Tle1 tumor suppressor negatively regulates inflammation in vivo and modulates NF-κB inflammatory pathway

Significance

This study provides the first physiological evidence, to our knowledge, that Tle1 (transducin-like enhancer of split 1) is a major negative regulator of inflammation. We show that the loss of Tle1 in mice leads to increased activity of the proinflammatory NF-κB pathway as well as decreased activity of Hes1 (hairy and enhancer of split-1), a negative regulator of inflammation. In addition, Tle1 loss resulted in decreased growth and survival with increased myelopoiesis and lung hypoplasia. Loss of Tle1 also sensitized mice to inflammatory stimuli and facilitated cancer progression. Our study opens the way to further investigations for a role of Tle1 in human inflammatory disease and cancer progression.

Abstract

Tle1 (transducin-like enhancer of split 1) is a corepressor that interacts with a variety of DNA-binding transcription factors and has been implicated in many cellular functions; however, physiological studies are limited. Tle1-deficient (Tle1^Δ/Δ) mice, although grossly normal at birth, exhibit skin defects, lung hypoplasia, severe runting, poor body condition, and early mortality. Tle1^Δ/Δ mice display a chronic inflammatory phenotype with increased expression of inflammatory cytokines and chemokines in the skin, lung, and intestine and increased circulatory IL-6 and G-CSF, along with a hematopoietic shift toward granulocyte macrophage progenitor and myeloid cells. Tle1^Δ/Δ macrophages produce increased inflammatory cytokines in response to Toll-like receptor (TLR) agonists and lipopolysaccharides (LPS), and Tle1^Δ/Δ mice display an enhanced inflammatory response to ear skin 12-O-tetradecanoylphorbol-13-acetate treatment. Loss of Tle1 not only results in increased phosphorylation and activation of proinflammatory NF-κB but also results in decreased Hes1 (hairy and enhancer of split-1), a negative regulator of inflammation in macrophages. Furthermore, Tle1^Δ/Δ mice exhibit accelerated growth of B6-F10 melanoma xenografts. Our work provides the first in vivo evidence, to our knowledge, that TLE1 is a major counterregulator of inflammation with potential roles in a variety of inflammatory diseases and in cancer progression.

Transducin-like enhancer of split 1 (TLE1) belongs to a family of corepressor proteins called transducin-like enhancer of split, or TLEs. Groucho, the TLE homolog in Drosophila, has crucial roles in neurogenesis, segmentation, and sex determination (1). These corepressors do not bind directly to DNA but rather interact with many different classes of transcription factors and help create a repressor complex (1, 2).

Vertebrates express five different TLEs (TLE1–4, AES), although the distinct functions of each of the TLEs has not been well determined. TLE1, the most studied among the Groucho family proteins in mammalian systems, is widely expressed in different tissues and cell types and has been implicated in neuronal differentiation (3, 4) and pancreatic beta cell development (5). TLE1 has tumor suppressor activity (6–8) as well as oncogenic functions in cancer (9, 10). TLE1 associates with many important transcription factors integral for cell proliferation and differentiation, including Runx2 to block rRNA expression (11), HES1 to suppresses MASH2 expression (12), and TCF/LEF proteins to block Wnt target gene activation (13). TLE1 also represses NF-κB activity (14, 15). Involvement in these diverse cellular functions and diseases was studied primarily in vitro or using in vivo overexpression systems. The major physiological function of TLE1, however, remains poorly understood.

A few recent studies suggest TLE1 might regulate immune function. For example, a single noncoding nucleotide polymorphism within the TLE1 locus was associated with inflammatory bowel diseases (16), and in human monocytes, increased TLE1 expression was important for zymosan-mediated inhibition of interleukin 12 (IL-12) p70 expression (17) as well as Bacillus anthracis toxin-induced immune suppression (18). However, the importance of TLE1 for overall immune function has not been characterized.

We previously identified TLE1 and TLE4 as tumor suppressor genes deleted in AML (acute myeloid leukemia) (19) whose loss cooperated with AML1-ETO (acute myeloid leukemia 1-eight twenty one) in leukemia development (6). We recently characterized Tle4 knockout mice and showed that Tle4 has a critical role in maintaining hematopoietic stem cell (HSC) function and in bone development (20). To better understand the physiological roles of Tle1, we developed Tle1 knockout mice. We found the loss of Tle1 leads to an excessive immune activation through a combination of constitutive activation of the NF-κB inflammatory pathway in skin, lung, and intestine as well as decreased Hes1-mediated immune suppression in macrophages. Tle1 deficiency also resulted in lung hypoplasia, decreased overall survival, and enhanced transplanted tumor growth. Our results demonstrate a critical and previously unidentified role of Tle1 in suppressing in vivo inflammation.

Results

Tle1^Δ/Δ Mice Have Growth Retardation and Reduced Fitness.

The creation of Tle1^Δ/Δ mice (Fig. S1) is described in SI Results. Tle1^Δ/Δ mice were born at normal frequency and appeared similar to wild-type (WT) and heterozygous (het) littermates at birth (Fig. 1 A, a). Delayed pigmentation was evident at 3 or 4 d of life (Fig. 1 A, b). Compared with WT and het littermates, Tle1^Δ/Δ mice generally became progressively runted (Fig. 1 A–C and Fig. S2). By day 5, Tle1^Δ/Δ mice on average weighed only 37% of WT and het mice. Knockout mice also exhibited significantly reduced fitness, with 50% of Tle1^Δ/Δ mice dying before day 60, the majority between 8 and 20 d of life (Fig. 1C). Although among the surviving Tle1^Δ/Δ mice there was some catch-up growth after weaning, older mice on average had 23–27% less body weight compared with age-matched het and WT mice (Fig. 1B). A similar effect was seen in females and males (Fig. 1B). In addition, Tle1^Δ/Δ mice have shorter intestinal length, fewer hair follicles with disorganized skin epidermis basal cell layer, and pulmonary lung hypoplasia (SI Results and Figs. S3 and S4).

Fig. S1.

Generation of Tle1^Δ/Δ mice. (A) Genomic structure of WT Tle1 locus, targeted allele with loxP sites flanking exon 2, and post–B-actin-cre Tle1 locus with exon 2 deletion. (B) Southern blot analysis of BamH1-digested genomic DNA with 5′ and 3′ Tle1 probes. (C) Sequenced RT-PCR product shows the loss of exon 2 and frame shift with resultant nonsense mutation in Tle1 exon 3. (D) Western blot experiment confirms the absence of Tle1 protein expression in the brain of Tle1^Δ/Δ mice.

Fig. 1.

Loss of Tle1 reduces growth and survivability. (A) Tle1^Δ/Δ mice appear similar to WT and het at birth (a) but by 4 d (b) have delayed pigmentation and at 19 d (c) are runted with poor body condition. (B) Growth of Tle1^Δ/Δ mice was significantly impaired relative to het and WT. A similar trend was observed in both male and female Tle1^Δ/Δ mice; n = 3–6; *P < 0.05, **P < 0.01; all error bars indicate mean ± SEM. (C) Kaplan–Meier curve for Tle1^Δ/Δ and WT mice. Fifty percent of mice died within 60 d, most between 8 and 20 d (n = 17–20).

Fig. S2.

Phenotype of severely affected Tle1^Δ/Δ mice at different days after birth.

Fig. S3.

Tle1^Δ/Δ mice have developmental abnormalities in intestine, skin, and lung. (A) Newborn Tle1^Δ/Δ mice show shortened intestine with delayed passage of meconium (M). (B) Hematoxylin and eosin staining of 11-d-old WT and Tle1^Δ/Δ mouse skin. (Scale bar, 100 μm.) Arrow points to hair follicles. (C) Hart’s elastin staining of 6-d-old WT and Tle1^Δ/Δ lung showed hypoplasia with poor septation. (Scale bar, 100 μm.)

Fig. S4.

Tle1^Δ/Δ mice have fewer proliferating cells in the lung, reduced lung septation, and macrophage infiltration in the skin. (A) Hart’s elastin staining of 23-d-old Tle1^Δ/Δ lung shows hypoplasia with poor septation. (B) K_i 67 staining of 6-d-old lung. (C) Immunostaining of 11-d-old mice skin with pan macrophage marker CD68. (Scale bar, 100 μm.) Arrow points to CD68-positive macrophage.

Tle1 Deficiency Skews Hematopoiesis Toward the Myeloid Lineage by Hematopoietic Cell Extrinsic Stimuli.

Complete blood counts (CBCs) showed a higher neutrophil count in the Tle1^Δ/Δ mice (Fig. 2A). We saw a trend toward a decrease in lymphocytes that was not significant, although as demonstrated below, we did find a significant decrease in B cells. The cellularity of the bone marrow (BM) was comparable between WT and Tle1^Δ/Δ mice (Fig. S5A); however, Tle1-deficient BM showed a trend toward a higher percentage of granulocyte macrophage progenitors at 2 wk (Tle1^Δ/Δ, 0.296; Het, 0.179; P = 0.2819) and 4 wk (Tle1^Δ/Δ, 0.344; Het, 0.175; P = 0.1183) and a significant increase at 12 wk (Tle1^Δ/Δ, 0.238; Het, 0.173; P = 0.034) (Fig. 2B). By contrast, the frequencies of HSCs (Fig. S5B), common myeloid progenitors (CMPs) (Fig. S5C), megakaryocyte-erythroid progenitors (MEPs) (Fig. S5D), common lymphoid progenitors (CLPs) (Fig. S5E), or CD3e-positive T-cell populations (Fig. S5F) were not significantly altered. Tle1 knockout mice also displayed increased myeloid (Mac1⁺Gr1⁺) and macrophage (Gr1⁻Mac1⁺F4/80⁺) and decreased B-cell populations (B220) in their BM, spleen, and peripheral blood (PB) (Fig. 2 C–E). BM methylcellulose assays also revealed a higher number of granulocyte macrophage and macrophage colonies (Fig. 2F). These data suggest that loss of Tle1 results in myeloid expansion. To determine whether this hematopoietic phenotype is cell-intrinsic or -extrinsic, one million CD45.2 BM cells from Tle1^Δ/Δ or WT mice were transplanted into lethally irradiated CD45.1 recipient mice. Tle1^Δ/Δ-derived BM cells reconstituted all of the lineages similar to WT, indicating the myeloid expansion observed in the Tle1 knockout mice was not cell-autonomous (Fig. S6 A–C).

Fig. 2.

Hematopoiesis in Tle1^Δ/Δ mice is skewed toward the myeloid lineage with increased serum IL-6 and G-CSF. (A) PB counts for the indicated population at day 23 (n = 5–6). (B) Percentage of BM GMP-Lin⁻Sca-1⁻c-Kit⁺CD34⁺FcγR^hi granulocyte macrophage progenitor cells. (C) Percentage of Gr1- and Mac1-positive myeloid cells. (D) Percentage of B220-positive B cells. (E) Percentage of Gr1⁻Mac1⁺F4/80⁺-positive macrophage cells. (F) CFU-C assay for myeloid progenitors. (G) Serum IL-6 level. (H) Serum G-CSF level. n = 3–6 mice per genotype in A–H; *P < 0.05, **P < 0.01; all error bars indicate mean ± SEM.

Fig. S5.

Loss of Tle1 affects only the myeloid and B-cell hematopoietic cell populations. (A) Hematoxylin and eosin staining of BM reveals normal cellularity. (Scale bar, 100 μm.) (B) Percentage of Lin⁻c-Kit⁺Sca⁺CD150⁺CD48⁻ HSCs. (C) Percentage of Lin⁻c-Kit⁺Sca-1⁻CD34⁺FcγR^lo CMPs. (D) Percentage of Lin⁻c-Kit⁺Sca-1⁻CD34⁻FcγR^lo MEPs. (E) Percentage of Lin⁻ IL-7R⁺Sca-1^loc-Kit^lo CLPs. (F) Percentage of CD3e⁺ T cells. Error bars indicate mean ± SEM.

Fig. S6.

The myeloid cell predominance seen in Tle1^Δ/Δ mice is due to influences extrinsic to the hematopoietic cell compartment. Tle1^Δ/Δ mice and WT mice BM was transplanted to CD45.1 syngeneic mice and analyzed after 16 wk. (A) Percentage of BM chimeras (n = 6). (B) Percentage of indicated mononuclear cells in BM. (C) Percentage of indicated mononuclear cells in blood. Error bars indicate mean ± SEM.

Tle1-Deficient Mice Have Increased Expression of Inflammatory Cytokines and Chemokines in the Intestine, Skin, and Lung.

Because myeloid cells and macrophages are key regulators of inflammation, we next asked whether the myeloid expansion observed in Tle1^Δ/Δ mice is associated with increased inflammation. We analyzed mRNA expression of inflammatory cytokines IL-6 (interleukin-6), IL-1β, and TNF-α (tumor necrosis factor-α) and chemokines M-CSF (macrophage colony stimulating factor), GM-CSF (granulocyte macrophage-CSF), and G-CSF (granulocyte-CSF) at 2, 8, and 21 d after birth of the severely affected animals from the lung, liver, small intestine, skin, and BM. At day 2, the small intestine of Tle1^Δ/Δ mice had significantly higher IL-1β, IL-6, GM-CSF, and G-CSF expression compared with WT (Fig. 3A). IL-1β, IL-6, and G-CSF expression was also significantly higher in the skin, and TNF-α levels were higher in the small intestine and BM of 8-d-old Tle1^Δ/Δ mice (Fig. 3B). At day 21, IL-6, IL-1β, TNF-α, and chemokines M-CSF, GM-CSF, and G-CSF were significantly increased in the lungs, and IL-6 was also increased in the skin and small intestine (Fig. 3C). The analysis of serum from 21-d-old mice revealed a corresponding increase of the inflammatory cytokine IL-6 (Fig. 2G) and the chemokine G-CSF (Fig. 2H) in Tle1^Δ/Δ mice compared with WT mice. We also observed increased expression of IL-1β in 1.5-y-old Tle1-deficient mice (Fig. S7). Thus, Tle1 deficiency leads to increased local and systemic production of inflammatory cytokines and chemokines.

Fig. 3.

Deficiency in Tle1 results in systemic inflammation with increased inflammatory cytokines and chemokine expression. Expression of IL-1β, IL-6, TNF-α, M-CSF, GM-CSF, and G-CSF was analyzed from liver, lung, skin, small intestine, and BM in WT (open bar) and Tle1^Δ/Δ mice (closed bar). All error bars are expressed as mean + SEM (n = 3); *P < 0.05, **P < 0.01. (A) Two-day-old mice. (B) Eight-day-old mice. (C) Mice 21 d old.

Fig. S7.

Older Tle1^Δ/Δ mice express increased inflammatory cytokines in the lung. Expression of IL-1β in 1.5-y-old mice lung is shown. n = 3; **P < 0.01; error bars indicate mean ± SEM.

Tle1 Modulates NF-κB–Mediated Inflammation.

Because Tle1 overexpression has been shown to block NF-κB activation (14, 15), we further studied the effect of TLE1 on NF-κB nuclear translocation and phosphorylation (21). Overexpression of TLE1 inhibited both basal as well as LPS-induced NF-κB activation in THP-1 human monocyte cells, as evidenced by a decrease in nuclear NF-κB translocation using ImageStream analysis. The mean difference between nuclear and cytoplasmic (mean nuc–cyto diff) NF-κB was –24 ± 0.8 in unstimulated cells with empty vector, whereas it was –28 ± 0.9 in TLE1-expressing cells. After LPS stimulation, the mean nuc–cyto diff in cells with empty vector increased to 13 ± 1.2, whereas it is –7 ± 0.8 in the TLE1-transduced cells, indicating a greater proportion of NF-κB remains in the cytoplasm (Fig. 4A). Overexpression of Tle1 decreased LPS-mediated NF-κB phosphorylation in the Raw264.7 mouse macrophage cell line (Fig. 4B, Bottom). We found increased expression of phosphorylated NF-κB (serine 536) in the skin, lung, and intestine of 10-d-old Tle1^Δ/Δ mice compared with WT mice (Fig. 4B, Top) and in LPS-treated Tle1^Δ/Δ BMDMs compared with WT BMDMs (Fig. 4B, Middle). We also found an increase in many NF-κB pathway genes involved in inflammation and antiapoptosis in the lungs of 21-d-old Tle1^Δ/Δ mice (Fig. 4C). Most significantly, angiotensinogen (Agt) was expressed sixfold higher in Tle1^Δ/Δ mice compared with WT mice. We found overexpression of the neutrophil chemoattractants CXCL1 and CXCL3 as well as the inflammatory cytokines IL-1β and myeloid growth factor CSF2. We also observed increased expression of inflammatory genes C-Rel, Snap25, and Tnfrsf1b. These data are consistent with Tle1 modulation, NF-κB nuclear translocation, phosphorylation, and increased expression of NF-κB inflammatory target genes, which in turn promotes inflammation in Tle1^Δ/Δ mice. Next, we asked whether suppression of NF-κB activation using a bortezomib proteasome inhibitor can ameliorate inflammation in Tle1^Δ/Δ mice and decrease the expansion of myeloid cells. Repeated injection of bortezomib twice a week for 5 wk significantly reduced the percentage of Gr1⁺mac1⁺ myeloid cells in the spleen of Tle1^Δ/Δ mice (2.54) compared with PBS vehicle (7.38)-treated mice (Fig. 4D). These results demonstrate that Tle1 modulates NF-κB activation and expression of inflammatory target genes and provide evidence that at least part of this inflammatory phenotype in Tle1^Δ/Δ mice can be ameliorated by NF-κB inhibition in vivo.

Fig. 4.

TLE1 modulates NF-κB–mediated inflammation. (A) Inhibition of nuclear translocation of NF-κB. Empty vector or TLE1-transduced THP-1 cells were exposed to LPS for 1 h. The level of NF-κB protein in the nucleus and cytoplasm was visualized (Left) and measured by ImageStream before or after LPS exposure. The mean nuc–cyto diff is shown (Right). A negative value indicates a relatively greater cytoplasmic concentration. (B) Western blots showing the level of phosphorylated NF-κB (S536) in 10-d-old Tle1^Δ/Δ and WT mice (Top) and in Tle1^Δ/Δ and WT BMDMs after exposure to 100 ng/mL LPS for the indicated time (Middle) and RAW264.7 macrophage cell line transfected with vector or TLE1 after exposure to 10 ng/mL LPS for the indicated time (Bottom). Data are representative of three independent experiments. (C) Expression of NF-κB target genes in the lungs of Tle1^Δ/Δ 21-d-old mice compared with WT mice measured using PCR array (n = 3). (D) Percentage of Gr1⁺Mac1⁺ myeloid cells in the Tle1^Δ/Δ and WT mice after 5 wk of intraperitoneal injection of PBS or bortezomib 0.75 mg/kg (n = 4). *P < 0.05, **P < 0.01.

Tle1 Loss Enhances Inflammatory Response to TPA and Accelerated B6-F10 Tumor Growth.

To confirm that Tle1 regulates inflammation, we examined models involving inflammation. First, we analyzed the sensitivity of Tle1^Δ/Δ and WT mouse ears to 12-O-tetradecanoylphorbol-13-acetate (TPA), an inflammation-inducing compound. Repeated treatment with TPA produced only mild swelling and minimal redness in the ears of WT mice. In contrast, an enhanced inflammatory response was seen in Tle1^Δ/Δ mice with severe redness, scaling (Fig. S8A), and swelling (Fig. 5A). The histology of untreated ears of Tle1^Δ/Δ mice was similar to WT. The ears of TPA-treated Tle1^Δ/Δ mice had a thicker epidermal layer, with increased hematopoietic cell infiltration (Fig. 5B). Analysis of expression of inflammatory cytokine and chemokine mRNA levels using a PCR array revealed that TPA-treated Tle1^Δ/Δ ears produced higher inflammatory cytokines and chemokines compared with treated WT ears (Fig. 5C). Thus, the loss of Tle1 enhanced an inflammatory response to TPA through the increased production of inflammatory cytokines and chemokines and attraction of inflammatory cells.

Fig. S8.

Tle1-deficient mice exhibit an enhanced inflammatory response to TPA treatment and accelerated B6-F10 tumor growth. (A) Application of TPA to ears resulted in moderate redness in WT compared with very severe scaling and swelling with increased ear thickness in Tle1^Δ/Δ mice. (B) B16-F10 cells were injected intradermally, and representative pictures are of excised tumors after 17 d.

Fig. 5.

Tle1-deficient mice exhibit an enhanced inflammatory response to TPA treatment and accelerated B6-F10 tumor growth. (A) Ear thickness in the Tle1^Δ/Δ and WT mice after repeated application of TPA on the indicated day for 11 d. n = 6; *P < 0.05, **P < 0.01; error bars indicate mean ± SEM. (B) Ears treated with TPA for 11 d and stained with hematoxylin and eosin had marked thickening of the dermis and epidermal layers with dense immune cell infiltrates in Tle1^Δ/Δ mice compared with WT mice. (Scale bar, 100 μm.) (C) Expression of various cytokine and chemokine mRNAs in the Tle1^Δ/Δ mice ear compared with WT mice ear after application of TPA for 11 d (n = 3–4). (D) B16-F10 cells were injected intradermally, and the tumor volume (mm3) was measured (n = 4–6). *P < 0.05, **P < 0.01; error bars indicate mean ± SEM.

As a second model, we used an orthotopic melanoma cancer model. Because cancer growth is regulated by immune and inflammatory components (22), we asked whether absence of Tle1 affects tumor progression. We compared B6-F10 melanoma growth in Tle1^Δ/Δ mice and WT mice using an orthotopic transplant model. Indeed, as shown in Fig. 5D and Fig. S8B, loss of Tle1 accelerated primary tumor growth.

Altogether our data support that the loss of Tle1 enhances the inflammatory response to inflammatory stimuli and accelerates tumor growth.

Tle1-Deficient Hematopoietic Cells and Macrophages Produce Enhanced Inflammatory Response to Inflammatory Stimuli.

Although we did not observe increased myelopoiesis when Tle1^Δ/Δ BM was transplanted into WT mice, we performed further experiments to determine if there are intrinsic defects in the regulation of inflammatory pathways in Tle1^Δ/Δ hematopoietic cells. For these experiments, we examined the response of BM-transplanted mice, as well as isolated macrophages, to LPS and Toll-like receptor (TLR) agonist inflammatory stimuli.

Lethally irradiated WT mice were transplanted with BM from WT or Tle1^Δ/Δ mice. After 16 wk, injection of LPS resulted in increased septic shock-induced death in Tle1^Δ/Δ BM-transplanted mice (Fig. 6A). Given the critical role of macrophages as mediators of inflammation, we also isolated thioglycolate-induced peritoneal macrophages from Tle1^Δ/Δ mice. These macrophages were stimulated with TLR agonists, and inflammatory cytokine expression was quantified. We found significantly increased expression of cytokines in response to stimulation for Tle1^Δ/Δ macrophages compared with WT (Fig. 6B). Specifically, the expression of IL-6 and TNF-α in response to PAM3, the expression of IL-6 in response to Flagellin, and the expression of TNF-α in response to heat-killed Listeria Monocytogenes were greater in Tle1^Δ/Δ macrophages. We confirmed this increased macrophage responsiveness using bone marrow-derived macrophages (BMDMs). Stimulating BMDMs with LPS resulted in increased expression of inflammatory cytokines IL-6 and IL-1β (Fig. 6C). As shown in Fig. 6 D and E, the expression of Hes1 is decreased in LPS-treated and untreated Tle1^Δ/Δ BMDMs compared with WT. These data show that the loss of Tle1 with resultant decreased expression is associated with increased expression of inflammatory cytokines. Significantly higher Hes1mRNA in Tle1-deficient macrophage compared with WT BMDM (Fig. S9) indicates that decreased Hes1 protein expression is not due to decreased mRNA.

Fig. 6.

Tle1 deficiency in the hematopoietic cells increases inflammatory response to LPS and TLR agonist. (A) Kaplan–Meier survival plots of syngenic lethally irradiated WT mice transplanted with Tle1^Δ/Δ or WT BM for 16 wk and challenged with LPS (n = 10–13). (B) Thioglycolate-induced peritoneal macrophages were cultured and stimulated with increasing concentrations of purified TLR agonist for 24 h. Supernatants were assayed for expression of TNF-α and IL-6 using ELISA. Data (mean ± SEM) from four replicates from two independent experiments are presented after normalizing for cell numbers in each condition. (C) BMDMs were generated by culturing the BM cells with mouse M-CSF for 7 d. BMDMs from either WT or Tle1^Δ/Δ mice were stimulated with LPS for the indicated time, and the expression of IL-1β and IL-6 mRNA was quantified using quantitative RT-PCR on triplicate samples from three independent experiments; P < 0.05, **P < 0.01; error bars indicate mean ± SEM. (D) Western blot analysis of Hes1 expression in the BMDMs. Data are representative of two independent experiments. (E) Level of Hes1 protein in the Tle1^Δ/Δ and WT BMDMs after exposure to 100 nm LPS for the indicated time. Data are representative of two independent experiments.

Fig. S9.

Tle1 deficiency in BMDMs results in increased Hes1 mRNA expression. BMDMs from either WT or Tle1^Δ/Δ mice were analyzed for expression of Hes1 mRNA using quantitative RT-PCR on triplicate samples from three independent experiments; *P < 0.05, **P < 0.01; error bars indicate means ± SEM.

SI Results

Creation of Tle1-Null (Tle1^Δ/Δ) Mice.

Conditional Tle1^Δ/Δ mice were created containing LoxP sites flanking exon 2, and exon 2 was constitutionally deleted by crossing these mice with β-actin:Cre mice (Fig. S1A). Resultant mice were backcrossed to C57BL/6 for over six generations and inbred to generate Tle1^Δ/Δ mice. Excision of exon 2 removes a leucine zipper motif critical for the function of the Q-oligomerization domain (47) and results in a frame shift leading to a predicted truncated protein of nine amino acids. Correct targeting and exon 2 deletion were verified by Southern blot (Fig. S1B). Successful recombination and deletion of exon 2 were confirmed by cDNA sequencing (Fig. S1C). Western blot of Tle1 immunoprecipitation from the brain tissue of Tle1^Δ/Δ mice confirmed the absence of Tle1 protein expression (Fig. S1D).

Loss of Tle1 Leads to Intestine, Skin, and Lung Defects.

In search of a cause for the impaired development and early death of Tle1^Δ/Δ mice, we analyzed the gross pathology of all organs. Despite similar overall size, many newborn Tle1^Δ/Δ mice had significantly shorter intestinal length with apparent slower motility, as evidenced by delayed passage of meconium (Fig. S3A). The skin of severely affected Tle1^Δ/Δ younger mice between the ages of 8 and 20 d had fewer hair follicles and disorganization of the basal cell layer of the epidermis along with macrophage infiltration and thickening of the epidermal wall (Figs. S3B and S4C). Staining of the lungs with Hart’s elastin stain revealed pulmonary hypoplasia with decreased secondary septation in Tle1^Δ/Δ mice at 6 d and 20 d after birth compared with controls (Figs. S3C and S4A). K_i 67 staining showed fewer proliferating cells in the lungs of Tle1^Δ/Δ mice (Fig. S4B). No major anatomical defects were observed in the other organs studied.

Discussion

Inflammation is exquisitely balanced by immune activators and suppressors. Tle1 knockout mice exhibited both increased chronic inflammation as well as heightened susceptibility to inflammatory stimuli, as evidenced by enhanced cytokine expression by stimulated macrophages and enhanced inflammatory response of Tle1^Δ/Δ mice ears to TPA treatment. Our data indicate that loss of Tle1 results in excessive activation of NF-κB–mediated inflammatory pathways.

TLE1 has previously been shown to negatively regulate NF-κB using several in vitro systems (14–16, 23), although this has not been previously examined using animal models. NF-κB is fundamental in controlling many cellular processes, including inflammation and immune response, cell proliferation, apoptosis, and development (24–26). Our results indicate that loss of Tle1 results in increased NF-κB phosphorylation with increased expression of inflammatory cytokines and chemokines in the skin, lung, and intestine and increased expression of NF-κB inflammatory target genes in the lung. This is associated with increased committed myeloid progenitors and myeloid cells and decreased B cells in the BM along with increased circulating neutrophils, IL-6, and G-CSF in Tle1-deficient mice. It is similar to the effects seen with dysregulation of inflammatory pathways and infections (27) and the increase in G-CSF and granulopoiesis seen with constitutive activation of NF-κB in IκBa-deficient mice (28). In addition, in vitro overexpression of Tle1 in monocyte/macrophage cell lines blocks LPS-induced NF-κB nuclear translocation and phosphorylation. We also found bortezomib-mediated suppression of NF-κB activation in Tle1^Δ/Δ mice decreased myeloid cell expansion. This finding provides evidence for the physiological role of Tle1 in negatively regulating the NF-κB inflammatory pathway and in regulating the immune response. Given that the TLEs are capable of interacting with other signaling pathways, we cannot exclude the possibility that pathways other than NF-κB influence the severe inflammatory phenotype in Tle1^Δ/Δ mice. However, we did not find evidence of activation of other pathways such as Wnt, TGF-β, and IFN gamma in lung and TPA-treated ears using NF-κB and inflammatory/cytokine expression arrays. In addition, we did not see increased β-catenin protein, a marker of Wnt activation (Fig. S10).

Fig. S10.

β-catenin expression in Tle1^Δ/Δ mice. Western blots show the level of β-catenin expression in 10-d-old Tle1^Δ/Δ and WT mice. Data are representative of three independent experiments.

Although WT recipients of Tle1^Δ/Δ BM did not exhibit increased myelopoiesis or decreased survival or body weight as seen in Tle1-null mice, they do demonstrate increased susceptibility to LPS-induced septic shock. Isolated Tle1^Δ/Δ macrophages from Tle1^Δ/Δ mice expressed higher inflammatory cytokines in response to LPS and TLR agonists. Our data indicate that along with dysregulation of inflammatory pathways intrinsic to Tle1^Δ/Δ hematopoietic cells, the inflammatory signal from nonhematopoietic cells contributes to the severe inflammatory phenotype of Tle1 knockout mice. Indeed, Tle1 is known to be expressed in endothelial cells and epithelial cells (29), which play a major role in the innate immune response (30, 31).

TLR-mediated activation of the Notch pathway and its target genes Hes1 and Hey1 negatively regulates TLR-mediated inflammatory cytokine production in macrophages (32, 33), providing an important counterregulatory feedback to prevent excessive inflammation. Notch activation also up-regulates TLEs (1, 34, 35). Tle1 forms a very stable complex with Hes1 protein, and it is generally through the association of TLEs that HES1 gains its repressive function (1, 34, 35). TLE–Hes1 interaction has been shown to be important for immune suppression in human monocytes (18) and has been suggested to play a critical role in limiting inflammation (18, 33, 36), although in vivo evidence for this was lacking. Decreased Hes1 protein in the Tle1-deficient macrophage suggests defective immune suppression may be contributing to the increased inflammatory response to the TLR agonist and LPS in these cells. The lower Hes1 protein levels in Tle1-deficient macrophages compared with WT BMDMs, despite higher Hes1 mRNA levels (Fig. S8), suggest a decrease in Hes1 protein stability. Although we did not examine the mechanism of this observed decrease in Hes1 protein levels, Hes1 protein is known to be stabilized by phosphorylation (37) or targeted for degradation by interaction with Hes6 (38). Our results suggest a combination of increased NF-κB activation with decreased Hes1-mediated immune suppression in Tle1^Δ/Δ macrophages might be responsible for enhanced sensitivity of Tle1^Δ/Δ BM to LPS-induced inflammation.

Along with inflammation, we observed a number of apparent developmental abnormalities in severely affected Tle1^Δ/Δ mice, including delayed skin pigmentation, fewer hair follicles with disorganization of the epidermis, poor lung septation, and shortening of the intestine and dysmotility, in addition to growth retardation and early mortality. Although these effects may reflect an important role of Tle1 in the development of these organs, the inflammatory state caused by constitutive activation of NF-κB and expression of inflammatory cytokines and chemokines could be responsible for many of these findings. This concept is supported by prior work from several groups that found inflammation in mice can lead to delayed pigmentation (39), alveolar hypoplasia, and poor postnatal growth and decreased survival (28, 40).

The accelerated B6-F10 melanoma xenograft growth in Tle1-null mice could be due to increased inflammation. Chronic inflammation has been associated with cancer (41). Tle1^Δ/Δ mice had elevated serum IL-6 and G-CSF with higher circulating myeloid cells. It has been shown that neutrophils exposed to a combination of G-CSF and IL-6 converted BM neutrophils from a tumor-suppressive to a tumor-promoting phenotype (42). NF-κB–mediated increases in myeloid-derived suppressor cells were shown to have a major role in tumor angiogenesis and immune modulation (43). This could underlie the advantage for the tumor cell growth we see in Tle1^Δ/Δ mice.

The inflammatory phenotype in these Tle1^Δ/Δ mice may have relevance to a variety of human inflammatory states. Increased NF-κB activation and its target genes Agt and IL-1β are observed in lungs with pulmonary fibrosis (44, 45). A genome-wide association study previously reported that an SNP variation in TLE1 was associated with colitis, and a case was made that this might be due to TLE1 interacting with NOD2 and inhibiting NF-κB activation (16). In support of a possible regulatory role of TLE1 in inflammatory bowel disease, we observed an increase in IL-1β, IL-6, and TNF-α in the intestine of Tle1^Δ/Δ mice.

In summary, our work provides evidence that Tle1 negatively regulates the NF-κB inflammatory pathway and together with Hes1 provides an important role in suppressing excessive inflammation. Through the characterization of Tle1 knockout mice, we provide the first in vivo evidence, to our knowledge, that TLE1 is a major counterregulator of inflammation and could play a role in a variety of inflammatory diseases and potentially cancer growth.

Materials and Methods

Materials and procedures for all experiments are described in SI Materials and Methods. Included are the generation of Tle1^Δ/Δ mice, flow cytometry analysis, PCR array analysis, bortezomib treatment, LPS-induced septic shock, TPA treatment, generation and culturing of macrophage, and statistical analysis. Also provided are additional results and figures. This research was approved by the Massachusetts General Hospital Institutional Animal Care and Use Committee.

SI Materials and Methods

Generation of Tle1^Δ/Δ Mice.

A conditional Tle1 knockout mouse was generated by targeting LoxP sites to flank exon 2 via homologous recombination using the 129S6/SvEvTac ES cell line (Fig. 1A). Resultant mice were crossed with β-actin:Cre mice (a gift from Gail Martin, University of California, San Francisco) to delete exon 2 in all tissues. Het mice were backcrossed to C57BL/6 background for over six generations and then interbred to generate Tle1 knockout (Tle1^Δ/Δ) mice.

Genotyping.

The Tle1 targeting constructs, ES clones, and mice lines were verified by Southern blotting and PCR genotyping. The Tle1^Δ/Δ mice and WT mice were genotyped with primers GGGTTCAATCAATGGCTCG, CCCCTCCCACTGGATGTCTA, and ACCCAGAACAGAGGTGAACG.

Flow Cytometry Analysis.

The mononuclear cells from BM, spleen, or blood were stained with the appropriate antibody in 1:100 dilutions for 30 min on ice and analyzed using a LSRII flow cytometer. Data were analyzed using FlowJo software (Treestar). For lineage detection, BM cells were first stained with a mixture of biotinylated anti-mouse antibodies to Mac1, Gr1, Ter119, CD3, CD4, and B220 followed by streptavidin-conjugated APC-Cy7. Cells were subsequently stained with SLAM HSCs and progenitor markers Sca PE, c-Kit APC, CD150 PE-Cy7, CD48 Pacific Blue, IL-7 PE-Cy7, CD16/32-PE, and CD34 FITC. For detection of individual lineages, Gr1 PE, Mac1 APC, CD3e FITC, and B220 PE-Cy7 antibodies were used. Spleen and BM cells were first incubated with Fc block (1:100 Fc block in 0.5 μm EDTA and PBS) and then stained with Gr1 APC-Cy7, Mac1 APC, and F4/80 FITC for detection of macrophages. For congenic strain discrimination, we used anti-CD45.1 PE-Cy7 and anti-CD45.2 Alexa Flour-700 antibodies.

PCR Array Analysis.

The cytokine and chemokine expression profiles of TPA-treated mice ears were analyzed using mouse cytokine and chemokine signaling pathway RT² profiler PCR array (PAMM150Z, SABiosciences) containing 84 inflammatory pathway target genes. The expression profile of 84 key NF-κB target genes was analyzed using cDNA from the lungs of 23-d-old mice using mouse NF-κB signaling targets RT² Profiler PCR array (PAMM225Z, SABiosciences). The genes in these arrays are listed in SI Materials and Methods. PCR was performed with BioRad MyIQ PCR machine. For data analysis, the ΔΔ^Ct method was used; a positive value indicates gene up-regulation, and a negative value indicates gene down-regulation.

Bortezomib Treatment.

Tle1^Δ/Δ and WT mice were injected with 0.75 mg/kg bortezomib (Velcade:Santa Cruz: SC_217785) twice a week for 5 wk. Controls and untreated mice were injected intraperitoneally with the same volume of PBS for the same period.

TPA Treatment.

Tle1^Δ/Δ and WT mice were treated with 20 µL of 0.01% TPA (Sigma) dissolved in 2% (vol/vol) DMSO and 98% (vol/vol) acetone on the left side of the ear on days 0, 1, 3, 5, 7, and 9. Ear thickness was measured using Traceable Digital Calipers (Fisher Scientific) near the top of the ear distal to the cartilaginous ridges. The day 0 data were used as the baseline to subtract from day 1, 3, 5, 7, 9, and 11 data.

LPS-Induced Septic Shock Model.

We injected 1 × 10⁶ BM cells from Tle1^Δ/Δ and WT BM tail vein into lethally irradiated WT recipient mice. After 16 wk of transplantation, BM chimera was challenged with 25 mg/kg LPS (L2630, Sigma) in PBS.

Ex Vivo Cytokine Assay.

Thioglycollate-elicited peritoneal macrophages were prepared by i.p. injection of 1 mL of 4% (wt/vol) thioglycollate broth. Four days after injection, cells were harvested by peritoneal lavage with ice-cold PBS, and 10⁵ macrophages were plated on each well of 96-well tissue culture plastic dishes in RPMI media supplemented with 10% (vol/vol) heat-inactivated FCS, 500 IU/mL of penicillin–streptomycin, and 2 mM glutamine. After 6 h, plates were washed three times to remove nonadherent cells and stimulated overnight with indicated concentrations of TLR agonists. Tissue culture supernatants were collected and stored at –80 °C for cytokine analysis.

Levels of IL-6 and TNF-α (BD Biosciences) were determined by ELISA according to the manufacturer’s instructions.

Generation of BMDMs.

BM was flushed from the tibia and femur, and red blood cells were lysed with ACK (ammonium–chloride–potassium) lysis buffer (Bio Whittaker). BM cells were cultured in DMEM/F12 media with 10% FCS, 1% penicillin–streptomycin (10,000 U/mL; Invitrogen), and 20 ng/mL M-CSF (Peprotech) for 7 d in corning nontreated sterile plates.

In Vitro LPS Stimulation.

Tle1^Δ/Δ and WT BMDMs (1 × 10⁶ cells per 60 mm tissue culture-treated plates) were cultured overnight and stimulated with 100 ng/mL LPS (L2630, Sigma). After LPS treatment, media was removed and the cells were used for RNA extraction and protein analysis. Raw 264.7 mouse macrophage vector and TLE1 transfected cells (1 × 10⁶ cells per 60 mm tissue culture-treated plates) were cultured for 48 h with 2 μg/mL puromycin and treated with 10 ng/mL of LPS. The stimulated cells were lysed with Triton-X lysis buffer and used for Western blotting.

Plasmid Construction and Cell Line Generation.

Full-length human Tle1 (gift from Stefano Stifani, Montreal Neurological Institute, Montreal) cDNAs were cloned into pEntry vector using Gateway Technology (Invitrogen) and transferred to C-terminal HA-Flag destination vector following the manufacturer's instructions. Raw264.7 mouse macrophage cells were transfected with MSCV-HA-Flag and MSCV-Tle1 HA-Flag vector using lipofectamine 3000 and experiment was performed using cells selected with 4 μg/mL puromycin. Tle1 cDNAs were also cloned into MSCV-IRES-GFP retroviral vector. Virus-containing supernatants were collected 48 and 72 h after the transduction. THP-1 cells (1 × 10⁶) were transduced twice in a 24-h interval. Experiments were performed 5 d after transduction, with gating on GFP-positive cells to assess transduced cells.

Immunoprecipitation and Blotting.

Lysate was obtained from brain, lung, skin, and intestine tissues by homogenization in Triton-X lysis buffer and Halt protease and phosphatase inhibitor (Thermo Scientific, 1861280) using Qiagen TissueLyser. BDMD cells were lysed with Triton-X lysis buffer and Halt protease and phosphatase inhibitor. All of the lysates were sonicated with Bioruptor Plus (Diagenode) before clearing the debris by centrifugation. To verify loss of Tle1 at the protein level, brain lysates (1.2 mg/mL) were incubated with 5 µg of mouse anti-TLE1 antibody (sc-13368 Santa Cruz) or mouse IgG for 2 h at 4 °C and precipitated with protein G Sepharose beads and immunoblotted with anti-rabbit TLE1 (ab125183) antibody. Expression of NF-κB (Cell Signaling, cat. no. 8242), phospho–NF-κB p65 (Ser536) (Cell Signaling, cat. no. 3636s), Hes1 (Santa Cruz, sc-25392x), β-catenin (Cell Signaling, cat. no. 9562), and GAPDH (Santa Cruz, sc-25778) was detected by immunoblotting equal amounts of lysate.

NF-κB Activation and Immunostaining.

Empty vector or TLE1-expressing THP-1/mL cells were cultured in the presence or absence of 10 ng/mL LPS-K12 (InVivoGen) for 60 min and were permeabilized and stained with anti–NF-κB antibody (Santa Cruz p65, clone sc-372) followed by anti-rabbit IgG Cy5-conjugated (ZyMax) secondary antibody using the IntraPrep intracellular staining kit. Cells were fixed with 1% paraformaldehyde/PBS and stained with DAPI before analysis using the ImageStream flow cytometer (Amnis Corp). Data were analyzed using ImageStream Data Exploration and Analysis Software.

Survival and Growth Analysis.

Mice were genotyped at birth, and WT, het, and knockout littermates were monitored for body weight and survival from birth for over 2 mo.

Lung Analysis.

C57BL/6 mice were euthanized, the neck and chest were carefully opened to avoid injury to the lungs, a catheter was secured in the right atrium, and the pulmonary vessels were perfused with 0.1 M sodium citrate at 15 cm H₂O pressure for 5 min. After cannulation of the trachea, the lungs were inflated at 20 cm H₂O pressure for 10 min with 0.4% low-melt agarose in culture media-maintained liquid by radiant warming. Subsequently, the lungs were carefully excised from the body and immersed in cold PBS to solidify the agarose. Only the left lungs were used for analysis.

Hart’s Elastin Staining.

After fixing with 0.1% glutaraldehyde and 4% (vol/vol) paraforamaldehyde and imbedding the lung in paraffin, 5-µm slices were obtained, dewaxed, and stained with Hart’s reagents using standard techniques (46).

Haematoxylin and Eosin Staining.

Lung, skin, and bone from mice were fixed with neutral buffered formalin. The bones were decalcified with EDTA solution. Paraffin tissues were sectioned to 5 µm thickness, dewaxed, stained with hematoxylin and eosin, and examined with an Olympus BX51 light microscope.

Immunohistochemistry.

Formalin or 0.1% glutaraldehyde and 4% paraformaldehyde-fixed lung paraffin sections were deparaffinized and rehydrated. Following sodium citrate antigen retrieval and blocking with 10% BSA in PBS, sections were incubated with K_i 67 (clone SP6, Abcam) or CD68 (ab955, Abcam) antibodies overnight at 4 °C, followed by biotinylated secondary antibodies (Vector Labs), avidin–biotin complexed peroxidase, and 3,3′-diaminobenzidine (Vector Labs).

RNA Extraction and Semiquantitative and Real-Time RT PCR.

Mouse tissues were homogenized using TissueLyser (Qiagen), and RNA was isolated with RNAeasy mini kit (Qiagen) and treated with DNase to remove the genomic DNA. cDNA synthesis was done with M-MLV reverse transcriptase (Ambion). Semiquantitative RT-PCR was performed using DNA Taq polymerase (Roche). RT-PCR was carried out with IQ SYBR green or TaqMan IQ supermix (BioRad) using a BioRad MyIQ PCR machine. TaqMan probes were used for Csf-2, Mm01290062_m1; Csf-1, Mm00432686_m1; and csf-3, Mm00438335_g1. Invitrogen custom primers were used for IL-6 (TAGTCCTTCCTACCCCAATTTCC, TTGGTCCTTAGCCACTCCTTC), IL-1β (TTCGACACATGGGATAACGAGG, TTTTTGCTGTGAGTCCCGGAG), TNF-α (CCCTCACACTCAGATCATCTTCT, GCTACGACGTGGGCTACAG), and Hes1 (CCAGCCAGTGTCAACACGA, AATGCCGGGAGCTATCTTTCT).

BM Cell Isolation.

Mice were euthanized by CO₂ asphyxiation, and the right femur bone was crushed with pestle and mortar and washed with PBS without calcium and magnesium with 3% FBS. Cells were filtered through a 40-µm filter and used for staining after lysing the red blood cells or for transplant experiments.

Spleen Cell Extraction.

Spleens were minced and then incubated with Hank's Balanced Salt Solution (HBBS) buffer containing 2 mg/mL collagenase IV and 4 U/mL DNase I at 37 °C for 30 min. The digestion was stopped by adding 15 mL of PBS with 3% FBS followed by spinning the cells at 300 × g for 10 min. The red blood cells were lysed with ACK red blood cell lysis buffer.

Blood Analysis.

Blood samples were collected with EDTA-treated tubes. The differential blood count was done with 50 µL of blood using a VetScan HM5 Hematology Analyzer (Abaxis). For flow cytometry analysis, 100 µL of blood sample was treated with ACK lysis buffer, washed, and stained. For blood cytokine and chemokine analysis, blood serum was obtained by allowing blood to clot at room temperature for 30 min and centrifuging at 1,000 × g for 10 min in a refrigerated centrifuge. Blood serum was used for analyzing inflammatory cytokine and chemokine using a mouse cytokine/chemokine magnetic bead panel (MCYTOMAG-70, Millipore).

Transplantation Experiments.

BM cells were collected from CD45.2 Tle1^Δ/Δ mice and their littermate controls, and 1 × 10⁶ cells were injected into the tail veins of lethally irradiated (1,000 cGy) CD45.1 mice.

Methylcellulose Colony Assay.

We mixed 2 × 10⁴ whole BM cells with 1 mL of methylcellulose (M3434, Stem Cell Technology) on a 35-mm plate in duplicate, and colonies were scored after 10 d of culturing.

Tumor Model.

For the orthotopic tumor model, 1 × 10⁵ B6-F10 melanoma cells were suspended in 50 µL of PBS and injected intradermally on the right flank of 6–12-wk-old Tle1^Δ/Δ or WT mice. After 7 d, tumors were measured with a caliper (length and width). Tumor volume was calculated using the formula (L × W × H)/2.

Statistical Analysis.

Statistical analysis was performed with GraphPad Prism 6 software. Two-tailed Mann–Whitney nonparametric U test or Student’s t test was used to test the statistical significance difference of two means between groups. Survival analysis was performed with the Gehan–Breslow–Wilcoxon test. Differences were considered significant when the P value was <0.05. Data shown are mean ± SEM.

PCR Array Gene List.

Inflammatory cytokine and chemokine array—PAMM150Z (SABiosciences).

The following chemokines were used: Ccl1, Ccl11, Ccl12, Ccl17, Ccl19, Ccl2, Ccl20, Ccl22, Ccl24, Ccl3, Ccl4, Ccl5, Ccl7, Cx3cl1, Cxcl1, Cxcl10, Cxcl11, Cxcl12, Cxcl13, Cxcl16, Cxcl3, Cxcl5, Cxcl9, Pf4, Ppbp, and Xcl1.

The following interleukins were used: Il-10, Il-11, Il-12a, Il-12b, Il-13, Il-15, Il-16, Il-17a, Il-17f, Il-18, Il-1a, Il-1b, Il-1rn, Il-2, Il-21, Il-22, Il-23a, Il-24, Il-27, Il-3, Il-4, Il-5, Il-6, Il-7, and Il-9.

Interferons Ifna2 and Ifng were used.

The following growth factors were used: Bmp2, Bmp4, Bmp6, Bmp7, Cntf, Csf1, Csf2, Csf3, Gpi1, Lif, Mstn, Nodal, Osm, Thpo, and Vegfa.

The following TNF Superfamily members were used: Cd40lg, Cd70, Fasl, Lta, Ltb, Tnf, Tnfrsf11b, Tnfsf10, Tnfsf11, and Tnfsf13b.

Other cytokines used were as follows: Adipoq, Ctf1, Hc, Mif, Spp1, and Tgfb2.

The following anti-inflammatory cytokines were used: Ccl19, Il-10, Il-11, Il-12a, Il-12b, Il-13, Il-18, Il-2, Il-22, Il-23a, Il-24, Il-4, Il-6, and Tgfb2.

NF-κB signaling target array—PAMM225Z (SABiosciences).

The following cytokines and chemokines were used: Ccl12, Ccl22 (Mdc), Ccl5 (Rantes), Csf1 (M-CSF), Csf2 (GM-CSF), Csf3 (G-CSF), Cxcl1, Cxcl10 (Inp10), Cxcl3, Cxcl9 (Mig), Fasl (Tnfsf6), Ifnb1, Ifng, Il-12b, Il-15, Il-1a, Il-1b, Il-1rn, Il-2, Il-4, Il-6, Lta (Tnfb), Ltb, Tnf, and Tnfsf10 (Trail).

For inflammation, the following genes were used: acute—C3, C4a, Ccl5 (Rantes), Cfb (Bf), F3, F8, Il-1a, Il-1b, Il-6, Ins2, Stat3, and Stat5b; other inflammatory genes—Agt, Akt1, Ccl12, Ccl22 (Mdc), Ccr5, Cd40 (Tnfrsf5), Cxcl1, Cxcl10 (Inp10), Cxcl3, Cxcl9 (Mig), Il-15, Il-1rn, Il-2, Il-2ra (Cd25), Myd88, Ptgs2 (Cox-2), Sele, Selp, Tnf, and Tnfrsf1b; apoptosis—Agt, Birc2 (c-IAP2), Cd74, Egfr, Fasl (Tnfsf6), Gadd45b, Ifnb1, Ifng, Il-12b, Il-2ra (Cd25), Il-4, Ins2, Lta (Tnfb), Map2k6 (Mek6), Mitf, Mmp9 (Gelatinase B), Nqo1, Nr4a2 (Nur77), Ptgs2 (Cox-2), Stat1, Tnfrsf1b, Tnfsf10 (Trail), Trp53, and Traf2; and antiapoptosis—Adm, Akt1, Bcl2a1a (Bfl-1/A1), Bcl2l1 (Bcl-x), Birc3 (c-IAP1), Ccl12, Cdkn1a (p21Cip1/Waf1), Csf2 (GM-CSF), F3, Fas (Tnfrsf6), Il-1a, Il-1b, Il-2, Il-6, Myd88, Nfkbia (IkBa/Mad3), Sod2, Stat5b, Tnf, and Xiap.

For immune response, the following were used: innate—C3, C4a, Ccl12, Cfb (bf), Ifnb1, Il-12b, Il-6, Ins2, Myd88, Nfkbia (IkBa/Mad3), Stat5b, and Tnf; adaptive—C3, C4a, Ccl12, Cd40 (Tnfrsf5), Cd74, Cd80, Icam1, Ifng, Il-12b, Il-1b, Il-2, Il-4, and Traf2; other immunity-related genes—Cd83, Fas (Tnfrsf6), Fasl (Tnfsf6), Il-1r2, Lta (Tnfb), Ltb, and Tnfsf10 (Trail); and type I IFN-responsive genes—Adm, Ccl12, Ccl5 (Rantes), Cd80, Cdkn1a (p21Cip1/Waf1), Cfb (BF), Cxcl10 (Inp10), Cxcl9 (Mig), Il-15, Il-1rn, Irf1, Myd88, Ncoa3, Stat1, and Tnfsf10 (Trail).