B6.SJL-Ptprca Pepcb/Boy (CD45.1) mice have been used in hundreds of congenic competitive transplants, with the presumption that they differ from C57BL/6 mice only at the CD45 locus. In this study, we describe a point mutation in the natural cytotoxicity receptor 1 (Ncr1) locus fortuitously identified in the CD45.1 strain. This point mutation was mapped at the 40th nucleotide of the Ncr1 locus causing a single amino acid mutation from cysteine to arginine at position 14 from the start codon, resulting in loss of NCR1 expression. We found that these mice were more resistant to CMV due to a hyper innate IFN-γ response in the absence of NCR1. In contrast, loss of NCR1 increased susceptibility to influenza virus, a result that is consistent with the role of NCR1 in the recognition of influenza Ag, hemagglutinin. This work sheds light on potential confounding experimental interpretation when this congenic strain is used as a tool for tracking lymphocyte development.
The most common approach used for tracking immune cell development in vivo takes advantage of polymorphisms in the extracellular domain of the transmembrane receptor tyrosine phosphatase protein CD45 (Ptprc), a 220-kDa protein expressed on all subsets of leukocytes. Two isoforms have been identified in mice: the common form is CD45.2, which is expressed by the C57BL/6 (B6) strain and is encoded by the Ptprcb allele; and an additional allelic variant Ptprca, which encodes the CD45.1 isoform, was identified in the SJL mouse strain. CD45.1 and CD45.2 alleles differ by only 5 aa within the extracellular domain, resulting in epitope changes that permit specific recognition by mAbs. Backcrossing of mice expressing the CD45.1 allele into the B6 background has resulted in the development of the mouse strain B6.SJL-Ptprca Pepcb/Boy (CD45.1). As CD45.1 mice have been backcrossed during many generations into the B6 background, these mice have been termed congenic, with the presumption that they differ from the B6 strain only at the CD45 locus. Surprisingly, despite extensive backcrossing, genotypic analysis revealed that the congenic interval in which CD45.1 differs from CD45.2 mice is almost 43 mbp encoding 306 genes and at least 124 genetic polymorphisms (1).
In this study, we describe a point mutation in the natural cytotoxicity receptor 1 (Ncr1) locus of CD45.1 strain, resulting in loss of expression. A point of critical consequence of this mutation is the different susceptibility of this strain to viral infection. Thus, using this “congenic” strain, under the presumption that it differs from the B6 strain only at the CD45 locus, probably has been, and most likely will be, conducive to confounding experimental interpretation.
Materials and Methods
CD45.2 and CD45.1 mice were originally obtained from The Jackson Laboratory and maintained in our animal facility at the University of Michigan. For bone marrow mixed chimeras, CD45.2 recipient mice were lethally irradiated (1200 rad) and injected (i.v) with 10 million donor bone marrow cells containing a mixture of CD45.2 and CD45.1 cells provided at a 1:1 ratio. For the transfer of gut microbiota, gut content was harvested and provided to recipient mice by oral gavage. As indicated, mice were either infected with murine CMV (MCMV; 3500 PFU of Smith strain per gram of body weight by i.p. injection), influenza A virus (50 PFU of H1N1 per mouse by intranasal inhalation), or Citrobacter rodentium (1010 CFU per mouse by oral gavage). All experiments were performed in accordance with the University of Michigan Animal Care and Use Committee, and approval to use mice was granted by the University of Michigan in accordance with the National Institutes of Health requirements for the care and use of animals.
Sequencing, cloning, and transfection of the Ncr1 gene
The Ncr1 open reading frame region was amplified from cDNA using the following primers, and amplified products were sequenced by Sanger method (University of Michigan): 5′-GTTCAGCACTGGTCTGGCCACTGG-3′ and 5′-GCCAAACTTGGTAACACTCCTACC-3′. For cloning and transfection of the Ncr1 gene, the Ncr1 open reading frame was amplified from cDNA using the following primers: 5′-GGAATTAGAGAGTTTCATGCTGCCAACACTCACTGC-3′ and 5′- GAATTGTGGAAGTTTCTCACAAGGCCCCAGGAGTTG-3′. Amplified Ncr1 gene was cloned into pMSCV-neo vector, transfected into Bosc cells for viral packaging, infected into MEF cells, and then selected with 500 μg/ml Geneticin for analysis.
Results and Discussion
CD45.1 and CD45.2 strains are not functionally equivalent
CD45.1 and CD45.2 mice were originally obtained from The Jackson Laboratory and then maintained by breeding in our animal facility. While performing experiments with CD45.2 and CD45.1 mice, we observed a pattern of responses that suggested inherent differences in their susceptibility to infection (Fig. 1). In response to MCMV infection, we found that CD45.1 mice were better equipped to fight a lethal dose of MCMV. In response to influenza virus infection, we observed opposite results, with CD45.1 mice being less protected than CD45.2 mice. However, in response to C. rodentium infection, both CD45.1 and CD45.2 mouse strains were equally resistant to the enteric bacteria. Based on these results, we conclude that CD45.1 and CD45.2 mouse strains—that are housed in our animal facility—are not functionally equivalent.
Point mutation in the Ncr1 locus identified in the CD45.1 strain
Detailed analysis of the NK cell compartment landed one major difference between CD45.1 and CD45.2 strains: NCR1 expression was lost in the CD45.1 strain (Fig. 2). Loss of NCR1 expression in the CD45.1 strain was identified in NK cells from various organs (Fig. 2A, 2B) and innate lymphoid cells from the intestine (Fig. 2C). We confirmed that the loss of NCR1 expression was not the result of intracellular sequestration, as shown by surface and intracellular staining and with monoclonal and polyclonal anti-NCR1 Abs (Fig. 2D, 2I). Given the increasing evidence of associations between host gene expression and gut microbiota (2), we considered the possibility that the loss of NCR1 in the CD45.1 strain is the result of a specific gut microbiota acquired during breeding in our animal facility. However, the transfer of gut content from CD45.2 mice failed to restore the expression of NCR1 in CD45.1 recipient NK cells (Fig. 2E). Finally, to demonstrate that the loss of NCR1 in CD45.1 strain is not driven by the environment, we assessed the expression of NCR1 in mixed bone marrow chimeras. Results showed no expression of NCR1 on CD45.1-derived NK cells, indicating a cell-autonomous defect (Fig. 2F).
Because levels of the Ncr1 transcripts were comparable in both mouse strains (Fig. 2G), we sequenced the mRNA and identified a point mutation, TGT to CGT, in the 40th nucleotide from the start codon leading to a single amino acid mutation from cysteine to arginine at the 14th aa (C14R), which is localized in the region of the signal peptide of NCR1 (Fig. 2H). To evaluate the effects of this signal peptide mutation, we constructed expression vectors with mutant (NCR1C14R) or wild-type (NCR1WT) forms and assessed outcomes on protein expression. Unlike wild-type controls, no protein was detected from the mutated construct, indicating that this signal peptide mutation was sufficient to abrogate NCR1 expression (Fig. 2J, 2K). Signal peptides are responsible for numerous functions, including recognition and binding of the protein to the cytosolic signal recognition particle, its insertion into the endoplasmic reticulum (ER) protein–conducting channels, and protein maturation processes such as quality control and addition of N-linked glycans (3). Examples from signal peptide mutations showed that changes in the signal peptide hydrophobicity interfere with the binding of the mutant preproteins to the signal recognition and translocation into the ER (4). Even for the fraction of the mutant preproteins that enter the ER, their cleavage by the signal peptidase is inefficient; as a result, the uncleaved preproteins could not reach the glycosylation machinery and therefore were retained in the ER (4). Alternatively, the mutant preproteins are removed from the ER by a process of active degradation via the proteasome pathway (5). Although further studies are required to identify the exact mechanism, our results imply a similar mechanism for the C14R signal peptide mutation resulting in degradation and loss of NCR1. Accordingly, we termed this mutant strain Ncr1C14R mice.
Mutant Ncr1C14R confers hyper IFN-γ response without altering the lytic activity
NCR1 contains two extracellular Ig domains, a type I transmembrane domain, and a short cytoplasmic tail lacking ITAMs. Given the lack of inherent signaling motifs, NCR1 relies on adaptor molecules FCεRIγ and CD3ζ to impart positive signals through their ITAMs (6). Based on this structure, NCR1 has been assumed to act solely as an activating receptor delivering positive signals to NK cells (7–9). However, such a role has been challenged by opposite results arguing instead for a regulatory role of NCR1 in NK cell activities (10). Given the controversial data reported from Ncr1gfp/gfp and Ncr1Noe/Noe mice (11, 12), we sought to re-examine the role of NCR1 using the Ncr1C14R mutant mice.
Although the efficacy of lytic activity was not altered in the absence of NCR1 (Fig. 3A–C), the IFN-γ response was higher in Ncr1C14R compared with wild-type NK cells (Fig. 3D–F). To examine how this gain of function translates in vivo, mutant and wild-type mice were infected with a lethal dose of MCMV (Fig. 3G–M). We found that Ncr1C14R mice were more resistant to MCMV infection than were wild-type mice, as indicated by higher survival rates (Fig. 1B) and reduced viral loads in the liver and spleen (Fig. 3G). Analysis of NK cells on 1.5 and 5 d postinfection showed higher production of IFN-γ both in the liver and spleen of Ncr1C14R compared with wild-type mice, and both in Ly49H+ and Ly49H− NK cell subsets lacking NCR1 expression (Fig. 3H–M). Notably, neither proliferation nor production of granzyme B (GzmB) was altered in the absence of NCR1 (Supplemental Fig. 1A), indicating a selective role of NCR1 in the regulation of innate IFN-γ. One could raise the concern that, in these experiments, the control mice are not the mutant’s littermates. To address this issue, we repeated the above experiments using mutant (F2 Ncr1C14R) and wild-type (F2 Ncr1WT) littermates derived from F2 progeny (Fig. 3N–R, Supplemental Fig. 1B–E) and confirmed the gain of NK cell IFN-γ response in the absence of NCR1. Interestingly, expression of NCR1 in F1 heterozygotes was intermediate, indicating that normal expression requires both functional alleles. Consistent with our data, stimulation of human NK cells in the presence of anti-NCR1 blocking Ab resulted in copious amounts of IFN-γ (13). Of note, this gain of function was specific to NCR1, as blocking NCR2 or NCR3 showed no effects on IFN-γ (13).
Because levels of T-bet, Eomes, and Helios were not different in both mouse groups, (Supplemental Fig. 2), we tested the possibility that augmented IFN-γ response in the absence of NCR1 is mediated via downstream of kinase (Dok)-1 (Fig. 3S). By recruiting enzymes transducing negative signals such as RasGAP and SHIP1, Dok-1 inhibits Ras/ERK and PI3K/Akt signaling pathways, resulting in negative regulation of leukocyte activation (14). In T cells, Dok-1 is tyrosine phosphorylated, resulting in negative regulation of TCR-induced T cell activation. Overexpression of Dok-1 decreased IL-2 production (15), and loss of Dok-1 enhanced TCR-mediated signaling (16). Likewise, overexpression of Dok-1 in NK cells reduced IFN-γ production whereas Dok-1 gene ablation augmented the IFN-γ response (17). Consistent with these results, we found that the loss of NCR1 in NK cells was sufficient to prevent Dok-1 activation downstream of Ly49H engagement, resulting in increased MAPK activation and augmented IFN-γ response in NK cells (Fig. 3S). Although it is tempting to speculate that NCR1 may serve as a regulator of Dok-1 activation downstream of NK cell receptor engagement, further studies are required to elucidate the specific nature of the interplay between NCR1 and Dok-1 in control of NK cell IFN-γ.
Given the frequent use of CD45.1 strain as a tool for tracking lymphocyte development, our finding raised one important question: did the Ncr1C14R mutation identified in our congenic mouse colony originate from the vendor or did it occur in our animal facility as a result of genetic drifting? Using flow cytometry data from our archives, we were able to trace this mutation backward and have identified the loss of NCR1 in CD45.1 mice from experiments dated from 2009 (Fig. 4). Conversely, analysis of recent studies performed with CD45.1 mice purchased after 2014 showed normal expression of NCR1 (Fig. 4). Colony maintenance information from The Jackson Laboratory indicates that the CD45.1 strain (stock no. 002014) had been imported into The Jackson Laboratory in 1990 from Dr. Edward Boyse at generation N22. From that time until 2009, the strain was maintained by pedigrees and filial matings. Between 2009 and 2010, the B6 CD45.1 congenic mouse colony underwent three backcrosses to the inbred mouse strain C57BL/6J (N25). It is possible that this backcrossing may have removed the Ncr1C14R mutation identified in mice obtained prior to 2010; however, this has not been directly tested. Notably, as the Ncr1C14R point mutation results in loss of NCR1 expression, determining the status of your congenic mouse colony will be easily achievable by flow cytometry. Otherwise, using these mice, as a tool for tracking lymphocyte development under the presumption that they differ from the B6 strain only at the CD45 locus, will certainly be conducive to confounding experimental interpretation.
We thank Dr. Low-Marchelli from The Jackson Laboratory for providing helpful information on the colony generation of the CD45.1 congenic strain.
This work was supported by National Institutes of Health Grants R01 AI102893 and R01 CA179363 (to S.M.), R01 HL119682 (to B.B.M.), and R01 AI083642 (to Y.L.), the Midwest Athletes Against Childhood Cancer Fund (to S.M.), the Gardetto Family Foundation (to S.M.), the Nicholas Family Foundation (to S.M.), and American Association of Immunologists Careers in Immunology Grant N023607 (to Y.L.).
The online version of this article contains supplemental material.
The authors have no financial conflicts of interest.