Cell Metabolism
Letters NLRP1a Expression in Srebp-1a-Deficient Mice Motti Gerlic,1,2 Ben A. Croker,3 Louise H. Cengia,1 Mahtab Moayeri,4 Benjamin T. Kile,1,2 and Seth L. Masters1,2,* 1The
Walter and Eliza Hall Institute of Medical Research, Parkville 3052, Australia of Medical Biology, The University of Melbourne, Parkville 3010, Australia 3Division of Hematology/Oncology, Boston Children’s Hospital, Harvard Medical School, Boston MA 02115, USA 4Microbial Pathogenesis Section, Laboratory of Parasitic Diseases, NIAID, NIH, Bethesda, MD 20892, USA *Correspondence:
[email protected] http://dx.doi.org/10.1016/j.cmet.2014.02.002 2Department
The sterol regulatory element binding protein-1a (SREBP-1a) is a lipogenic transcription factor that differentially regulates the expression of inflammationrelated genes. One of the most profoundly downregulated genes in Srebp-1a-deficient mice was recently reported to be the inflammasome sensor Nlrp1a (Im et al., 2011). We obtained Srebp-1adeficient mice to cross with mice containing an Nlrp1a activating mutation to see if the Srebp-1a deficiency could reduce the incidence of their spontaneous inflammatory disease (Masters et al., 2012). We noticed while breeding the Srebp-1a/Nlrp1a mutant mice that no pups in any of the first litters had both Srebp-1a deleted and Nlrp1a mutated on the same allele. This suggested that the genes are in close proximity, and not recombining. Indeed, Srebp-1a is on chromosome 11: 60, 209, 846, and Nlrp1a is next door at chromosome 11: 71, 117, 950. This drew our attention to the fact that the original ES cells used to create Srebp-1a-deficient mice were from the 129/Ola strain. We and others have previously shown that in all 129 strains of mice so far tested, Nlrp1a is not expressed (Sastalla et al., 2013). We therefore genotyped the Nlrp1 locus using strain-specific primers for Nlrp1b, which distinguish 129/Ola from C57BL/6, and confirmed that in Srebp-1a-deficient mice the Nlrp1 locus is from the parental 129/Ola ES cell, not fully backcrossed to C57BL/6 (Figure S1A, lane 2). After considerable effort, we generated Srebp-1a-deficient mice, where the Nlrp1 locus is from C57BL/6, not 129/Ola (Figure S1A, lane 4). We find that in this backcrossed line of C57BL/6 Srebp-1a/ mice, Nlrp1a expression is now normal (Figure S1B). Genetic activation of NLRP1a triggers inflammasome formation and IL-1b production; however, stimuli that trigger NLRP1a are not yet known (Masters
et al., 2012). The Osborne laboratory (Im et al., 2011) reported that IL-1b production from Srebp-1a/ BMDM was reduced when stimulated by LPS+ATP, which is unusual as this treatment is known to activate NLRP3, not NLRP1 (Kovarova et al., 2012). Indeed, we verified that IL-1b production from NLRP3deficient cells treated with LPS+ATP was absent; however, macrophages from Srebp-1a/ mice, or backcrossed C57BL/6 Srebp-1a/ mice, had normal IL-1b production when stimulated with LPS, LPS+ATP (NLRP3 activation), or LPS+poly dAdT (AIM2 activation) (Figure S1C). The region of chromosome 11 where Srebp-1a is situated is quite gene dense, and so strain differences in nearby alleles may account for some of the other observations made with these mice. For example, in a microarray study of liver tissue, two of the five most differentially expressed genes (Atox1 and Ulk2) are in this region (Im et al., 2009). A second microarray study found that Centb1, Grap, Ccl8, and St6galnac1 are also differentially expressed, and these are near Srebp-1a, as well (Im et al., 2011). It is still possible that Srebp-1a can bind to the Nlrp1a promoter, as shown in the 2011 paper by Osborne and colleagues (Im et al., 2011), so our results are not directly contradictory. Moreover, the loss of a transcription factor such as Srebp-1a could have context-dependent effects on Nlrp1a expression that we have not investigated. However, the more fully backcrossed Srebp-1a-deficient mice we generated, with Nlrp1a inherited from the C57BL/6 strain, clearly show that Nlrp1a expression is unaltered at baseline in macrophages and that IL-1b production from these macrophages is also normal following inflammasome activation. Future studies of Srebp-1a/ mice should ensure that the
phenotype of interest is not due to insufficient backcrossing, and the mice we have generated will be useful in this endeavor. Similar caution is warranted in the study of mice knocked out for any gene in the vicinity of the Nlrp1 cluster on chromosome 11. As an example, our laboratories have genotyping data indicating that the iNOS knockout mice provided by Taconic Farms through mid-2005 carried an Nlrp1 locus inherited from the 129 mouse, while the knockout from Jackson Laboratories carried an Nlrp1 locus inherited from the C57BL/6 strain. Discrepancies in findings between iNOS knockout mice across various laboratories may be linked to differences in this region of chromosome 11. Another example with implications for the metabolism community is mice lacking arachidonate 15-lipoxygenase (ALOX15). The Nlrp1 locus in these mice is from the parental 129 ES cell and will be difficult to backcross as ALOX15 is only 600 kb away. Together with critical differences that inactivate caspase-11 in the 129 mouse and impact on analyses of caspase-1deficient mice (Kayagaki et al., 2011), our findings emphasize the importance of genetic fidelity in the analysis of mouse phenotypes, especially those that are inflammatory or metabolic in nature. SUPPLEMENTAL INFORMATION Supplemental Information includes one figure, Supplemental Experimental Procedures, and Supplemental Acknowledgments and can be found with this article online at http://dx.doi.org/ 10.1016/j.cmet.2014.02.002. REFERENCES Im, S.S., Hammond, L.E., Yousef, L., Nugas-Selby, C., Shin, D.J., Seo, Y.K., Fong, L.G., Young, S.G., and Osborne, T.F. (2009). Mol. Cell. Biol. 29, 4864–4872.
Cell Metabolism 19, March 4, 2014 ª2014 Elsevier Inc. 345
Cell Metabolism
Letters Im, S.S., Yousef, L., Blaschitz, C., Liu, J.Z., Edwards, R.A., Young, S.G., Raffatellu, M., and Osborne, T.F. (2011). Cell Metab. 13, 540–549. Kayagaki, N., Warming, S., Lamkanfi, M., Vande Walle, L., Louie, S., Dong, J., Newton, K., Qu, Y.,
Liu, J., Heldens, S., et al. (2011). Nature 479, 117–121. Kovarova, M., Hesker, P.R., Jania, L., Nguyen, M., Snouwaert, J.N., Xiang, Z., Lommatzsch, S.E., Huang, M.T., Ting, J.P., and Koller, B.H. (2012). J. Immunol. 189, 2006–2016.
Masters, S.L., Gerlic, M., Metcalf, D., Preston, S., Pellegrini, M., O’Donnell, J.A., McArthur, K., Baldwin, T.M., Chevrier, S., Nowell, C.J., et al. (2012). Immunity 37, 1009–1023. Sastalla, I., Crown, D., Masters, S.L., McKenzie, A., Leppla, S.H., and Moayeri, M. (2013). BMC Genomics 14, 188.
Response to Gerlic et al. Seung-Soon Im,1,4 Stephen G. Young,2 Manuela Raffatellu,3 and Timothy F. Osborne1,* 1Metabolic
Disease Program, Sanford-Burnham Medical Research Institute, Orlando, FL 32827, USA of Medicine and Human Genetics, David Geffen School of Medicine, University of California, Los Angeles, Los Angeles, CA 90095, USA 3Department of Microbiology and Molecular Genetics and Institute for Immunology, University of California, Irvine, Irvine, CA 92617, USA 4Present address: Keimyung University, School of Medicine, Department of Physiology, 1905 Shindang-Dong, Dalgubeol-Daero, Dalseogu, Daegu 704-701, Korea *Correspondence:
[email protected] http://dx.doi.org/10.1016/j.cmet.2014.02.016 2Department
Gerlic et al. contacted us a few months ago regarding the concerns expressed in the accompanying letter. We have communicated with them about their observations relative to conclusions from our original paper (Im et al., 2011) and for the field moving forward. Based on their data, we performed a separate DNA sequence analysis of the Nlrp1a locus in our SREBP-1a-deficient (SREBP-1aDF) mice and confirmed that it is derived from the 129 strain. However, the issues raised by this observation are almost certainly more complicated than the tight linkage between the Srebf1 and Nlrp1a loci and the extent of backcrossing. In fact, similar complications could affect observations regarding Nlrp1a function from another recent report from Gerlic and colleagues (Masters et al., 2012). In this study, the Nlrp1a locus from the C57BL/6 strain, a line where Nlrp1a and Nlrp1c are expressed in bone marrow-derived macrophages (BMDMs), was inserted into the BALB/c strain where Nlrp1a/Nlrp1c also appear to be silent in BMDMs (similar to the 129 strain). In addition to this study, there are numerous publications that suggest that strain differences at the murine Nrlp1 locus significantly influence responses to pathogens and inflammatory stimuli.
However, a major relevant issue that has not been solved is why the Nlrp1a locus in the 129 strain, and several others, including BALB/c, seems to be silenced at least in BMDMs cultured in vitro. To begin to address this issue, we compared the Nlrp1a DNA sequence and putative mRNA coding regions from the C57BL/6 and 129 strains (Figure S1). The alignment predicts almost complete identity at the protein level between the two strains, with the exception of only two amino acid differences, both of which correspond to residues that display variations between different mouse strains (Sastalla et al., 2013). Interestingly, the Nlrp1a 50 -flanking sequences from the two strains are even more highly conserved. The extensive conservation strongly suggests that the Nlrp1a coding sequence is intact in both strains and that major structural alterations surrounding the Nlrp1a gene are an unlikely explanation for the absence of Nlrp1a expression in 129 BMDM. This is important because it suggests the Nlrp1a locus is under evolutionary pressure to maintain the coding integrity and thus predicts that it is expressed under the appropriate circumstances in 129 mice, we would argue in response to SREBP-1a activation. Also, it should be noted that the absence
346 Cell Metabolism 19, March 4, 2014 ª2014 Elsevier Inc.
of 129-derived Nlrp1a transcripts in cultured macrophages is not reflective of Nlrp1a expression in an in vivo context where the SREBP-1aDF mice exhibit a profound inflammatory phenotype (Im et al., 2011). The data in Figures 5 and 6 of our original paper demonstrate that Nlrp1a is directly activated by SREBP-1a because reintroduction of SREBP-1a into the SREBP-1aDF macrophages through either adenovirus vector delivery or plasmid transfection activates Nlrp1a mRNA expression and restores LPSdependent IL-1b secretion to wild-type levels (Im et al., 2011). We recently repeated the activation experiment with identical results. These observations demonstrate that the Nlrp1a locus in the SREBP-1aDF strain can be expressed in isolated macrophages when SREBP-1a is reintroduced. Why Nlrp1a is not expressed in 129 BMDMs as well as from several other strains is an intriguing issue, one that deserves more investigation. Similarly, why further backcrossing to the C57BL/6 strain (which would alter many loci on chromosome 11 and elsewhere throughout the genome) would restore Nlrp1a expression deserves more study as well. With regard to a second issue, IL-1b secretion, data in our original paper
