Widespread discrepancy in Nnt genotypes and genetic backgrounds complicates granzyme A and other knockout mouse studies
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
This paper is of interest not only for immunologists studying the inflammation, but also for biomedical researchers studying various biological processes using C57BL/6 mice. The data in this paper indicate that genetic differences between C57BL/6 substrains can affect reproducibility and generalizability in a broad range of biological studies with mouse models reported to date.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
Abstract
Granzyme A (GZMA) is a serine protease secreted by cytotoxic lymphocytes, with Gzma -/- mouse studies having informed our understanding of GZMA’s physiological function. We show herein that Gzma -/- mice have a mixed C57BL/6J and C57BL/6N genetic background and retain the full-length nicotinamide nucleotide transhydrogenase ( Nnt ) gene, whereas Nnt is truncated in C57BL/6J mice. Chikungunya viral arthritis was substantially ameliorated in Gzma -/- mice; however, the presence of Nnt and the C57BL/6N background, rather than loss of GZMA expression, was responsible for this phenotype. A new CRISPR active site mutant C57BL/6J Gzma S211A mouse provided the first insights into GZMA’s bioactivity free of background issues, with circulating proteolytically active GZMA promoting immune-stimulating and pro-inflammatory signatures. Remarkably, k-mer mining of the Sequence Read Archive illustrated that ≈27% of Run Accessions and ≈38% of BioProjects listing C57BL/6J as the mouse strain had Nnt sequencing reads inconsistent with a C57BL/6J genetic background. Nnt and C57BL/6N background issues have clearly complicated our understanding of GZMA and may similarly have influenced studies across a broad range of fields.
Article activity feed
-
-
Evaluation Summary:
This paper is of interest not only for immunologists studying the inflammation, but also for biomedical researchers studying various biological processes using C57BL/6 mice. The data in this paper indicate that genetic differences between C57BL/6 substrains can affect reproducibility and generalizability in a broad range of biological studies with mouse models reported to date.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
-
Reviewer #1 (Public Review):
Rawle, Le et al. investigated the influence of genetic background of murine models of granzyme A absence or inactivity on the arthritic foot swelling phenotype induced by chikungunya virus. By using CRISPR/Cas knockouts, the authors show that the reduced foot swelling previously attributed to the absence of granzyme A (GzmA-/-), was in truth due to the presence of intact nicotinamide nucleotide transhydrogenase in GzmA-/-, since the experimental controls used in previous studies bear a truncated Nnt, which results in the lack of activity of the corresponding enzyme. It is worth noting that these results challenge the previously stated role of granzyme A in the arthritic foot swelling phenotype induced by chikungunya virus infection. Moreover, an interesting analysis of previous literature reveals a …
Reviewer #1 (Public Review):
Rawle, Le et al. investigated the influence of genetic background of murine models of granzyme A absence or inactivity on the arthritic foot swelling phenotype induced by chikungunya virus. By using CRISPR/Cas knockouts, the authors show that the reduced foot swelling previously attributed to the absence of granzyme A (GzmA-/-), was in truth due to the presence of intact nicotinamide nucleotide transhydrogenase in GzmA-/-, since the experimental controls used in previous studies bear a truncated Nnt, which results in the lack of activity of the corresponding enzyme. It is worth noting that these results challenge the previously stated role of granzyme A in the arthritic foot swelling phenotype induced by chikungunya virus infection. Moreover, an interesting analysis of previous literature reveals a significant number of studies comparing mice with truncated Nnt to mice with full length Nnt. Concerns over Nnt genotype in experimental settings are not new and have been raised in several studies since the description of the truncated Nnt in C57BL/6J by Toye et al. (PMID: 15729571). However, this is the first study to analyze Sequence Read Archive (SRA- NCBI) data in order to demonstrate that inappropriate comparisons regarding the Nnt genotype also occur due to listing errors and inadequate backcrossing in knock-out mice. Thus, this study emphasizes the importance of characterizing genetic backgrounds when conducting experiments on mice.
Strengths
By using CRISPR/Cas knockouts, the authors provide reliable data that strongly support the conclusions. Experimental controls and comparisons are well-conceived and suitable for the aim of the study.
Weakness
Despite the presence of a section dedicated to the re-investigation of the physiological role of granzyme A, the authors did not reach a conclusion on this subject. Experimental settings carried out to investigate the role of granzyme A were poorly explored. Moreover, the possible role of granzyme A in other chikungunya virus-induced phenotypes was not investigated, as acknowledged by the authors. Additionally, the analysis of SRA-NCBI data did not comprise the presence of heterozygous carriers of the truncated Nnt allele, which could also generate a distinct phenotype due to gene-dose effects.
-
Reviewer #2 (Public Review):
In a previous paper, the authors demonstrated that the chikungunya virus (CHIKV)-induced arthritis (foot swelling) was attenuated in Gzma gene knock-out (C57BL/6J-Gzma-/-) mice compared to C57BL/6J (B6J) mice. In this paper the authors created C57BL/6J-GzmaS211A knock-in mice that have a mutation at GzmA active site to verify the role of GzmA in CHIKV-induced arthritis. The authors observed that foot swelling following CHIKV infection was not different between B6J-GzmaS211A and B6J. The authors then conducted whole-genome sequencing of B6-Gzma-/-and found that this mouse has a mixed (mosaic) genetic background of C57BL/6J and C57BL/6N (B6N). Of note, B6J uniquely has a loss-of-function deletion mutation of exons 8-12 of the nicotinamide nucleotide transhydrogenase (Nnt) gene that plays a critical role in …
Reviewer #2 (Public Review):
In a previous paper, the authors demonstrated that the chikungunya virus (CHIKV)-induced arthritis (foot swelling) was attenuated in Gzma gene knock-out (C57BL/6J-Gzma-/-) mice compared to C57BL/6J (B6J) mice. In this paper the authors created C57BL/6J-GzmaS211A knock-in mice that have a mutation at GzmA active site to verify the role of GzmA in CHIKV-induced arthritis. The authors observed that foot swelling following CHIKV infection was not different between B6J-GzmaS211A and B6J. The authors then conducted whole-genome sequencing of B6-Gzma-/-and found that this mouse has a mixed (mosaic) genetic background of C57BL/6J and C57BL/6N (B6N). Of note, B6J uniquely has a loss-of-function deletion mutation of exons 8-12 of the nicotinamide nucleotide transhydrogenase (Nnt) gene that plays a critical role in scavenging mitochondrially generated ROS, while both B6-Gzma-/-and B6N have an intact (functional) Nnt gene. The authors then created B6N-NntΔexon8-12 mice and demonstrated that CHIKV arthritic foot swelling was ameliorated in the mice, reinforcing the contention that the amelioration of foot swelling in B6-Gzma-/-mice was rather due to functional Nnt gene introgressed into the strain from B6N. The authors also conducted RNA-seq analysis of the CHIKV foot and revealed that signatures of CHIKV arthritis in B6-Gzma-/-mice are reduced ROS and more importantly reduced cell (leukocyte) migration. The authors then re-evaluated the physiological role of GzmA by administration of polyinosinic:polycytidylic acid (poly(I:C)) to C57BL/6J-GzmaS211A knock-in mice and concluded that one of the roles of circulating GzmA is to activate monocyte/macrophages. Finally, the authors undertook a k-mer mining approach to transcriptome data deposited in the public database to find that ~27% of NCBI Sequence Read Archive (SRA) Run accessions and ~38% of BioProjects are labeled erroneously as collected from C57BL/6J. Based on these data the authors concluded that widespread discrepancy in Nnt genotypes complicates granzyme A and other knockout mouse studies.
This is a well-written paper containing interesting results. The conclusions of this paper are well supported by ample data obtained from an appropriately and carefully designed methodology.
This reviewer agrees with the authors' contention that "the C57BL/6J-GzmaS211A knock-in mice should allow assessment of the physiological function of GzmA without the confounding influence of differences in Nnt or other genes associated with the mixed genetic background" (Line 516). However, this reviewer considers that the more appropriate model would be C57BL/6N-GzmaS211A knock-in mice with the functional Nnt gene, given that most humans have a functional NNT gene (Line 547).
Also, the authors may be able to describe how C57BL/6J-GzmaS211A (or C57BL/6N-GzmaS211A) knock-in mice can be used to resolve the critical issues concerning granzyme A, such as target(s) of the extracellular target of the enzyme and molecular mechanisms of the enzyme on activation of monocyte/macrophage for the benefit of the reader. -
