allergens. Allergy. 2001;56:478–90.
Radauer C, Willerroider M, Fuchs H, Hoffmann-Sommergruber K, Thalhamer J, Ferreira F, et al. Cross-reactive and species-specific immunoglobulin E epitopes of plant profilins: an experimental and structure-based analysis. Clin Exp Allergy. 2006;36:920–9.
Mukae T, Yoshii K, Watanobe T, Tagami T, Oishi I. Production and characterization of eggs from hens with ovomucoid gene mutation. Poult Sci. 2021;100:452–60.
Sugano S, Hirose A, Kanazashi Y, Adachi K, Hibara M, Itoh T, et al. Simultaneous induction of mutant alleles of two allergenic genes in soybean by using site-directed mutagenesis. BMC Plant Biol. 2020;20:513.
Kim B, Lee YE, Yeon JW, Go GY, Byun J, Lee K, et al. A novel therapeutic modality using CRISPR-engineered dendritic cells to treat allergies. Biomaterials. 2021;273:120798.
Kanemaru K, Noguchi E, Tahara-Hanaoka S, Mizuno S, Tateno H, Denda-Nagai K, et al. Clec10a regulates mite-induced dermatitis. Sci Immunol. 2019;4:eaax6908.
Adachi K, Hirose A, Kanazashi Y, Hibara M, Hirata T, Mikami M, et al. Site-directed mutagenesis by biolistic transformation efficiently generates inheritable mutations in a targeted locus in soybean somatic embryos and transgene-free descendants in the T(1) generation. Transgenic Res. 2021;30:77–89.
Biswal AK, Wu TY, Urano D, Pelissier R, Morel JB, Jones AM, et al. Novel mutant alleles reveal a role of the extra-large G protein in rice grain filling, panicle architecture, plant growth, and disease resistance. Front Plant Sci. 2021;12:782960.
Brackett NF, Pomes A, Chapman MD. New frontiers: precise editing of allergen genes using CRISPR. Front Allergy. 2021;2:821107.
Assou J, Zhang D, Roth KDR, Steinke S, Hust M, Reinard T, et al. Removing the major allergen Bra j I from brown mustard (Brassica juncea) by CRISPR/Cas9. Plant J. 2022;109:649–63.
Brackett NF, Davis BW, Adli M, Pomes A, Chapman MD. Evolutionary biology and gene editing of cat allergen, Fel d 1. CRISPR J. 2022;5:213–23.
Charlesworth CT, Deshpande PS, Dever DP, Camarena J, Lemgart VT, Cromer MK, et al. Identification of preexisting adaptive immunity to Cas9 proteins in humans. Nat Med. 2019;25:249–54.
Ferdosi SR, Ewaisha R, Moghadam F, Krishna S, Park JG, Ebrahimkhani MR, et al. Multifunctional CRISPR-Cas9 with engineered immunosilenced human T cell epitopes. Nat Commun. 2019;10:1842.
Simhadri VL, McGill J, McMahon S, Wang J, Jiang H, Sauna ZE. Prevalence of pre-existing antibodies to CRISPR-associated nuclease Cas9 in the USA population. Mol Ther Methods Clin Dev. 2018;10:105–12.
Wagner DL, Amini L, Wendering DJ, Burkhardt L-M, Akyüz L, Reinke P, et al. High prevalence of Streptococcus pyogenes Cas9-reactive T cells within the adult human population. Nat Med. 2019;25:242–8.
Pickar-Oliver A, Gersbach CA. The next generation of CRISPR-Cas technologies and applications. Nat Rev Mol Cell Biol. 2019;20:490–507.
El-Mounadi K, Morales-Floriano ML, Garcia-Ruiz H. Principles, applications, and biosafety of plant genome editing using CRISPR-Cas9. Front Plant Sci. 2020;11:56.
Nordstrom EA, Teixeira C, Montelius C, Jeppsson B, Larsson N. Lactiplantibacillus plantarum 299v (LP299V((R))): three decades of research. Benef Microbes. 2021;12:441–65.
Gonczi NN, Strang O, Bagi Z, Rakhely G, Kovacs KL. Interactions between probiotic and oral pathogenic strains. Biol Fut. 2021;72:461–71.
Gross LA, Baird GS, Hoffman RC, Baldridge KK, Tsien RY. The structure of the chromophore within DsRed, a red fluorescent protein from coral. Proc Natl Acad Sci USA. 2000;97:11990–5.
Tsien RY. The green fluorescent protein. Annu Rev Biochem. 1998;67:509–44.
Wall MA, Socolich M, Ranganathan R. The structural basis for red fluorescence in the tetrameric GFP homolog DsRed. Nat Struct Biol. 2000;7:1133–8.
Labas YA, Gurskaya NG, Yanushevich YG, Fradkov AF, Lukyanov KA, Lukyanov SA, et al. Diversity and evolution of the green fluorescent protein family. Proc Natl Acad Sci USA. 2002;99:4256–61.
Richards HA, Han CT, Hopkins RG, Failla ML, Ward WW, Stewart CN Jr. Safety assessment of recombinant green fluorescent protein orally administered to weaned rats. J Nutr. 2003;133:1909–12.
Strack RL, Strongin DE, Bhattacharyya D, Tao W, Berman A, Broxmeyer HE, et al. A noncytotoxic DsRed variant for whole-cell labeling. Nat Methods. 2008;5:955–7.
Tao W, Evans BG, Yao J, Cooper S, Cornetta K, Ballas CB, et al. Enhanced green fluorescent protein is a nearly ideal long-term expression tracer for hematopoietic stem cells, whereas DsRed-express fluorescent protein is not. Stem Cells. 2007;25:670–8.
Matsushima H, Ogawa Y, Miyazaki T, Tanaka H, Nishibu A, Takashima A. Intravital imaging of IL-1beta production in skin. J Invest Dermatol. 2010;130:1571–80.
Nordin O, Donald W, Ming WH, Ney TG, Mohamed KA, Halim NA, et al. Oral ingestion of transgenic RIDL Ae. aegypti larvae has no negative effect on two predator Toxorhynchites species. PLoS ONE. 2013;8:e58805.
United States of America Environmental Protection Agency. Human health and environmental risk assessment for the new product OX5034 containing the tetracycline-repressible transactivator protein variant (tTAV-OX5034; new active ingredient) protein, a DsRed2 protein variant (DsRed2-OX5034; new inert ingredient), and the genetic material (Vector pOX5034) necessary for their production in OX5034 Aedes aegypti. Washington DC; 2020.
Lacroix R, McKemey AR, Raduan N, Kwee Wee L, Hong Ming W, Guat Ney T, et al. Open field release of genetically engineered sterile male Aedes aegypti in Malaysia. PLoS ONE. 2012;7:e42771.
Carvalho DO, McKemey AR, Garziera L, Lacroix R, Donnelly CA, Alphey L, et al. Suppression of a field population of Aedes aegypti in Brazil by sustained release of transgenic male mosquitoes. PLoS Negl Trop Dis. 2015;9:e0003864.
United States of America Environmental Protection Agency. Human health and environmental risk assessment of OX5034 Aedes aegypti containing tetracycline-repressible transactivator protein variant (tTAV-OX5034, new active ingredient), DsRed2-OX5034 protein (new inert ingredient), and the genetic material necessary (vector pOX5034) for production of the proteins in vivo. Data and information provided in support of an extension and amendment to a FIFRA. Washington, DC; 2022.
Meghani Z. Regulation of genetically engineered (GE) mosquitoes as a public health tool: a public health ethics analysis. Global Health. 2022;18:21.
Netherlands National Institute for Public Health and the Environment. Technical evaluation of a potential release of OX513A Aedes aegypti mosquitoes on the island of Saba. Bilthoven: RIVM; 2017.
Matsuoka H, Sano G, Hattori R, Tomita H, Yamamoto DS, Hirai M. One injection of DsRed followed by bites from transgenic mosquitoes producing DsRed in the saliva elicits a high titer of antibody in mice. Trop Med Health. 2012;40:47–52.
Long JZ, Lackan CS, Hadjantonakis AK. Genetic and spectrally distinct in vivo imaging: embryonic stem cells and mice with widespread expression of a monomeric red fluorescent protein. BMC Biotechnol. 2005;5:20.
Barbier M, Damron FH. Rainbow vectors for broad-range bacterial fluorescence labeling. PLoS ONE. 2016;11:e0146827.
van den Broek MF, van den Berg WB, van de Putte LB, Severijnen AJ. Streptococcal cell wall-induced arthritis and flare-up reaction in mice induced by homologous or heterologous cell walls. Am J Pathol. 1988;133:139–49.
van Bruggen MC, van den Broek MF, van den Berg WB. Streptococcal cell wall-induced arthritis and adjuvant arthritis in F344 → lewis and in Lewis → F344 bone marrow chimeras. Cell Immunol. 1991;136:278–90.
Schwab JH, Brown RR, Anderle SK, Schlievert PM. Superantigen can reactivate bacterial cell wall-induced arthritis. J Immunol. 1993;150:4151–9.
Esser RE, Stimpson SA, Cromartie WJ, Schwab JH. Reactivation of streptococcal cell wall-induced arthritis by homologous and heterologous cell wall polymers. Arthritis Rheum. 1985;28:1402–11.
Shaler CR, Choi J, Rudak PT, Memarnejadian A, Szabo PA, Tun-Abraham ME, et al. MAIT cells launch a rapid, robust and distinct hyperinflammatory response to bacterial superantigens and quickly acquire an anergic phenotype that impedes their cognate antimicrobial function: defining a novel mechanism of superantigen-induced immunopathology and immunosuppression. PLoS Biol. 2017;15:e2001930.
Fiedel BA, Jackson RW. Immunogenicity of a purified and carrier-complexed streptococcal lipoteichoic acid. Infect Immun. 1976;13:1585–90.
Kim S, Koo T, Jee HG, Cho HY, Lee G, Lim DG, et al. CRISPR RNAs trigger innate immune responses in human cells. Genome Res. 2018;28:367–73.
Arca B, Colantoni A, Fiorillo C, Severini F, Benes V, Di Luca M, et al. MicroRNAs from saliva of anopheline mosquitoes mimic human endogenous miRNAs and may contribute to vector-host-pathogen interactions. Sci Rep. 2019;9:2955.
Polack FP, Thomas SJ, Kitchin N, Absalon J, Gurtman A, Lockhart S, et al. Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine. N Engl J Med. 2020;383:2603–15.
Baden LR, El Sahly HM, Essink B, Kotloff K, Frey S, Novak R, et al. Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine. N Engl J Med. 2021;384:403–16.
Sadoff J, Gray G, Vandebosch A, Cardenas V, Shukarev G, Grinsztejn B, et al. Safety and efficacy of single-dose Ad26.COV2.S vaccine against Covid-19. N Engl J Med. 2021;384:2187–201.
Wolff JA, Malone RW, Williams P, Chong W, Acsadi G, Jani A, et al. Direct gene transfer into mouse muscle in vivo. Science. 1990;247:1465–8.
Wang F, Zuroske T, Watts JK. RNA therapeutics on the rise. Nat Rev Drug Discov. 2020;19:441–2.
Sahin U, Kariko K, Tureci O. mRNA-based therapeutics–developing a new class of drugs. Nat Rev Drug Discov. 2014;13:759–80.
Pandey M, Ojha D, Bansal S, Rode AB, Chawla G. From bench side to clinic: potential and challenges of RNA vaccines and therapeutics in infectious diseases. Mol Aspects Med. 2021;81:101003.
Nishida F, Zanuzzi CN, Sisti MS, Falomir Lockhart E, Camiña AE, Hereñú CB, et al. Intracisternal IGF-1 gene therapy abrogates kainic acid-induced excitotoxic damage of the rat spinal cord. Eur J Neurosci. 2020;52:3339–52.
Kaustio M, Nayebzadeh N, Hinttala R, Tapiainen T, Åström P, Mamia K, et al. Loss of DIAPH1 causes SCBMS, combined immunodeficiency, and mitochondrial dysfunction. J Allergy Clin Immunol. 2021;148:599–611.
Nakajima K, Kazuno AA, Kelsoe J, Nakanishi M, Takumi T, Kato T. Exome sequencing in the knockin mice generated using the CRISPR/Cas system. Sci Rep. 2016;6:34703.
Wagner S, Breiteneder H, Simon-Nobbe B, Susani M, Krebitz M, Niggemann B, et al. Hev b 9, an enolase and a new cross-reactive allergen from hevea latex and molds. Purification, characterization, cloning and expression. Eur J Biochem. 2000;267:7006–14.
Sanz ML, Garcia-Aviles MC, Tabar AI, Anda M, Garcia BE, Barber D, et al. Basophil Activation Test and specific IgE measurements using a panel of recombinant natural rubber latex allergens to determine the latex allergen sensitization profile in children. Pediatr Allergy Immunol. 2006;17:148–56.
Peng Z, Simons FE. Comparison of proteins, IgE, and IgG binding antigens, and skin reactivity in commercial and laboratory-made mosquito extracts. Ann Allergy Asthma Immunol. 1996;77:371–6.
Peng Z, Li H, Simons FE. Immunoblot analysis of salivary allergens in 10 mosquito species with worldwide distribution and the human IgE responses to these allergens. J Allergy Clin Immunol. 1998;101:498–505.
Arca B, Lombardo F, Struchiner CJ, Ribeiro JM. Anopheline salivary protein genes and gene families: an evolutionary overview after the whole genome sequence of sixteen Anopheles species. BMC Genomics. 2017;18:153.
Hammond A, Galizi R, Kyrou K, Simoni A, Siniscalchi C, Katsanos D, et al. A CRISPR-Cas9 gene drive system targeting female reproduction in the malaria mosquito vector Anopheles gambiae. Nat Biotechnol. 2016;34:78–83.
Roberts AL, Connolly KL, Kirse DJ, Evans AK, Poehling KA, Peters TR, et al. Detection of group A Streptococcus in tonsils from pediatric patients reveals high rate of asymptomatic streptococcal carriage. BMC Pediatr. 2012;12:3.
Shaikh N, Leonard E, Martin JM. Prevalence of streptococcal pharyngitis and streptococcal carriage in children: a meta-analysis. Pediatrics. 2010;126:e557–64.
Levy RM, Leyden JJ, Margolis DJ. Colonisation rates of Streptococcus pyogenes and Staphylococcus aureus in the oropharynx of a young adult population. Clin Microbiol Infect. 2005;11:153–5.
Carapetis JR, Steer AC, Mulholland EK, Weber M. The global burden of group A streptococcal diseases. Lancet Infect Dis. 2005;5:685–94.
Hammond B, Kough J, Herouet-Guicheney C, Jez JM, ILSI International Food Biotechnology Committee Task Force on use of mammalian toxicology studies in safety assessment of GM Foods. Toxicological evaluation of proteins introduced into food crops. Crit Rev Toxicol. 2013;43(Suppl 2):25–42.
Rusk N. Human immunity to Cas9. Nat Methods. 2019;16:286.
Mehta A, Merkel OM. Immunogenicity of Cas9 Protein. J Pharm Sci. 2020;109:62–7.
Romeis J, Bartsch D, Bigler F, Candolfi MP, Gielkens MM, Hartley SE, et al. Assessment of risk of insect-resistant transgenic crops to nontarget arthropods. Nat Biotechnol. 2008;26:203–8.
Connolly JB, Romeis J, Devos Y, Glandorf DCM, Turner G, Coulibaly MB. Gene drive in species complexes: defining target organisms. Trends Biotechnol. 2023;41:154–64.
Stephens D, Brown J, Ydenberg R. Foraging: behavior and ecology. Chicago, IL: Univeristy of Chicago Press; 2007.
We thank Camilla Beech, Silke Fuchs, and John Mumford for their comments on earlier versions of this manuscript and are indebted to Karen Logan and Austin Burt for their support of this work. We are also grateful for anonymous comments received during peer review, which greatly strengthened the conclusions of this manuscript.
Supported by a grant from the Bill & Melinda Gates Foundation and Open Philanthropy.
Ethics approval and consent to participate
Consent for publication
JBC is employed by Imperial College London in a position which is solely funded through a grant from the Bill & Melinda Gates Foundation, member of Target Malaria not-for-profit research consortium, and awardee of travel grants from FNIH. AQ is employed by Imperial College London in a position which is solely funded through a grant from Open Philanthropy and a member of Target Malaria not-for-profit research consortium.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Qureshi, A., Connolly, J.B. Bioinformatic and literature assessment of toxicity and allergenicity of a CRISPR-Cas9 engineered gene drive to control Anopheles gambiae the mosquito vector of human malaria. Malar J 22, 234 (2023). https://doi.org/10.1186/s12936-023-04665-5
- Population suppression
- Gene drive
- Environmental risk assessment (ERA)
- Genetically modified mosquitoes (GMM)