Ubiquitin-conjugating enzyme Ubc13 is a critical component of TNF receptor-associated factor (TRAF)-mediated inflammatory responses
- Toru Fukushima,
- Shu-ichi Matsuzawa,
- Christina L. Kress,
- Jean Marie Bruey,
- Maryla Krajewska,
- Sophie Lefebvre,
- Juan M. Zapata,
- Ze'ev Ronai, and
- John C. Reed*
-
Edited by Tak Wah Mak, University of Toronto, Toronto, ON, Canada, and approved February 16, 2007 (received for review January 19, 2007)
Abstract
Ubc13 is a ubiquitin-conjugating enzyme responsible for noncanonical ubiquitination of TNF receptor-associated factor (TRAF)-family adapter proteins involved in Toll-like receptor and TNF-family cytokine receptor signaling, which are regulators of innate immunity. Gene ablation was used to study the function of Ubc13 in mice. Whereas homozygous ubc13 gene disruption resulted in embryonic lethality, heterozygous ubc13 +/− mice appeared normal, without alterations in immune cell populations. Haploinsufficient ubc13 +/− mice were resistant to lipopolysaccharide-induced lethality, and demonstrated reduced in vivo ubiquitination of TRAF6. Macrophages and splenocytes isolated from ubc13 +/− mice exhibited reduced lipopolysaccharide-inducible cytokine secretion and impaired activation of TRAF-dependent signal transduction pathways (NF-κB, JNK, and p38 MAPK). These findings document a critical role for Ubc13 in inflammatory responses and suggest that agents reducing Ubc13 activity could have therapeutic utility.
Ubc13 is a ubiquitin-conjugating enzyme (E2) that catalyzes the attachment of unusual polymers of ubiquitin (E1) onto target proteins, where the ubiquitin chain is linked by noncanonical lysine 63 (1–3) instead of canonical lysine 48 (4). Unlike K48-linked ubiquitin polymers, these K63-linked ubiquitin chains are not substrates for the proteasome (5, 6). Instead, K63-linked polyubiquitination plays roles in protein activation, protein interactions, and subcellular targeting of proteins (7–9).
Ubc13 has been implicated in two cellular processes. First, in collaboration with cofactor Uve1A, it binds the RING domains of TNF receptor (TNFR)-associated factor (TRAF)-family adapter proteins and promotes activation of protein kinases involved in signaling by TNFRs (10, 11) and by Toll-like receptors (TLRs) (12). Second, in collaboration with cofactor Mms2, Ubc13 participates in DNA replication and repair pathways, probably by modifying proliferating cell nuclear antigen (PCNA), Rad5, and Pol30 (13–15). With regard to the role of Ubc13 in TRAF-mediated signaling, cytokines that activate TRAF2 result in its noncanonical ubiquitination and subsequent activation of the kinase JNK. Ubc13 modification of TRAF2 was linked to the translocation of TRAF2 to insoluble membrane domains (9), suggesting a role for noncanonical ubiquitination in targeting TRAFs to specified signaling compartments. Ubc13 was also associated with the activation of NF-κB through its effect on TRAF6, resulting in activation of the kinase transforming growth factor β-activated kinase 1 (TAK1) (8).
Genes encoding six different TRAF-family adapter proteins are found within the human genome. TRAFs are multidomain proteins that contain an N-terminal RING domain (with the exception of TRAF1), followed by a series of zinc finger domains, and then the TRAF domain, a β-strand fold that binds several TNFRs or other proteins that become recruited to TNFR and TLR receptor complexes (16). Binding of Ubc13 to TRAFs depends on the RING domain, and mutant versions of TRAF2, 5, or 6 that lack the RING domain function instead as dominant-negative inhibitors of signaling by the TNFRs and TLRs/interleukin receptors (ILRs) that bind them (7, 9, 17, 18). Reducing Ubc13 expression by using RNA interference (RNAi) has demonstrated its role in the regulation of stress kinases and of the NF-κB pathway (9).
To explore the cellular functions of Ubc13 in the context of innate immune responses, we used the method of targeted gene ablation in mice. We show here that, whereas homozygous ablation of the gene encoding Ubc13 results in embryonic lethality, heterozygous mice have normal phenotypes, despite reduced levels of Ubc13 protein. However, hemizygous ubc13 +/− mice exhibit marked reductions in responsiveness to challenge with lipopolysaccharide (LPS), which was used as a representative TRAF-dependent stimulus. Macrophages and lymphocytes cultured from these mice also displayed blunted responses to LPS and TNF with respect to cytokine secretion and various TNFR/TLR-mediated signal transduction events. The findings therefore document a rate-limiting role for Ubc13 in signaling by TNFRs and TLRs and furthermore suggest that Ubc13 may be an attractive target for design of new therapeutic agents for treatment of certain inflammatory and autoimmune disorders.
Results
Ubc13+/− Mice Are Born Normally and Exhibit No Abnormalities in Immune Cell Populations.
To test the in vivo function of Ubc13 in mice, we turned to a commercially available collection of murine ES cell clones that contain single retroviral insertions into the mouse genome, identifying a clone in which a retroviral provirus is integrated into the first intron of the ubc13 gene in reverse orientation (Fig. 1 A). These ES cells were used to generate heterozygous mice, and the retrovirus-targeted ubc13 gene was confirmed by PCR analysis of genomic DNA by using specific primers (Fig. 1 B) and by Southern blot analysis (Fig. 1 C). Breeding together of heterozygous ubc13 +/− mice never resulted in any homozygous ubc13 −/− progeny in >36 matings. Analysis of embryos from timed pregnancies suggested that ubc13 −/− embryos die or fail to progress very early, probably before embryonic day 5.5. In contrast, heterozygous ubc13 +/− were born with normal Mendelian frequencies and appeared normal at birth, showing no gross developmental abnormalities or differences in size or weight compared with wild-type ubc13 +/+ littermates. Organ sizes were also not different, including thymus and spleen, which contained similar numbers of mononuclear cells in ubc13 +/− and ubc13 +/− mice (Fig. 1 D).
Haploinsufficiency of the ubc13 gene does not alter immune cell populations. (A) The targeting vector, wild-type ubc13 allele, and targeted allele are depicted. The vector was targeted to the first intron of the ubc13 locus. The closed rectangles denote exons of ubc13. (B) PCR analysis of genomic DNA extracted from mouse tails. Primers were designed to amplify the regions of wild-type and mutant alleles. PCR products of wild-type and mutated (KO) alleles are shown for two mice of each genotype. The genotypes of mice are presented above the lane. (C) Southern blot analysis was performed by using genomic DNA extracted from mouse embryo fibroblasts (MEFs) and digested with BamHI. The blot was hybridized with the probe shown in A. The expected sizes of the bands corresponding to wild-type and KO alleles are indicated (in kilobase pairs). (D) Comparison of total cell numbers of the thymocytes and splenocytes in ubc13 +/+ and ubc13 +/− mice (mean ± SD; n = five female 2- to 3-month-old mice). (E–H) Flow cytometry analysis of lymphocyte subpopulations in bone marrow (E), thymus (F), spleen (G), and peritoneum (H) of ubc13 +/+ and ubc13 +/− mice. (E) IgM and B220 expression of the total bone marrow population. (F) Thymocytes were stained with antibodies to CD4, CD8, TCRγδ, DX5, B220 (lineage marker, Lin), CD25, and CD44. (Upper) CD4 and CD8 expression of the total thymocyte population. (Lower) CD44 and CD25 expression (gated in Lin− cells). (G) (Top) CD3 and B220 expression of the total splenocyte population. (Middle) IgM and IgD expression of the total splenocyte population. (Bottom) CD21 and CD23 expression of B220+ population. (H) CD5 and B220 expression of the peritoneal population. Numbers in quadrants indicate percentage of positive cells in that region. Data are representative of three different experiments using age- and sex-matched mice.
We assessed various lymphocyte populations in ubc13 +/− mice by immunofluorescence-based detection of surface markers, in conjunction with flow cytometry analysis. No differences between wild-type ubc13 +/+ and heterozygous ubc13 +/− mice were observed with respect to the proportions of pre-B cells in bone marrow (B220/sIgM) (Fig. 1 E), thymocyte subtypes defined by CD4, CD8, CD25, and CD44 surface markers (Fig. 1 F), and mature T cells (CD3), B cells (IgM/IgD), and mantle zone B cells (B220+CD21+CD23−) in spleen (Fig. 1 G), or B1 B cells (B220/CD5) in peritoneum (Fig. 1 H). We conclude, therefore, that ubc13 haploinsufficiency does not appear to alter normal immune system development. Furthermore, ubc13 +/− mice had normal lifespans and remained healthy typically for ≥2 years, displaying no tendencies toward increased incidences of infection, autoimmunity, or malignancy compared with wild-type littermates.
Ubc13 Haploinsufficient Mice Are Resistant to LPS Challenge.
TRAF6 plays an essential role in signaling by several TLRs, including the LPS receptor TLR4 (19, 20). Because Ubc13 is required for TRAF6 activation (7), we compared the effects of LPS challenge on age-matched female ubc13 +/+ and ubc13 +/− mice by using a dose of LPS reported to be lethal for normal mice (21). These experiments showed that heterozygous ubc13 +/− mice are significantly more resistant to LPS-induced lethality, with approximately half as many mice dying after challenge with 250 mg/kg LPS compared with wild-type mice (<30% ubc13 +/− vs. >60% ubc13 +/+; P < 0.001 by ANOVA) (Fig. 2). Thus, ubc13 heterozygosity was associated with resistance to the lethal effects of LPS.
Ubc13 +/− mice are resistant to LPS-induced lethality. Female mice (aged 4–6 months) were injected i.p. with 250 mg/kg LPS and then were monitored hourly. The proportion of surviving ubc13 +/− (open ovals) and ubc13 +/+ (filled ovals) mice is indicated (n = 13–14 mice per group). Statistical significance was determined by ANOVA; P < 0.001.
Ubc13+/− Cells Secrete Less Cytokines in Response to LPS.
LPS induces massive cytokine production contributing to its lethal effects in vivo. We therefore assessed the effects of LPS on cytokine production in cultures of primary macrophages differentiated in vitro from bone marrow of ubc13 +/+ and ubc13 +/− mice (Fig. 3 A) and primary splenocytes (Fig. 3 B). LPS induced significantly less TNF, IL-6, and INFγ production in cultures of splenocytes and macrophages derived from heterozygous ubc13 +/− mice compared with homozygous ubc13 +/+ animals. Differences in cytokine production were more striking for macrophages than for splenocytes but statistically significant in both cases.
Ubc13 is required for LPS-induced cytokine production. Production of TNF, IL-6, and IFNγ in cultures of ubc13 +/+ and ubc13 +/− macrophages (A) and splenocytes (B) was analyzed 24 h after stimulation with LPS (10 μg/ml). Data represent mean ± SD of triplicate samples, and are representative of three independent experiments. ∗, P < 0.05; ∗∗, P < 0.005, compared with wild-type mice (Student's t test).
The effects of ubc13 haploinsufficiency on lymphocyte proliferation were also assessed by BrdU labeling of cultured splenocytes stimulated with antigen receptor agonists: anti-IgM for B cells or anti-CD3/anti-CD28 for T cells. The ubc13 +/− lymphocytes displayed a partial defect in proliferation [see supporting information (SI) Fig. 6 and SI Methods]. LPS-stimulated proliferation of ubc13 +/− B cells was also slightly reduced (SI Fig. 6). However, LPS- or anti-IgM-stimulated expression of activation antigen CD69 on B cells and anti-CD3/CD28-stimualted expression of CD25 expression on T cells was not reduced in ubc13-haploinsufficient cells (SI Fig. 7).
Impaired Activation of Signal Transduction Pathways in ubc13+/− Mice Stimulated with LPS.
Because Ubc13 is required for TRAF-dependent activation of a variety of downstream kinases (8, 9), we compared the status of the JNK and p38 MAPK in ubc13 +/− and ubc13 +/+ mice after LPS challenge. For these experiments, mice were injected i.p. with LPS, were killed 6 h later, and their spleens were recovered for preparation of protein lysates, which were analyzed by immunoblotting using phospho-specific antibodies specific for phospho-JNK and phospho-p38 MAPK. Compared with wild-type ubc13 +/+ mice, spleens of haploinsufficient ubc13 +/− mice contained less phospho-JNK and phospho-p38MAPK after LPS challenge (Fig. 4 A). Total levels of JNK and p38 MAPK proteins however were similar in ubc13 +/− and ubc13 +/+ mice, as shown by immunoblots using antibodies that react with phospho-independent epitopes on these proteins. Note that immunoblotting also revealed lower levels Ubc13 protein in the spleens of heterozygous ubc13 +/− mice compared with homozygous, wild-type ubc13 +/+ mice (Fig. 4 A), as expected.
Reduced LPS stimulation of phosphorylation of JNK and p38 MAPK and diminished polyubiquitination of TRAFs in haploinsufficient ubc13 +/− mice. (A) Wild-type and ubc13 +/− mice were injected i.p. with 150 mg/kg LPS and killed 6 h later. Spleen tissue lysates were prepared, normalized for total protein content, and analyzed by SDS/PAGE/immunoblotting by using phospho-specific or pan-specific JNK and p38 MAPK antibodies or anti-Ubc13 antibody. (B) Spleen lysates from 150 mg/kg LPS-treated (+) or untreated (−) ubc13 +/+ or ubc13 +/− mice were subjected to immunoprecipitation by using anti-TRAF6 antibody. Immune complexes (Upper) or lysates (Lower) were analyzed by immunoblotting with anti-ubiquitin (Upper) and anti-TRAF6 (Lower) antibodies. (C) Splenocytes of ubc13 +/+ and ubc13 +/− mice were cultured with (+) or without (−) LPS (20 μg/ml) for 15 min. Lysates were prepared and subjected to immunoprecipitation with anti-TRAF6 antibody. Immune complexes (Upper) or lysates (Lower) were analyzed by immunoblotting with anti-ubiquitin (Upper) and anti-TRAF6 (Lower) antibodies.
In vitro experiments have shown that TRAFs become polyubiquitinated in a Ubc13-dependent manner upon ligation of TNFRs and TLRs (7, 22). To determine whether Ubc13 is required for the polyubiquitination of TRAF6 in vivo, we immunoprecipitated TRAF6 from spleen lysates prepared from LPS-treated ubc13 +/− and ubc13 +/+ mice and then analyzed the immunoprecipitated proteins by immunoblotting with anti-ubiquitin antibody. Indeed, the extent of ubiquitin conjugation to TRAF6 was greater in lysates prepared from wild-type ubc13 +/+ compared with haploinsufficient ubc13 +/− mice (Fig. 4 B). Immunoblot analysis revealed similar total levels of TRAF6 in spleen tissue of these mice. These data thus provide previously unrecognized evidence that Ubc13 is required for polyubiquitination of TRAF6 in vivo. Consistent with the in vivo results, LPS-induced polyubiquitination of TRAF6 in cultured ubc13 +/− splenocytes was also reduced (Fig. 4 C). In contrast to TRAF6, in cultured splenocytes, LPS did not stimulate polyubiquitination of NF-κB essential modulator (NEMO) [IκB kinase γ (IKKγ)] (data not shown), a recognized substrate of TRAF6.
Experiments Using Cultured Cells from ubc13+/− Mice Reveal Selective Impairment in Signal Transduction Initiated by Members of the TNFR and TLR Family.
To further define the signaling defect in immune cells of ubc13 +/− heterozygous mice, we performed experiments using primary splenocytes and primary macrophages, stimulating these cells in vitro with ligands that engage TRAF-dependent TNFRs (e.g., TNF) and TLRs (e.g., LPS). Comparisons were made with the TRAF-independent ligands, namely, H-Ala-d-γ-Glu-diaminopimelic acid (γTriDAP), a component of peptidoglycan that stimulates the Nod-like receptor (NLR)-family protein Nod1 (23), and UV-irradiation, a potent inducer of JNK and p38 MAPK activation (24). Signal transduction events examined in cultured cells included assessment of phosphorylation of JNK and p38 MAPK by using phospho-specific antibodies (as above) and determination of levels of IκBα, a suppressor of p65/p50 NF-κB that becomes degraded in the context of TNFR/TLR signaling as a result of IκB kinase (IKK) activation (25–29).
Macrophages and splenocytes from heterozygous ubc13 +/− mice showed clear reductions in one or both of these signal transduction end points compared with wild-type ubc13 +/+ cells after stimulation with LPS or TNF. Macrophages, for example, displayed a marked decrease in IκBα degradation after LPS treatment, whereas differences in phosphorylation of JNK and p38 MAPK were less robust (Fig. 5 A). In contrast, cultured splenocytes exhibited a clear reduction in phosphorylation of JNK and p38 MAPK, in addition to differences in LPS-inducible IκBα degradation (Fig. 5 B). Consistent with impaired degradation of IκBα, LPS-stimulated splenocytes from ubc13 haploinsufficient mice produced less NF-κB DNA binding activity than ubc13 wild-type animals on the basis of EMSAs (SI Fig. 8). TNF-induced IκBα degradation was reduced in ubc13 +/− macrophages and splenocytes compared with ubc13 +/+ cells (Fig. 5 C and D). In contrast, activation of stress kinase after TNF stimulation was only marginally different in ubc13 +/+ vs. ubc13 +/− macrophages and splenocytes (Fig. 5 C and D). Immunoblot analysis demonstrated comparable reductions in Ubc13 protein levels in macrophages and splenocytes of ubc13 +/− compared with ubc13 +/+ mice.
Impaired signal transduction in ubc13 haploinsufficient cells treated with TRAF-dependent stimuli. (A and B) Macrophages (A) and splenocytes (B) of ubc13 +/+ and ubc13 +/− mice were stimulated with LPS (20 μg/ml) for various times. Degradation of IκBα and phosphorylation of JNK and p38 were detected by immunoblot analysis using antibodies to IκBα, phospho-JNK (p-JNK), and phospho-p38 (p-p38). Total amounts of JNK and p38 are also presented as loading controls. Levels of Ubc13 were determined by using anti-Ubc13 antibody. (C and D) Macrophages (C) and splenocytes (D) of ubc13 +/+ and ubc13 +/− mice were stimulated with TNF (10 ng/ml) for various times. Degradation of IκBα and phosphorylation of JNK and p38 were detected by immunoblot analysis using antibodies to IκBα, p-JNK, and p-p38. Total amounts of JNK and p38 are also presented as loading controls. Levels of Ubc13 were determined by using anti-Ubc13 antibody. (E and F) Splenocytes of ubc13 +/+ and ubc13 +/− mice were stimulated with anti-IgM (10 μg/ml) (E) or anti-CD3 (1 μg/ml)/anti-CD28 (5 μg/ml) (F) for various times. (G and H) Splenocytes of ubc13 +/+ and ubc13 +/− mice were stimulated with γTriDAP (5 μg/ml) (G) and UV irradiation (10 J/m2) (H) for various times. Normalized cell lysates were analyzed by immunoblotting for degradation of IκBα and phosphorylation of JNK and p38 with antibodies to IκBα, p-JNK, and p-p38. Total amounts of JNK and p38 are also presented as loading controls.
Because TRAF6/Ubc13 complexes have been implicated in antigen receptor signaling in lymphocytes (30), we also examined these signaling events in splenocytes stimulated with anti-IgM or anti-CD3/CD28. These agonists of B cell and T cell receptors stimulated less phosphorylation of JNK and p38MAPK in ubc13 +/− splenocytes, whereas IκBα protein levels were only marginally different (Fig. 5 E and F).
In contrast to LPS, TNF, and antigen receptor agonists, stimulation of splenocytes from ubc13 +/− and ubc13 +/+ mice revealed little difference in induction of JNK and p38 phosphorylation by the Nod1 agonist γTriDAP or by UV irradiation (Fig. 5 G and H). The decline in IκBα levels induced by γTriDAP was also not substantially different in ubc13 +/− compared with ubc13 +/+ cells. UV irradiation caused little decline in IκBα levels in either ubc13 +/− or ubc13 +/+ cells. Taken together, these data demonstrate that ubc13 haploinsufficiency selectively blunts signaling in the context of TLR, TNFR, and antigen receptor activation.
The differences in signal transduction events observed in heterozygous ubc13 +/− versus ubc13 +/+ cells were not the consequence of impaired cell survival. In this regard, IκBα, phospho-JNK, and phospho-p38 MAPK levels were measured within 1 h after stimulation, before changes in cell survival. Also, analysis of the percentage of viable B220+ splenocytes by FITC-annexin V binding at 24 h after stimulation with LPS revealed only a modest (<10%) difference in the percentage of surviving cells (SI Fig. 9), distinguishing the blunted signaling events observed from a cell survival difference.
Discussion
Here, we provide evidence that Ubc13 plays a rate-limiting role in vivo in signal transduction mediated by TRAF-dependent receptors of the TNFR and TLR families, critical regulators of innate immunity and inflammation. Using haploinsufficient ubc13 +/− mice, we observed striking decrements in responses to LPS compared with wild-type ubc13 +/+ mice with respect to lethality, polyubiquitination of TRAF6, and activation of protein kinases that operate as downstream mediators of inflammatory signaling. Cells derived from heterozygous ubc13 +/− mice also displayed blunted responses with respect to LPS and/or TNF-induced cytokine secretion, phosphorylation of JNK and p38 MAPK, and IκBα degradation. In contrast, signaling induced by TRAF-independent stimuli was not altered by ubc13 haploinsufficiency.
The heterozygous ubc13 +/− mice used here show a clear reduction in levels of Ubc13 in macrophages and splenocytes compared with their wild-type ubc13 +/+ littermates. The extent of the decrement in Ubc13 protein levels in other types of cells in these haploinsufficient mice remains to be determined. Cell type-specific or developmental stage-specific differences in Ubc13 protein levels could impact the phenotype of animals. In our mice in which retroviral insertional mutagenesis inactivates one ubc13 allele, however, ubc13 haploinsufficiency did not cause apparent aberrations in lymphocyte differentiation or homeostasis. In this regard, although our analysis of ubc13 +/− mice was designed to focus predominantly on signaling events relevant to innate rather than acquired immunity, we also observed impaired activation of stress kinases in splenocytes stimulated in vitro with antigen receptor agonists, as well as a modest proliferation defect, without evidence of impaired IκBα degradation or reduced expression of activation antigens. Further analysis, however, is required to determine whether these blunted responses to antigen receptor agonists are direct consequences of impaired antigen receptor signaling versus an indirect consequence resulting from differences in secondary signaling by cytokine receptors activated as a result of antigen receptor-induced expression of cytokines or cytokine receptors in ubc13 +/− splenocyte cultures.
We found that homozygous disruption of ubc13 resulted in embryonic lethality at a very early stage of development. The explanation for this early lethality is unknown, but it could be related to the alternative role that Ubc13 plays in DNA replication and repair (14). The rapid cell division that characterizes the early embryo may necessitate that Ubc13 is available during development. Although ubc13 +/− splenocytes demonstrated only a modest decline in S-phase entry, the residual levels of Ubc13 protein in these haploinsufficient cells presumably are sufficient to sustain proliferative responses. This notion is also supported by the observation that ubc13 +/− mice contain normal numbers of immune cells, suggesting that any role of Ubc13 in DNA replication and repair is adequately preserved in the haploinsufficient state.
After submission of this paper, independent studies of mice were reported in which ubc13 was ablated by tissue-specific gene knockout in B cells, T cells, and macrophages, permitting production of viable mice with homozygous inactivation of ubc13 in one of these three cell lineages (31, 32). Comparisons of the phenotypes of haploinsufficient (ubc13 +/−) and homozygous-deficient (ubc13 −/−) macrophages and splenocytes reveal general agreement but also some differences that presumably reflect cell type-specific differences in the effects of reduced vs. complete loss of Ubc13 protein (see SI Table 1 and SI Fig. 10). For example, whereas Yamamoto et al. (31) found no impact of homozygous inactivation of ubc13 on TNF-induced IκBα degradation or stress kinase activation in mouse embryonic fibroblasts (MEFs), we observed a significant decrease in ubc13 heterozygous macrophages, thus illustrating a cell type-specific difference. Also, whereas homozygous disruption of ubc13 genes in T cells caused a profound inhibition of NF-κB activation by anti-CD3/CD28 (32), we observed a modest difference in IκBα degradation in ubc13 +/− splenocytes, suggesting a gene dosage-dependent difference. The only discrepant finding concerns IκBα degradation induced by LPS in lymphocytes, which we found to be reduced in ubc13 +/− splenocytes, whereas LPS-induced IκBα degradation was reported to be intact in ubc13 −/− B-cells (31). Further analysis of the impact of Ubc13 on LPS receptor signaling in B cells thus is required, including attention to issues such as the contribution of mouse strain differences. The characterization here of the in vivo response of ubc13 +/− mice to LPS provides previously unrecognized insights into the role of this ubiquitin-conjugating enzyme in innate immune response, complementing the recently reported results for mice containing ubc13 −/− lymphocytes and that focused on acquired immunity (31, 32).
The phenotype of haploinsufficient ubc13 +/− mice reported here suggests that the unusual E2 encoded by this gene may represent an attractive target for drug discovery efforts aimed at producing new therapies for inflammatory or autoimmune diseases. However, given the involvement of TRAFs in receptor-mediated signaling pathways important for innate immunity, the antiinflammatory benefits of Ubc13-targeted therapeutics must be weighed against risk of infection. Use of ubc13 haploinsufficient mice in clinically relevant models of acute and chronic inflammatory and autoimmune diseases may reveal the best scenarios for exploiting Ubc13 as a therapeutic target.
Materials and Methods
Generation of ubc13 Haploinsufficient Mice.
While searching the Omni Bank database for ES cell clones with retrovirus insertions in the ubc13 gene, we found a clone (OST374154) in which the integration site for the retrovirus was located in the first intron of the mouse ubc13 gene. This ES cell clone was used to derive mice on a C57Bl6/129SV background. The position of retrovirus insertion was confirmed by PCR using primers distinguishing the germ-line and mutant genes. The primers used to detect ubc13 germline were as follows: forward 5′-AACTACAGTGGTTATCCATCACAC-3′ and reverse 5′-TACCAAAGATTCTCTTGGCC-3′. Targeted ubc13 was detected by using the same reverse primer described above and the forward primer 5′-GGCGTTACTTAAGCTAGCTTGC-3′. Female mice of 4–6 months old were injected i.p. with LPS (Escherichia coli 026:B6, L8274; Sigma, St. Louis, MO) and monitored hourly to assess time of death.
Flow Cytometry.
Cells were prepared from thymus glands, bone marrow, spleens, and peritoneum of untreated mice and then were stained with FITC-, phycoerythrin (PE)-, cyanine dye-coupled peridinin chlorophyll protein- (PerCP-Cy5.5)-, or allophycocyanin (APC)-conjugated antibodies (BD Biosciences Pharmingen, Franklin Lakes, NJ) as described in ref. 33. Cells were flow-sorted with a FACSCanto (BD Biosciences Pharmingen), and data were analyzed with FlowJo software (Tree Star, Ashland, OR).
Cell Isolation and Culture.
Splenocytes were harvested from spleens of killed 2- to 3-month-old Ubc13 +/− and littermate control mice by using mouse erythrocyte lysing kit (R & D Systems, Minneapolis, MN). The resulting cells were cultured at 37°C with 5% CO2 in RPMI medium 1640 with 0.5% FCS and 50 μM 2-mercaptoethanol before stimulation with LPS (Sigma), anti-IgM (Jackson Immunoresearch Laboratories, West Grove, PA), or anti-CD3 plus anti-CD28 antibodies (BD Biosciences Pharmingen). Bone marrow cells were harvested from hind leg bones of killed 2- to 3-month-old Ubc13 +/− and littermate control ubc13 +/+ mice. Bone marrow cells were cultured at 37°C with 5% CO2 in RPMI medium 1640 with 10% heat-inactivated FCS. CSF-1 (Sigma) was added to the medium for the first week to promote differentiation. Adherent macrophages were stimulated for 24 h with various amounts of LPS.
Immunoblotting.
Cell extracts were prepared with lysis buffer containing 25 mM Tris·HCl (pH 7.4), 2 mM Na3VO4, 10 mM NaF, 10 mM Na2P2O7, 1 mM EGTA, 1 mM EDTA, 1% Nonidet P-40, and protease inhibitor mixture (Roche, Indianapolis, IN). Equal amounts of cell lysates were subjected to immunoblot analysis. Anti-JNK, anti-phospho-JNK (Thr-183/Thr-185), anti-p38, and anti-phospho-p38 (Thr-180/Tyr-182) antibodies were obtained from Cell Signaling (Beverly, MA). Anti-IκBα and anti-TRAF6 were purchased from Santa Cruz Biotechnology, (Santa Cruz, CA). Anti-Ubc13 was obtained from Zymed Laboratories (South San Francisco, CA). Tissue extracts were obtained by using a tissue grinder (Knotes Glass, Millville, NJ) with the lysis buffer and then were centrifuged at 12,000 × g for 5 min. Equal amounts of the lysates were subjected to immunoblot analysis.
TRAF6 Ubiquitination Analysis.
For in vivo studies, mice were injected i.p. with 150 mg/kg LPS and killed 6 h later. Spleen tissue lysates were prepared and normalized for total protein content. Protein extracts (1 mg) were subjected to immunoprecipitation with rabbit polyclonal anti-TRAF6 antibody (Santa Cruz Biotechnology) conjugated to protein A Sepharose 4B. Beads-bound proteins were then eluted in SDS sample buffer and subjected to immunoblotting with anti-ubiquitin antibody (Covance, Richmond, CA). For in vitro experiments, splenocytes were harvested from spleens of killed 2- to 3-month-old Ubc13 +/− and littermate control mice by using mouse erythrocyte lysing kit (R & D Systems). The resulting cells were cultured at 37°C with 5% CO2 in RPMI medium 1640 with 0.5% FCS and 50 μM 2-mercaptoethanol before stimulation with LPS (Sigma). Protein extracts were subjected to immunoprecipitation with rabbit polyclonal anti-NF-κB essential modulator (NEMO) antibody (Santa Cruz Biotechnology).
Cytokine Measurements.
Macrophages and splenocytes were cultured at 5 × 105 cells per milliliter with or without LPS for 24 h. Culture supernatants then were assayed for murine TNFα, IL-6, and INFγ by using ELISA kits from R & D Systems. Data were normalized for cell numbers, and assays were performed in triplicate.
Acknowledgments
We thank M. Hanai and J. Valois for manuscript preparation; S. Abbas for technical assistance; and the National Institutes of Health for generous funding through Grants CA69381, AI070859, and CA078419.
Footnotes
- *To whom correspondence should be addressed. E-mail: reedoffice{at}burnham.org
-
Author contributions: T.F and S.-i.M. contributed equally to this work; S.-i.M., Z.R., and J.C.R. designed research; T.F., S.-i.M., C.L.K., J.M.B., M.K., S.L., and J.M.Z. performed research; T.F., S.-i.M., J.M.Z., Z.R., and J.C.R. analyzed data; and S.-i.M., Z.R., and J.C.R. wrote the paper.
-
The authors declare no conflict of interest.
-
This article is a PNAS direct submission.
-
This article contains supporting information online at www.pnas.org/cgi/content/full/0700548104/DC1.
- Abbreviations:
- γTriDAP,
- H-Ala-d-γ-Glu-diaminopimelic acid;
- LPS,
- lipopolysaccharide;
- TLR,
- Toll-like receptor;
- TNFR,
- TNF receptor;
- TRAF,
- TNFR-associated factor.
- © 2007 by The National Academy of Sciences of the USA