and Immunity (2007) 8, 416–421 & 2007 Nature Publishing Group All rights reserved 1466-4879/07 $30.00 www.nature.com/gene

ORIGINAL ARTICLE Spontaneous mutations in the mouse Sharpin result in multiorgan inflammation, immune system dysregulation and dermatitis

RE Seymour1, MG Hasham1, GA Cox1, LD Shultz1, H HogenEsch2, DC Roopenian1 and JP Sundberg1 1The Jackson Laboratory, Bar Harbor, ME, USA and 2Department of Veterinary Pathobiology, Purdue University, West Lafayette, IN, USA

Homologues of the SHARPIN (SHANK-associated RH domain-interacting ) gene have been identified in the human, rat and mouse genomes. SHARPIN and its homologues are expressed in many tissues. SHARPIN protein forms homodimers and associates with SHANK in the post-synaptic density of excitatory neurotransmitters in the brain. SHARPIN is hypothesized to have roles in the crosslinking of SHANK and in enteric nervous system function. We demonstrate that two independently arising spontaneous mutations in the mouse Sharpin gene, cpdm and cpdmDem, cause a chronic proliferative dermatitis phenotype, which is characterized histologically by severe inflammation, eosinophilic dermatitis and defects in secondary lymphoid organ development. These are the first examples of disease-causing mutations in the Sharpin gene and demonstrate the importance of SHARPIN protein in normal immune development and control of inflammation. Genes and Immunity (2007) 8, 416–421; doi:10.1038/sj.gene.6364403; published online 31 May 2007

Keywords: Sharpin; lymphoid organ development; B cell; eosinophil

Introduction between markers D15Mit156 and D15Mit2 using an intraspecific cross of C57BL/KaLawRij-cpdm/cpdm with The chronic proliferative dermatitis mutation (cpdm) BALB/cJ. An intersubspecific cross using CAST/EiJ was was first observed as a spontaneous genetic disease in created to increase polymorphisms in this region for C57BL/KaLawRij mice at the TNO Institute, Rijswijk, high-resolution mapping; however, the skin phenotype The Netherlands. Mutant mice have a shortened lifespan, in some mice was highly variable, likely due to the runted appearance and a characteristic progressive contribution of CAST modifier genes.4 To avoid these dermatitis with alopecia that first becomes clinically complications, we developed a congenic strain, C.CAST- apparent at 3–4 weeks of age. Males and females are (D15Mit156-D15Mit2)/Sun, moving the CAST critical affected equally. The mutant phenotype includes multi- interval for cpdm onto the BALB/cJ background to organ inflammation with eosinophilia, defective TH1 advance fine mapping. This congenic strain retained a cytokine production, splenomegaly and dermatitis high rate of polymorphism in the m.Chr.15 interval of with granulocytic infiltrates, ortho- and parakeratotic 1,2 interest compared to C57BL/KaLawRij-cpdm/cpdm but hyperplasia, and follicular dystrophy. Peyer’s patches lacked the purported unlinked CAST-derived modifier are absent in adult mice, and remaining secondary genes altering the cpdm phenotype. Using the original lymphoid organs lack B-cell follicles, follicular dendritic (CAST/EiJ þ / þÂC57BL/KaLawRij-cpdm/cpdm) F2 in- cells and germinal centers. The B- and T-cell areas of tercross and over 1100 mice from the new C.CAST- the spleen are poorly defined, and the marginal zone (D15Mit156-D15Mit2)/Sun  C57BL/KaLawRij-cpdm/cpdm is absent. B-cell isotype switching and immuno- F2 intercross, seven critical recombinants were identified, globulin production is also affected, with significantly and the cpdm genetic interval was narrowed to o0.2 cM diminished serum immunoglobulins (Ig) IgG, IgA 3 (372 kb). The cpdm interval contained 21 identified genes and IgE. according to the Ensembl (http://www.ensembl.org/ index.html) mouse genome database annotation (Sup- plementary Table 1). The interval is located in a region of Results conserved synteny with human 8q24.3 We initially localized the cpdm gene to an approximately (h.Chr.8q24.3); comparison of the mouse and human 1.2 cM interval on mouse Chromosome 15 (m.Chr.15) genomic sequences demonstrates a complete conserva- tion of gene order and gene content between the two species. Correspondence: Dr RE Seymour, The Jackson Laboratory, An independent spontaneous mutation resulting in a 600 Main Street, Bar Harbor, ME 04609, USA. E-mail: [email protected] similar phenotype was identified in OcB-3/Dem recom- Received 20 February 2007; revised 1 May 2007; accepted 3 May binant congenic mice at The Netherlands Cancer 2007; published online 31 May 2007 Institute, Amsterdam. Spontaneous mutations in the mouse Sharpin gene RE Seymour et al 417 Allelism of this second mutation with cpdm was con- the 14-bp deletion disrupts the reading frame and results firmed by heterozygous matings (C57BL/KaLawRij þ / in an early stop codon beginning at position 552. To cpdm  Ocb-3/Dem þ /cpdm) that produced affected F1 ensure that this change represented a mutation from the offspring at expected Mendelian ratios. This second strain of origin, we obtained and sequenced genomic strain was given the gene symbol cpdmDem,4 and back- DNA from wild-type Ocb3/Dem mice (kind gift of Dr M crossed onto the BALB/cByJ inbred strain for 10 Lipoldova, Institute of Molecular Genetics, Academy of generations for colony maintenance purposes, creating Sciences, Czech Republic), and determined that these the CBy.OcB3 congenic strain. changes are not seen in the wild-type allele (Figure 2). To Genes in the critical region for cpdm were initially validate a relationship between the Ocb-3/Dem control screened by real-time quantitative PCR (QPCR) assay, DNA and the CBy.OcB3-cpdmDem/cpdmDem mice main- which suggested Sharpin as a strong candidate, the only tained at The Jackson Laboratory, we also sequenced the gene in the interval that was significantly downregulated nearby Slc39a4 gene amplified from the control DNA to in testes or spleen of both C57BL/KaLawRij-cpdm/cpdm look for the presence of 10 benign polymorphisms that and CBy.OcB3-cpdmDem/cpdmDem mice as compared to the we observed previously between CBy.OcB3-cpdmDem/ same tissues in their respective controls (data not cpdmDem mice and the other strains used (data not shown). We began to examine genes in the interval by shown). These polymorphisms were also confirmed in direct sequence analysis of cDNA, using sequence the Ocb-3/Dem control mouse DNA, verifying the close obtained from NCBI m34 via Ensembl for comparison. relationship to CBy.OcB3-cpdmDem/cpdmDem. Slc39a4, Gpaa1, Hsf1 and Sharpin were completed, and Western blot analysis of spleen tissue homogenates independent mutations were identified in exon 1 of from pre-clinical (o12-day old) mice with anti-SHARPIN Sharpin for each of the two cpdm alleles. Examination of antibody demonstrated an approximately 40 kDa the complete genomic sequence of Sharpin revealed that protein, present in control mice but absent in both the these changes were not found in control mice of the C57BL/KaLawRij-cpdm/cpdm and CBy.OcB3-cpdmDem/ parent strains nor in C57BL/10J or BALB/cByJ strains. In cpdmDem mutants (Figure 3). The size is consistent with the C57BL/KaLawRij-cpdm/cpdm mice, a single the predicted 39.9 kDa protein sequence for the mouse, deletion was observed at the 30-end of exon 1 (Figure 1), and slightly smaller than the 45-kDa major band resulting in a shift of the open reading frame predicted demonstrated by Lim et al. (2001) in rat tissues, probably to cause an early stop codon beginning at position reflecting species differences. Minor bands (o37 and 624 (numbered sequence is based on Ensembl cDNA 440 kDa) similar to those described for the rat were also transcript ENSMUST00000023211). The CBy.OcB3- observed in adult wild-type mice (data not shown). The cpdmDem/cpdmDem mouse showed a C to A transition at larger bands appear to be non-specific, as they were position 434 near the 50-end of Exon 1, followed by a visible in adult cpdm/cpdm mice and controls, though 14-bp deletion of nucleotides 438-451 (Figure 2). The barely detectable in mice o2 weeks of age. The smaller transition does not result in an amino acid change, but bands were not found in young mutant mice, and barely

Figure 3 Western blot using polyclonal rabbit anti-SHARPIN antibody demonstrates absence of the protein in mutant mice. (a) Anti-SHARPIN antibody. (b) Anti-BETA ACTIN antibody. Protein size marker is shown at top left; numbers below lanes indicate the Figure 1 Chromatogram showing the C57BL/KaLawRij-cpdm/ ratio of SHARPIN to BETA ACTIN loading control. Lanes 1 and 2, cpdm mutation. There is a single base pair deletion in exon 1 C57BL/KaLawRij þ / þ ; lane 3 CBy.OcB3/Dem þ / þ . Lanes 4 (box). (a) C57BL/KaLawRij þ / þ genomic sequence. (b) cpdm/ and 5, C57BL/KaLawRij-cpdm/cpdm; lane 6, CBy.OcB3-cpdmDem/ cpdm. cpdmDem.

Figure 2 Composite image of chromatograms demonstrating the C to A transition (box) and the 14-basepair deletion in exon 1 of the CBy.OcB3-cpdmDem/cpdmDem Sharpin gene. (a) BALB/cByJ þ / þ .(b) OcB3/Dem þ / þ .(c) CBy.OcB3-cpdmDem/cpdmDem.

Genes and Immunity Spontaneous mutations in the mouse Sharpin gene RE Seymour et al 418 detectable in adult mutants. To further confirm that Discussion mutations in Sharpin cause the cpdm phenotype, we used in vivo complementation with a bacterial artificial SHARPIN and its homologues are expressed in multiple chromosome (BAC) clone expressed in FVB/NJ trans- tissues in humans and mice (GNF SymAtlas, v.1.1.1, genic mice created for that purpose. The BAC clone http://www.wombat.gnf.org/SymAtlas/). Orthologues RP24-173I23 was selected, as it carries a C57BL/B6 (B6)- are also found in rats, dogs, chimpanzees (Mouse derived chromosome 15 segment including Sharpin Genome Informatics, http://www.informatics.jax.org/), (UCSC Genome browser http://www.genome.ucsc. and probably other species (Ensembl, http://www. edu/index.html?org ¼ Mouse, March 2005). Sharpin is ensembl.org/). SHARPIN is enriched in the post-synap- located in a gene-rich region on m.Chr.15 and the BAC tic density of excitatory synapses in the rat brain, where carries other additional genes: 5-oxoprolinase (Oplah), it may interact with SHANK protein.5 SHARPIN was exosome component 4 (Exosc4), cytochrome c-1 (Cyc1), also hypothesized to play a role in enteric motor GPI anchor attachment protein 1 (Gpaa1), brain protein16 transmission in humans and mice.6 Here we report two (Brp16), MAF1 homolog (S. cerevisiae)(Maf1), scleraxis independently occurring, spontaneous mutations in the (Scx), testis-specific serine kinase 5 (Tssk5) and mouse Sharpin gene, cpdm and cpdmDem. Cpdm/cpdm mice D330001F17Rik, with unknown function. Ensembl demonstrate defective secondary lymphoid organ devel-

further predicts two novel genes in the chromosomal opment, multiorgan inflammation with a TH2 dominated region found on the BAC. None of these genes has been cytokine milieu, dermatitis, decreased fertility and associated previously with normal immune function. shortened lifespan. We have not observed neurological Since the C57BL/KaLawRij sequence is nearly identical defects in cpdm/cpdm mice; however, the skin phenotype to B6 in this region, and the cpdm mutation is purely becomes obvious at 3–5 weeks of age and its rapid recessive, a B6 transgenic founder line would present progression may preclude any casual observation of difficulties in genotyping an F2 mutant mouse rescued subtle behavioral changes, particularly of symptoms by the B6 transgene vs a heterozygous mouse bearing the associated with learning or memory deficits. Affected transgene. The FVB/NJ strain was chosen to allow for mice over 12 weeks of age were not examined. selection of polymorphic markers for genotyping, for The function of SHARPIN is unknown; however, Lim large average litter production, relative ease in obtaining et al. (2001) reported that the N terminus bears a coiled- many synchronous embryos and larger egg pronuclei, coil domain that may be associated with dimerization or simplifying microinjection. DNA binding functions. The C terminus shares homol- BAC-positive founder animals were identified by ogy with the protein kinase C binding protein, RBCK1; in PCR using a primer set designed to amplify a B6/FVB RBCK1, this region is associated with DNA binding and polymorphic region located on the BAC close to Sharpin, transcriptional activity.5 SHARPIN also contains a C- in addition to primers specific for the bacterial region of terminal zinc-finger domain highly similar to that of a the BAC. None of the BAC-bearing transgenic mice Ran-binding protein. In Ran-BP, this domain binds to demonstrated gross abnormalities, and both males and Ran-GDP, and is associated with nuclear transport. This females were fertile. Three FVB transgenic lines were region is conserved between rat and mouse, and one established and subsequently crossed with mice bearing transcript of the human gene.6 The mouse mutations the C57BL/KaLawRij-cpdm allele to determine whether discussed here both result in early truncation; the cpdm the transgene could ‘rescue’ (prevent onset of) the mutation causes truncation immediately following the dermatitis and other immunologic defects. Any F2 coiled-coil domain, and in cpdmDem all of these predicted offspring of these crosses bearing two mutant Sharpin domains are lost (Figure 5). alleles and also carrying the transgene were expected to These are the first known examples of disease-causing have a wild-type phenotype. mutations in Sharpin and provide valuable new informa- Ten mice bearing two Sharpin mutant alleles and tion for the study of secondary lymphoid organ devel- positive for the BAC were examined, and all of them opment and control of inflammatory processes. The lack were grossly and histologically free of the disease of cellular organization in these tissues is likely to be a phenotype at 12 weeks of age. These included mice contributing factor in the defective humoral response of from each of the three founder lines. Another 10 mice cpdm/cpdm mice, since intimate communication between bearing two Sharpin mutant alleles but negative for B cells, T cells, and antigen-presenting cells is critical for the BAC developed the cpdm phenotype (Figure 4). The this function. Sharpin is highly expressed in B and T cells; cpdm phenotype was completely penetrant on the however, transplantation of mutant bone marrow into mixed F2 genetic background in absence of the trans- adult or neonatal wild-type mice does not confer the gene. Not surprisingly from our experience with the disease phenotype,1 suggesting a stromal component early mapping crosses, we observed an influence of as well. the FVB background on the severity of the disease We have confirmed two independent mutations of the phenotype, resulting in milder skin disease with mouse Sharpin gene in cpdm and cpdmDem mutant mice, delayed age of onset, and potentially extended lifespan demonstrated the loss of SHARPIN protein expression in in some mice. One mouse had two small submucosal these mutant mice, and established transgenic rescue of aggregates of lymphoid tissue in the distal small the disease phenotype by BAC microinjection. We are intestine, but no organized follicular structures presently working to create time- and tissue-specific were present (Figure 4e). We are presently aging cohorts targeted mutations of SHARPIN in mice to determine the of mice from this colony and developing a backcross of specific role of the protein in affected tissues. the BAC onto the C57BL/KaLawRij-cpdm/cpdm back- There are currently no reports of orthologous human ground to determine long-term effects of the mixed diseases mapping to the h.Chr 8q24.3 region, and no genetic background. disease known to be associated with mutations in human

Genes and Immunity Spontaneous mutations in the mouse Sharpin gene RE Seymour et al 419

Figure 4 BAC 173I23 rescues phenotypic defects in the skin and secondary lymphoid tissues of cpdm/cpdm mice. Hematoxylin and eosin. (a) The skin of wild type and BAC rescued cpdm/cpdm mutant mice (left columns) has a thin epidermis with no evidence of dermal inflammation. By contrast, the epidermis of BAC-negative cpdm/cpdm mutant mice (right columns) is typically hyperplastic with marked dermal inflammation, hyperkeratosis and numerous apoptotic keratinocytes. Background modifier genes from the FVB strain are likely responsible for the mild mutant phenotype occasionally observed in BAC-negative F2 mutant mice (far right). This effect is most evident in the skin, however mild inflammation is present histologically and scattered apoptotic keratinocytes are seen. (b) Organized Peyer’s patches are found in the antimesenteric wall of the small intestine in wild type and BAC-rescued cpdm/cpdm mice, but absent in BAC-negative cpdm/cpdm mice. (c) Cortical B cell follicles are seen in the lymphocyte-rich lymph nodes of wild type and BAC rescued cpdm/cpdm mice. These structures are effaced in BAC-negative cpdm/cpdm mice; reticuloendothelial cell hyperplasia and inflammatory cells are common. (d) The splenic white pulp is well defined in wild type and BAC rescued cpdm/cpdm mice. BAC-negative cpdm/cpdm mice lack this normal architecture. (e) Small aggregates of lymphoid tissue were observed in submucosa of the ileum of one BAC-negative F2 ‘mild’ cpdm/cpdm mutant mouse; however, no organized lymphoid tissues (that is, Peyer’s patches) were found.

SHARPIN. Future plans include screening of human heterozygote. Normal littermate controls were wild type patients with immunodeficiency syndromes and chronic as determined by test matings or sequencing of DNA, eosinophilic inflammation for mutations in the SHARPIN or heterozygotes, as indicated. gene. The original OcB-3/Dem mutation occurred at the Institute of Molecular Genetics, Academy of Sciences, Czech Republic. CBy.OcB3-cpdmDem/cpdmDem and their Materials and methods BALB/cByJ controls were maintained at The Jackson Laboratory under conventional SPF conditions. Both Mice male and female homozygotes of the CBy.OcB3 strain are Colonies of C57BL/KaLawRij-cpdm/cpdm and wild- infertile (data not shown) and the colony was maintained type mice were maintained at The Jackson Laboratory by heterozygous matings determined by genotype. under conventional specific pathogen-free (SPF) conditions. Female cpdm/cpdm mice are infertile,7 and BAC microinjection and rescue mutant colonies were maintained by mating homozy- The BAC was purchased from BACPAC resources gous male  heterozygous female, or heterozygote  (Oakland, CA, USA) and prepared using a Qiagen

Genes and Immunity Spontaneous mutations in the mouse Sharpin gene RE Seymour et al 420 (The Whitehead Institute for Biomedical Research). Primers were designed in accordance with universal thermocycling requirements for reverse transcriptase- PCR using the ABI 7900HT. Primer sequences were BLASTed using the UCSC Genome Browser (http:// www.genome.ucsc.edu/) to validate their specificity, and tested by PCR using control mouse genomic and cDNA. RNA was extracted from the spleen and testes of 3 C57BL/KaLawRij-cpdm/cpdm, 3 C57BL/KaLawRij þ / þ , 3 CBy.OcB3-cpdmDem/cpdmDem and 3 CBy.OcB3 þ / þ adult male mice, aged 9–11 weeks. Briefly, spleen and testis samples were stored in RNAlater (Ambion, Austin, TX, USA) as per the manufacturer’s instructions and homogenized in TRIzol (Invitrogen, Carlsbad, CA, USA). Total RNA was isolated by standard TRIzol methods, and quality was assessed using an Agilent 2100 Bioanalyzer instrument and RNA 6000 Nano LabChip assay. Total RNA was then treated with RNase-free DNase I (Qiagen) according to the manufacturer’s protocol.1 mg of total RNA was then reverse transcribed with random decamers and M-MLV RT using the Message Sensor RT Kit (Ambion). Each reaction consisted of 5 mlof2 SYBR green Figure 5 Mutant proteins predicted to result from frameshift master mix (ABI, Foster City, CA, USA), 3.0 ml dH2O, mutations of the Sharpin gene in cpdm mice. Black, source cDNA; 1.5 ml of 0.5 mm forward and reverse primers and 0.5 mlof green, C57BL/KaLawRij-cpdm/cpdm; red, CBy.OcB3-cpdmDem/ diluted (1:10) cDNA. ABI 384-well plates with optical Dem cpdm . *, amino acid change; #, stop codon; - - -, no sequence adhesive covers (ABI) were used for three technical predicted. Domain predictions published in Lim et al.5 and Daigo et al.6 ___, Coiled-coil domain; y, RBCK1-like domain; box replicates each of three mutants and three controls from indicates Ran-BP-like zinc-finger domain. each cpdm strain. Quantitative real-time PCR assay was run on the ABI Prism 7900HT Signal Detection System v2.0 using default conditions. Large-Construct Kit (Valencia, CA, USA) according to Global pattern recognition software was used to the manufacturer’s instructions. The presence of Sharpin compare the change of expression of each gene in in the BAC was confirmed before and after kit purifica- the interval normalized to every other gene in the tion by PCR, using primers previously designed for array. In this way, genes with significant expression sequencing the gene. Sequencing gel-purified PCR changes were readily identified from healthy baseline products following BAC purification verified that Sharpin profiles.8,9 was present in the final product. Microinjection of the BAC was accomplished through the TJL Cell Biology and Genomic sequence analysis Microinjection Service. This resulted in 434 FVB/NJ Genomic DNA was obtained from tail tips in accordance embryos implanted into 20 recipient females, yielding 37 with ACUC approved protocols. DNA was extracted pups, which were weaned and delivered to our colony. using a Dneasy Tissue Kit (Qiagen) in accordance with Founder mice were identified by PCR using primer sets the manufacturer’s instructions. Genomic sequence of specific for the vector (left primer – GGTCGTTTGACTG candidate genes was exported from Ensembl and over- GACGATT; right primer – GCCCTTAACCCTGTGAC lapping primer pairs were designed to cover the entire AAA) and a marker polymorphic between the B6 (125 sequence, using the Primer Express v2.0 (ABI) or Primer3 BP) and FVB (approximately 140 bp) strains, located software programs. DNA was amplified by PCR and gel- close to Sharpin on the BAC (left primer: 50-CAGCTTTT purified using a QIAquick Gel Extraction Kit (Qiagen). TAAACTGTGGGTCA-30; right primer: 50-AGCGTAG Purified DNA products were submitted with 6.6 pmol TATGTGTGTGCAAA-30). To establish rapidly a colony of appropriate primer to the DNA Sequencing Service while following segregation of the cpdm allele by at The Jackson Laboratory and analyzed by Big Dye genotype, heterozygous mice from the C.CAST- Terminator Cycle Sequencing Chemistry on an ABI (D15Mit156-D15Mit2)/Sun  C57BL/KaLawRij-cpdm/cpdm 3700. The resulting sequences were evaluated for mapping cross were bred to FVB/NJ BAC þ mice. Some potential mutations using Sequencher v4.1.4 software FVB/NJ BAC þ females were also bred to C57BL/ (Gene Codes Corp., Ann Arbor, MI, USA). KaLawRij-cpdm/cpdm males. F1 offspring genotyped at these loci as FVB/cpdm BAC þ were crossed with FVB/ Western blotting cpdm BACÀ mice, and F2 offspring were identified as Rabbit antisera containing primary antibody to FVB/FVB BACÀ, cpdm/cpdm BACÀ, cpdm/cpdm BAC þ SHARPIN was the kind gift of Dr Eunjoon Kim (rescue) or heterozygous (normal) mice with and without (Korea Advanced Institute of Science and Technology, the BAC. yuseong-gu, Daejeon, Korea). The anti-sharpin rabbit serum was purified using a 1 ml HiTrap protein G Quantitative real-time PCR column from Amersham Biosciences (Piscataway, NJ, Sequences for candidate genes were exported from USA). The serum was diluted 1:1 with 20 mM sodium Ensembl and copied into Primer3 software v0.1. beta 1a phosphate, 150 mM NaCl pH 7.0 and centrifuged for

Genes and Immunity Spontaneous mutations in the mouse Sharpin gene RE Seymour et al 421 10 min at 10 K g before filtering the supernatant through Acknowledgements a Gelman Sciences LC13 0.45 mm syringe filter. The serum was loaded onto the column using a peristatic This work was supported by grants from the National pump at 0.4 ml/min and washed with the phosphate Institutes of Health (Grant no. AI060707; AR49288; buffer at 0.7 ml/min until zero absorbance at 280 nm. The CA34196). We thank Drs D Serreze and K Johnson for column was then reversed and eluted with 0.1 M glycine their critical review of this manuscript, Drs E Kim and M (pH 2.5) collecting 0.5 ml fractions. Fractions were Lipoldova for contributing materials mentioned in the immediately neutralized with 20% volume of 1 M Tris text, and D Shaffer, A Brown and J Bubier for technical (pH 8.3). The 280 nm absorbing fractions were pooled assistance with real-time QPCR arrays. and dialyzed against phosphate-buffered saline with 0.02% azide. Cell lysates were obtained by homogenizing spleen References tissue of p2-week old, age and sex-matched mutant and control mice of each strain in Tris/NaCl western 1 HogenEsch H, Gijbels M, Offerman E, Hooft Jv, Bekkum Dv, blot lysis buffer with IGEPAL CA-630 lysis detergent Zurcher C. A spontaneous mutation characterized by chronic and a protease inhibitor cocktail (Roche, Indianapolis, proliferative dermatitis in C57BL mice. Am J Pathol 1993; 143: IN, USA). Proteins were quantitated using a Bio-Rad 972–982. DC Protein detection assay (Bio-Rad, Hercules, CA, 2 HogenEsch H, Torregrosa SE, Boggess D, Sundberg BA, Carroll USA) and read by Spectramax 250 (Molecular Devices, J, Sundberg JP. Increased expression of type 2 cytokines in Sunnyvale, CA, USA). Sodium dodecyl sulfate-polyacryl- chronic proliferative dermatitis (cpdm) mutant mice and amide gel electrophoresis (SDS-PAGE) gel electro- resolution of inflammation following treatment with IL-12. Eur J Immunol 2001; 31: 734–742. phoresis and membrane transfer were carried out by 3 HogenEsch H, Janke S, Boggess D, Sundberg JP. Absence of standard methods using 12% PAGE Gold Precast Peyer’s patches and abnormal lymphoid architecture in chronic gels (Cambrex BioScience, Rockland, ME, USA) and proliferative dermatitis (cpdm/cpdm) mice. J Immunol 1999; nitrocellulose membranes (Immobilon-P, Millipore, 162: 3890–3896. Bedford, MA, USA). Membranes were blocked with 4 Sundberg JP, Boggess D, Shultz LD, Fijneman R, Demant P, 5% dry milk in Tris-buffered saline Tween (TBST). HogenEsch H et al. The chronic proliferative dermatitis mouse Anti-SHARPIN antibody (1.84 mg/ml) was used at mutation (cpdm): mapping of the mutant gene locus. J Exp 1:500 in TBST. Anim Sci 2000; 41: 101–108. The secondary antibody was horseradish-peroxidase 5 Lim S, Sala C, Yoon J, Park S, Kuroda S, Sheng M et al. Sharpin, conjugated donkey anti-rabbit IgG (Research Diagnostics, a novel postsynaptic density protein that directly interacts with the shank family of proteins. Mol Cell Neurosci 2001; 17: Inc., Flanders, NJ, USA) at 1:25 000 dilution in TBST. 385–397. Membranes were developed using electrochemilumines- 6 Daigo Y, Takayama I, Ward SM, Sanders KM, Fujino MA. Novel cence þ Western Blot Detections System (Amersham human and mouse genes encoding a shank-interacting protein Biosciences). and its upregulation in gastric fundus of W/WV mouse. As a loading control, antibodies were stripped J Gastroenterol Hepatol 2003; 18: 712–718. from the membranes using a 100 mM 2-ME, 2% 7 Gijbels M, Zurcher C, Kraal G, Elliot G, HogenEsch H, sodium dodecyl sulfate, 62.5 mM solution at 501C for Schijff G et al. Pathogenesis of skin lesions in mice with chronic 30 minutes and reblotted as described above, using proliferative dermatitis (cpdm/cpdm). Am J Pathol 1996; 148: mouse IgG1 anti-b actin antibody at a 1:5000 dilution 941–950. 8 Akilesh S, Shaffer DJ, Roopenian D. Customized molecular (clone AC-15; Sigma, St Louis, MO, USA) and phenotyping by quantitative and pattern HRP-conjugated goat anti-mouse secondary at 1:10 000 recognition analysis. Genome Res 2003; 13: 1719–1727. (A-0412; Sigma). Western blot gel analysis was per- 9 Roopenian DC. Global pattern recognition algorithm. World formed using Image J 1.34 s software (http://www.rsb. Wide Web. http://www.jax.org/staff/roopenian/labsite/gene_ info.nih.gov/ij/). expression.html2005.

Supplementary Information accompanies the paper on Genes and Immunity website (http://www.nature.com/gene)

Genes and Immunity