Background

Why map and identify genes for rat disease phenotypes or related traits? As already pointed out, the laboratory rat (Rattus norvegicus) is more than a big mouse. The mouse is a species which has been the mammalian genetic model of choice for a long time, with an initial focus on monogenic traits. Rat models of monogenic traits and diseases have also been isolated but the rat has essentially been a key model for studies of complex traits in fields such as physiology, cardiovascular and diabetes research, arthritis, pharmacology, toxicology, oncology and neurosciences [1,2,3,4,5,6]. In some situations the rat seems to be a more relevant or faithful model. For instance, the physiology of the rat is extremely well documented, in part because its larger body size affords the opportunity for serial blood draws, which are almost impossible in the mouse; in cardiovascular research [7], sophisticated surgical manipulations, and physiological measurements such as blood pressure measurements by telemetry are easier to perform and more reliable in rats compared to mice [1, 3]. The rat has also long been a common choice for pharmacology and toxicology studies because it shares a similar pathway with humans for eradicating toxins [8]. With respect to cancer research [9, 10], and more precisely mammary cancer research, it is noteworthy that rat and human carcinomas show similar development and histopathological features [11, 12]; furthermore, rat mammary tumors are strongly hormone dependent for both induction and growth, thus resembling human breast tumors and no virus appears to be involved in rat and human mammary carcinogenesis, unlike mouse mammary carcinogenesis the etiological agent of which is the mouse mammary tumor virus. As stated by Russo “The rat mammary tumor model is well suited for studying in situ and invasive lesions […]. The classification of the tumors matches well with the criteria used in the human pathology, and provides an adequate model for understanding these phases of the human disease” [11]. In addition, there is extensive overlap between human breast and rat mammary cancer susceptibility genomic regions and “the laboratory rat will continue to be an important model organism for researching genetically determined mechanisms of mammary cancer susceptibility that may translate directly to human susceptibility” [13]. In neuroscience research, rats have significant anatomical and behavioral advantages over mice, because they are more sociable and skilled and have complex cognitive abilities; this wider range of social behaviors and a richer acoustic communication system confer the rat advantages in comparison to mouse models to study neuro-developmental disorders and in particular autism [14, 15]. The rat thus provides one with particularly reliable models of human traits or diseases [3, 8, 11,12,13,14,15,16,17] (multiple details emphasizing the value of rat models can be found in these articles).

Numerous rat strains have been created by selective breeding of animals expressing a desired phenotype, generating a very large collection of genetic models of pathological complex, polygenic traits, most of which are quantitative. Interestingly, these strains also provide one with additional phenotypes, which were not selected for. Just as the traits that were selected for, most of these phenotypes are polygenic. All these phenotypes can be used as models of human traits or diseases [18], implying that the genes underlying these traits or diseases should be identified. Information on rat strains and rat disease models, can be found at the Rat Genome Database (RGD, https://rgd.mcw.edu/) [19].

In order to give the rat the status of a valuable genetic model, and in particular to identify the genes underlying complex traits by forward genetic approaches and to analyze the relevant biological mechanisms, several tools had to be developed. This has been accomplished. Genetic and chromosome maps have been developed; the genomic sequence of dozens of rat strains has been established; a number of resources have been created to provide investigators with access to genetic, genomic, phenotype and disease-relevant data as well as software tools necessary for their research [3, 20]. Thanks to these resources, positional identification of numerous rat genes underlying monogenic or complex diseases and related traits could be achieved. On the other hand, reverse genetic tools have also been developed. Efficient methods to generate mutant rats became available; sperm N-ethyl-N-nitrosourea (ENU) mutagenesis followed by gene-targeted screening methods lead to the isolation of several mutants, including knockout (KO) strains ([21] and references therein). Rat ES were successfully derived and could be used for targeted mutations by homologous recombination; more importantly, several methods not relying on the use of ES cells were introduced to generated targeted mutations (often these are KO mutations), namely gene editing by zinc finger nucleases, by transcription activator-like effector nucleases and finally by the clustered regularly interspaced short palindromic repeat (CRISPR/Cas) system [22]. Transgenic rats can also be generated, including humanized rats carrying large chromosomic fragments (“transchromosomic humanized” rats) [23]. Development of these technologies provides the researcher with all the tools required to take advantage of the unique opportunities offered by the rat as leading model for studies different areas of biomedical research [3, 17]. In this review I made an inventory of the rat genes identified as responsible for monogenic or polygenic diseases and related traits. I took into account the rat genes identified by forward genetic methods as well as those inactivated by ENU-mutagenesis and by targeted mutations, the inactivation of which generated a disease or an abnormal phenotype. This update of the progress made in the identification of rat disease genes shows that a considerable number of conserved genes have similar effects on biological traits in rats and humans, establishing the rat as a valuable model in studies of the genetic basis of human diseases and thus providing one with a useful resource of disease models.

Materials and methods

The data (causal genes of rat diseases and related traits) were collected by regular and systematic screening of the biomedical literature, PubMed searches (https://www.ncbi.nlm.nih.gov/) and regular Google Scholar alerts based on the keywords “knockout”, “mutation”, “rat” (spread over several months). In addition, relevant data were retrieved from the RGD (“Disease Portals”), with advices from Jennifer R. Smith. Genes identified by forward genetic means (or by direct molecular sequencing) were considered as suggestive, solid or confirmed, respectively, as indicated in each case in Table 1 (one, two or three asterisks), on the basis of the criterions described in the legend to the table; these criterions are based on the standards discussed by Glazier and coworkers [24]. With respect to the induced mutants, they were included provided they were convincingly shown to be specifically altered. The official gene symbols are used in this article and were obtained from the National Center for Biotechnology Information (https://www.ncbi.nlm.nih.gov/), Gene section. In several instances the original publications did not use the official gene symbol; in these cases, the non-official symbol is indicated in parenthesis in the footnote to the table, where the full name of each gene is described. The position of every gene was also obtained from the NCBI.

Results

The core of this article is a list of the diseases and related traits or phenotypes the causal gene of which was identified in the rat (Table 1). The genes identified by forward genetic methods or, in a few instances, by direct molecular characterization are labeled by asterisks (see legend to table). Also listed are the phenotypes uncovered by reverse genetics methods, either by ENU-mutagenesis followed by selection of the desired mutated gene (these genes are labeled by the symbol ENU), or by targeted gene editing (these genes are labeled by T). Table 1A shows the monogenic traits, and Table 1B the complex traits (in a few cases this distinction is somewhat arbitrary, but in general this is a useful classification). Of note, when a gene was associated with several distinct phenotypes, an entry was created for each phenotype and the gene thus appears several times in the table. When the human homolog gene is known to be causal of the relevant disease or trait, it is also indicated in the table. Furthermore, entries in bold characters indicate that the human gene was found to be causal as a direct translation of the results obtained in the rat.

Identification of rat disease genes by forward genetic methods

The identification of gene(s) underlying a given phenotype typically starts with the mapping of the trait by linkage analysis (backcrosses, intercrosses). In the case of monogenic traits, this approach is generally sufficient to identify the causative gene (positional identification, as illustrated in Table 1A). Identifying genes controlling complex traits is much more difficult [24, 25]; indeed, linkage analyses of such traits lead to the localization of quantitative trait loci (QTLs), which are too large to allow the identification of the causative gene. Complementary strategies are thus required to narrow down the list of candidate genes, such as the generation of congenic lines or/and the use of integrative genomic approaches (as discussed in [26]). Alternative approaches rely on the use of panels of lines that show a higher level of recombinant events, as a result of crossing parental strains for multiple generations, such as recombinant inbred strains or heterogeneous stocks (as discussed in [27], for a striking harvest of results derived from the study of a heterogeneous stock, see [28]). The first complex-trait gene identified is the Cd36 gene, which causes insulin resistance, hyperlipidemia and hypertension in the spontaneously hypertensive rat (SHR) [29, 30]. This identification was based on a combined gene expression micro-array and linkage approach and was definitively proven by in vivo complementation, i.e. transgenic expression of normal Cd36 in the SHR [31]. Last but not least, association was then demonstrated between human CD36 and insulin resistance [32]. Subsequently, the tools of forward genetic studies as well as gene expression and/or computational analysis (integrative genomics) led to the identification of numerous genes underlying rat polygenic traits or diseases, such as blood pressure, cardiac mass, diabetes, inflammation (in particular arthritis, encephalomyelitis), glomerulonephritis, mammary cancer, neurobehavioral traits, proteinuria. In several instances, the results were translated to the human, as illustrated in Table 1 by bold entries. Interestingly, a recently discovered complex trait gene is a long non-coding RNA, itself contained within the 5′ UTR of the Rffl gene (Rffl-lnc1); Rffl-lnc1 shows a 19 bp indel polymorphism which is the precise variation underlying regulation of blood pressure and QT-interval. This work was based on fine and systematic congenic mapping and is the first one to identify quantitative trait nucleotides in a long non-coding RNA [33]. The human homologous region, on chromosome 17, has multiple minor alleles that are associated with shorter QT-intervals and, is some cases, hypertension [34].

Identifying rat disease genes is not only useful to discover the homologous human disease genes but also helps in studying the mechanisms underlying the pathological abnormalities. After all, this is the essence of an animal model. For instance, the study of the genetic basis of stroke in the stroke-prone SHR strain (SHRSP) led to the conclusion that mitochondrial dysfunction contributes to stroke susceptibility and to hypertensive target organ damage (such as vascular damage); this better understanding of the etiology of the disease can open the door to novel therapies, as briefly discussed below [8, 35, 36].

The importance of rat models in the era of human genetic studies and genome sequencing

The rat is also a useful model to decipher the biological significance of QTLs identified in human genome-wide association studies (GWAS) aimed at understanding the etiology of common human diseases [37, 38]. These studies pint-point human genomic regions controlling a complex trait, and generally contain several genes; the current methods lack the statistical power to pinpoint the human causative gene. Animal model such as the rat provides one with the possibility to knockout or to mutate in more subtle manner each of the rat genes homolog to the human genes contained in a given GWAS locus. In this way, the possible role of each gene can be evaluated. For instance, Flister and co-corkers [39], studying a multigene GWAS locus controlling blood pressure and renal phenotypes (AGTRAP-PLOD1 locus) used gene targeting in a rat model to test each of the genes contained in this locus. In this way these authors could show that several genes impact hypertension and that multiple causative gene variants cosegregate at this locus; several linked genes thus control blood pressure (Agtrap, Clcn6, Mthfr, Nppa, Plod1). Furthermore, each of the KO rat models so generated can be used to dissect the biological effects of the gene loss of function.

The genetic basis of human diseases is also actively analyzed by whole genome sequencing; such studies have uncovered several genes underlying diseases or related phenotypes [40, 41] and one can thus questioned the importance of genetic analyses in an animal model. As argued and illustrated above, animal models and the rat in particular, remain valuable tools to analyze the biological mechanisms underlying a phenotype. In addition, transgenesis or gene substitution can also be carried out, in which a human allele can be introduced in the relevant KO rat, in order to verify the role of the human mutation. Alternatively, the rat genome can be directly modified to specifically introduce a mutation similar to the one causing the human trait [40, 42]. If the modified rats exhibit defects similar to those observed in the human patients, it can be concluded that the tested human mutation indeed plays a causal role. In addition, similarly to examples mentioned above, such specifically modified rats provide one with models suitable to study the mechanisms responsible for the abnormalities generated by the mutation and also to carry out pharmacological tests and look for possible new therapies [42].

The need of relevant animal models is also illustrated by the fact that even when the human gene causing a disease is known, mutated rat strains (in particular KO strains) were created to analyze the gene function and the disease pathogenesis (see numerous examples of such gene targetings in Table 1). In 2008, Aitman and coworkers [2] reported a list of 21 rat disease genes (that had been identified by positional cloning). Here I updated the list of rat disease genes; this inventory added numerous genes identified (or deliberately mutated) after 2008, thereby evaluating progress made in the input of rat disease models. The total rat gene number listed in Table 1 exceeds 350, illustrating the vigor of the rat biomedical research which led to enrichment of numerous disease models, with the translation to humans of disease gene discoveries in rats.

Translation of the rat genetic studies into new treatments of human diseases

The identification of a human disease gene has the potential to develop new therapeutic approaches. For instance, the human gene NCF4 was found to be associated with arthritis as a translation of studies on a rat arthritis model. The gene encodes a component of the NADPH oxidase complex and these studies catalyzed the development of a new therapy for arthritis, based on the use of oxidative-burst inducing substances [43,44,45] (see Table 1B, Arthritis, Ncf1 gene). Another interesting example is that of the gene SHANK3: mutations in this gene lead to a neurodevelopmental disorder known as Phelan-McDermid syndrome; to date, no pharmaceutical compounds targeting core symptoms of this human disease are available. A Shank3-deficient rat model was generated, which showed disabilities similar to those seen in the Phelan-McDermid syndrome and interestingly, the deficits of the mutant rat could be ameliorated by intracerebroventricular oxytocin administration, implying that exogenous oxytocin administration might have therapeutic potential in human patients [42] (see Table 1A, Phelan-McDermid syndrome model). A third example is provided by the study of rat mutated in the Pde3a gene, which recapitulates the phenotype of HTNB (Hypertension with brachydactyly) human patients: the functional data suggest that soluble guanyly cyclase activation could be suitable for the treatment of HTNB patients [46] (See Table 1B, Blood pressure section).

Table 1 Alphabetical list of diseases and related traits with their causative rat genes and the human homologs

Conclusions

This evaluation of progress in the identification of genes causing monogenic or polygenic rat diseases or related phenotypes yielded a list containing over 350 genes. In several instances the result obtained in the rat model was translated to the human, demonstrating that a considerable number of conserved genes have similar effects on biological traits in rats and humans, and thus providing one with a rich and useful resource of disease models (Table 1, bold entries). For instance, the Inppl1 gene was first identified in the rat as a causative gene of type 2 diabetes and this discovery led to the identification of mutations in the homolog gene of diabetic patients [452, 534]. Similarly, a rat paralog of the Fcgr3 gene (Fcgr3-rs) was identified as causing glomerulonephritis, and the result was promptly translated to the human: low copy number of FCGR3B, an orthologue of rat Fcgr3, was associated with glomerulonephritis in the autoimmune disease systemic lupus erythematosus [471].

Also, as mentioned above, even when the human gene causing a disease had been identified (without resorting to a rat model), mutated rat strains, in particular KO strains, were created to analyze the gene function and the disease pathogenesis, and, potentially, to develop new therapies.

This review illustrates the vigor of the rat biomedical research and its value for understanding the etiology of human diseases and for suggesting new therapies.