Extract of Oregano, Coffee, Thyme, Clove, and Walnuts Inhibits NF-κB in Monocytes and in Transgenic Reporter Mice

Recent studies have identified NF-κB as a direct link between inflammation and cancer (1–3) and thus renders this family of transcription factors as a key molecular target for both prevention and treatments of cancers. The NF-κB family of transcription factors are crucial mediators of cellular stress, immune, and inflammatory responses (4). Although NF-κB is essential in normal physiology, several human disorders involve inappropriate regulation of NF-κB. Several chronic degenerative diseases (5–7) and various human cancers (8) have been associated with an aberrant upregulation of NF-κB activity. Both the NF-κB activation and cytokine profile of tumor-associated macrophages are closely linked to tumor growth (9, 10).

In vertebrates, the NF-κB family is comprised of five structurally related proteins [p65 (REL A), p50, p52, c-REL, REL B] that form heterodimers and homodimers. Normally, the NF-κB dimers are sequestered in the cytoplasm by binding to inhibitory factors. Two pathways lead to the delocalization of NF-κB dimers to the nucleus and consequent transcriptional regulation of target genes (11). The classic NF-κB signaling pathway is activated by proinflammatory stimuli, bacterial or viral infection, and various forms of stress such as UV radiation and environmental toxins (12) and is essential for the innate immune system and antiapoptotic signaling. The alternative NF-κB pathway plays a critical role for both the development and maintenance of the adaptive immune system and is activated by members of the tumor necrosis factor (TNF) family other than TNF-α (11). NF-κB target genes code for proteins that are central players in inflammation, activation of the immune system, and antiapoptotic signaling, which can be involved in both the promotion and progression of cancers (13). With the central role of NF-κB activation in inflammation, cancer, and antiapoptosis, substances with the ability to inhibit NF-κB activation may be valuable in a generally healthy diet, in chemoprevention (14), and in combination with chemotherapy (15).

Plants and plant-based diets contain thousands of phytochemicals (16) in addition to well-known macronutrients and micronutrients. In epidemiologic studies, diets rich in plant-based foods reduce the risk of several cancers (17) and several chronic diseases (18). Clinical trials with single compounds have, however, not been able to reproduce these risk reductions (17, 19). For these reasons, it is likely that several of these compounds work in additive, synergistic, or antagonistic fashions to create the observed health effects and thus foods or combinations of foods may be better suited as the basic unit for research on these preventive effects. The preventive properties of such diets are likely to be complex combinations of numerous mechanisms, including reduced inflammation, cell proliferation and migration, and growth factor activity, and increased phase II detoxification and endogenous antioxidant defense, of which the latter two processes are orchestrated through the regulation of the transcription factor nuclear factor erythoid 2–related factor 2 (Nrf2; ref. 14). Aberrant cellular signaling is a hallmark of carcinogenesis and targeting crucial signaling pathways, such as the NF-κB and Nrf2 signaling pathways, yield valuable progress toward the elucidation of the mechanisms underlying chemoprevention. We previously screened a wide range of dietary plants and phytochemicals for their effect on NF-κB activation and found clove, thyme, oregano, coffee, and walnuts to be the most potent inhibitors (20). In this study, we show that an extract of these five dietary plants synergistically inhibit NF-κB activation in vitro and can also inhibit NF-κB activation and induce Nrf2 activity in vivo in transgenic mice. Interestingly, lowered NF-κB activation was detected in the liver and intestine, as well as in testis and epididymis of male mice, indicating organ-specific effects of the dietary plants. In addition, we found the dietary plants to reduce the expression of genes related to inflammation and cell migration and proliferation in the livers of these mice.

U937-κB cells are previously described (21). Cell culture conditions and experimental set up is previously described (20). Briefly, NF-κB activity was induced by lipopolysaccaride (LPS; 1 μg/mL) 30 minutes before addition of extracts and cells were coincubated with LPS and extract for an additional 6 hours. Cell viability was determined by trypan blue exclusion. A cutoff value of 10% nonviable cells was used.

HepG2 cells (hepatocellular carcinoma cell line) were cultured and transiently transfected with an construct containing two EpRE sequences coupled to the luciferase gene as previously described (22). The cell culture medium was replaced before the addition of extract and the cells were incubated with the indicated concentrations for 17 hours. Luciferase activity was measured in cell lysates as previously described (22).

Coffee (Arabica, Medium roast), walnuts, oregano (dried), clove (dried), and thyme (dried) were purchased from a grocery store in Oslo, Norway. All samples were pulverized and extracted and concentrated by use of water/methanol (50:50,v/v) as previously described (20). The final extract was diluted in corn oil (extract used for mice) or a 50:50 mix of PBS/DMSO (v/v; for cells). The extract to be used for cell culture was thereafter sterile filtered. When measured as redox-active compounds by the ferric reducing ability of plasma (FRAP) assay, the extract administered to the mice corresponds to ∼10 times the average daily intake in the Norwegian diet (data not shown).

The mice were housed in accordance with the guidelines of the Federation of European Laboratory Animal Science Associations and animal experiments were done according to national guidelines for animal welfare. The mice had ad libitum access to water and chow (except fasting 3 h before per os feeding). Transgenic mice with B6 background carrying a transgene with binding sites for either NF-κB (with three κB sites; ref. 21) or Nrf2 [with two electrophile response elements (EpRE); ref. 22] coupled to the luciferase gene were used. A single dose of an extract corresponding to 60 mg each of thyme, oregano, and clove, and 600 mg each of walnuts and coffee or vehicle control (corn oil) was administered by oral gavage to the mice. NF-κB transgenic mice for both groups were n = 16 (10 female and 6 male) and 11 to 24 weeks of age. EpRE transgenic mice for both groups were n = 5 female and 12 to 17 weeks of age. No differences in NF-κB activity were found between the female and male NF-κB mice and thus the results from both sexes were pooled. For the NF-κB mice, LPS (2.5 mg/kg) in 100 μL PBS was injected s.c. on the back near the tail 3 hours after oral gavage. In vivo imaging was done at the given time points. Directly following imaging at 6 hours, the NF-κB mice were euthanized; ex vivo imaging of the intestine was done; and all other organs were frozen in liquid nitrogen for further analysis. Luciferase activity in homogenates of organs was measured as previously described (21).

Luciferase activity was measured by use of an IVIS 100 Imaging System (Caliper Life Sciences). d-luciferin (0.2 mg/mL medium) was added to the cell culture medium and cells were incubated at 37°C for 4 minutes. Mice were anesthetized using 2.5% isoflurane and were shaved on the abdomen. d-luciferin (Biosynth AG; 160 mg/kg) in PBS was injected i.p. and mice were imaged 7 minutes after injection of d-luciferin. Cell culture plates or mice were placed in a light proof chamber and the luminescence emitted from the cells or mice was detected for 1 minute. Ex vivo imaging of the intestine was done 20 minutes after the injection of d-luciferin. The number of photons emitted was calculated by use of the Living Image Software (Caliper Life Sciences). Grayscale images were used for reference of position.

Quantitative real-time PCR was done on liver tissue from NF-κB mice included in the experiment (n = 5 for both LPS and LPS + extract), plus two nontreated NF-κB mice for reference to the basal expression of mRNA. Total RNA was isolated by use of the RNeasy Mini kit (Qiagen) and cDNA synthesis was done using the High capacity cDNA reverse transcription kit (Applied Biosystems, Inc.). Quantitative real-time PCR was done in a 7900HT Fast Real-Time PCR System (Applied Biosystems, Inc.) using the Taqman Mouse Immune Array (Applied Biosystems, Inc.). RQ Manager v.1.2 (Applied Biosystems, Inc.) was used for data analysis. The amount of target gene relative to reference gene was quantified using the cycle threshold (Ct) with glucuronidase β as the endogenous control gene. Thus, −ΔCt represents Ct_{target gene} − Ct_{glucuronidase β} and was used to calculate 2^−ΔCt.

The U937-κB cell cultures may vary in their NF-κB response to LPS depending on the passage of the cells. Therefore, the control (no extract) response to LPS was idiosyncratic for each experiment and the most stable set of experimental conditions is within a particular experiment. For these reasons, we set the control LPS level as 1, meaning 100% of NF-κB activity, and computed the responses to the LPS plus extract treatments as fold changes compared with controls.

One-way ANOVA was used to examine effects of extracts on luciferase activity in U937-κB and HepG2 cells. Differences were identified using Dunnett's comparisons.

For the calculations of synergy, we used the natural logarithm of the NF-κB activity as the response metric, which takes a value of 0 for the control situation (response to LPS alone), a negative number for inhibition of the NF-κB activity to LPS plus extract, and a positive number for accentuation of the activity. We ran a linear regression model with dependent variable ln(NF-κB), with no intercept, which was included as independent variables all five main effects of extracts, and a single term representing the combined extract. As the main test of synergy, we then estimated the contrast for the combined extract effect minus the sum of the five corresponding main effects.

For in vivo experiments, the area under the curve (AUC) of LPS-induced luciferase activity from 0 to 6 hours was calculated for each mouse. Student's t tests were performed to compare the data from AUC, at each time point, from ex vivo imaging, and from luciferase activity in tissue homogenates. Due to a lower number of animals in the Nrf2 experiment, the two treatments were compared statistically using nonparametric Mann-Whitney Test. Principal component analysis was performed using The Uscrambler v.9.8 (Camo Software AS, http://www.camo.com) on the whole gene expression data set to identify possible intervariable relations between genes that may identify patterns caused by the treatment. Statistically significant difference was set to P < 0.05 for all analysis. Results are presented as mean ± SEM, unless otherwise noted.

An extract made from the combination of clove, oregano, thyme, walnuts, and coffee inhibited LPS-induced NF-κB activity in the U937-κB cells in a dose-dependent manner (Fig. 1A). In the U937-κB cells, basal NF-κB activity was 0.05-fold ± 0.009-fold (5%) of LPS-induced NF-κB activity (Fig. 1A). The extract made from 0.3 mg/mL coffee and walnut, and 0.03 mg/mL thyme, oregano, and clove reduced LPS-induced NF-κB activation to 0.74-fold ± 0.05-fold (P < 0.001) of LPS-controls, whereas the extract made of 0.6 mg/mL coffee and walnut, and 0.06 mg/mL thyme, oregano, and clove further reduced luciferase activity to 0.51-fold ± 0.04-fold (P < 0.001) of LPS controls. Increasing the concentration to 1.5/0.15 mg/mL and 3/0.3 mg/mL (coffee and walnut/thyme, oregano, clove) almost completely blocked LPS-induced NF-κB activation to, respectively, 0.09-fold ± 0.02-fold and 0.08-fold ± 0.01-fold (P < 0.001 for both) without cytotoxicity.

To test for synergy, we calculated the difference between the effects of the combined extract and the sum of effects for the five single ingredients tested separately. Effects of single ingredients are shown in Fig. 1B. When comparing equal concentrations of the ingredients (3 mg/mL each of coffee and walnut, and 0.3 mg/mL each of thyme, oregano, and clove), the combined extract inhibited NF-κB significantly more than the expected additive inhibitory effects from the five ingredients tested separately (Table 1A), indicating synergistic effects by the combination extract. At combined extract concentration of 1.5 mg/mL each of coffee and walnut, and 0.15 mg/mL each of thyme, oregano, and clove, the combined extract was still significantly more potent in inhibiting NF-κB activation compared with the expected sum of inhibitions from the five ingredients at 3 mg/mL coffee and walnut, and 0.3 mg/mL thyme, oregano, and clove (Table 1B). The combined extract concentration of 0.6 mg/mL of coffee and walnut, and 0.06 mg/mL of thyme, oregano, and clove was equally potent as the expected sum of the effects for 3 mg/mL of coffee and walnut, and 0.3 mg/mL of thyme, oregano, and clove (Table 1C). A way of expressing the synergy is to note that one fifth of the total volume of food extract was needed to obtain equal NF-κB inhibition using the combined extract as was expected based on the separate effects of the individual food extracts.

Next, we investigated whether the combination extract could modulate LPS-induced NF-κB activity in vivo. An s.c. injection of LPS (160 mg/kg) increased NF-κB activity to a maximum of 15.7-fold ± 2.4-fold (n = 16) after 4 hours as measured by whole body in vivo imaging of the luciferase activity. The extract or control vehicle (corn oil) was administered by oral gavage to mice 3 hours (at time −3 h) before s.c. LPS injection (at time 0 h). Whole body in vivo imaging was done at 0, 2, 4, and 6 hours (Fig. 2A and B). At 2 hours (P = 0.043) and 6 hours (P = 0.041) after LPS injection, the extract significantly inhibited the LPS-induced NF-κB activity, compared with mice treated with LPS only. Based on the AUC, the extract decreased LPS-induced NF-κB activity the first 6 hours by 35% (P = 0.040) compared with control mice (Fig. 2C).

Following in vivo imaging at 6 hours, the mice were euthanized and luminescence was measured in the intestine by ex vivo imaging and in homogenates of other tissues. There was a significant reduction in NF-κB activity to 0.53-fold (P = 0.028) of controls, in the intestine of mice receiving LPS plus the extract compared with mice receiving LPS only, as measured by ex vivo imaging (Fig. 3A). Representative intestines from the LPS only– and the LPS + extract–treated mice are shown in Fig. 3B. Furthermore, comparing luciferase activity by luminometry of tissue homogenates of organs from mice receiving the extract to mice receiving LPS only, revealed significantly decreased NF-κB activity in the liver (0.67-fold of control; P = 0.046), testis (0.66-fold of controls; P = 0.007), and epididymis (0.48-fold of control; P = 0.021; Fig. 4C). In the spleen, the NF-κB activity was further increased in the mice receiving the extract (1.57-fold of control; P = 0.047). Although there was a suggestive but not significant increase in heart NF-κB activity, no other differences in luciferase activity were found in brain, thymus, lung, kidney, fat, muscle, or uterus.

We proceeded to investigate the expression levels of genes related to immune responses, cancer development, and NF-κB activity in livers of mice receiving LPS + extract versus LPS only. Two nontreated mice were included for reference to non–LPS-induced expression levels. The principal component analysis score plot (Fig. 4A) shows that the LPS and LPS + extract treatments are separated by PC1 and PC2, which totally explain 51% of the variation in the gene expression data set. The correlation loading (Fig. 4B) indicate that a set of genes are important in terms of explaining the variance in the data and thus the effect of the treatment. This group of genes important for the PC1 and PC2 (Fig. 4B, enhanced area) include chemokine ligand 3 (Fig. 4C, Ccl3), Cd80 antigen (Fig. 4D, CD80), Cd40 antigen (Fig. 4E, CD40), chemokine ligand 10 (Fig. 4G, Cxcl10), chemokine ligand 11 (Fig. 4H, Cxcl11), interleukin 1 α (Fig. 4K, Il1α), interleukin 6 (Fig. 4M, Il6), interleukin 12b (Fig. 4N, Il12b), interleukin 15 (Fig. 4O, Il15), nuclear factor κB 1 (Fig. 4P, Nfkb1; p50/p105), nuclear factor κB 2 (Fig. 4Q, Nfkb2; p52/p100), nitric oxide synthase 2 inducible (Fig. 4R, Nos2), protein tyrosine phosphatase receptor type C (Fig. 4S, Ptprc), TNF α (Fig. 4U, Tnf), and vascular cell adhesion molecule 1 (Fig. 4V, Vcam1), which are also all significantly differently expressed between the two treatments. In addition, colony stimulating factor 1 (macrophage) (Fig. 4F, Csf1), endothelin 1 (Fig. 4I, Edn1), heme oxygenase 1 (Fig. 4J, Hmox1), interleukin 2 receptor α chain (Fig. 4L, Il2ra), and transforming growth factor β 1 (Fig. 4T, Tgfb1) were all differentially expressed in the livers from the LPS + extract–treated mice compared with in the livers of mice treated with LPS only. Of the 20 above-mentioned genes, all except interleukin 15 show a significantly lower mRNA expression in the livers of LPS + extract mice.

Previous research has found correlations between the ability to induce Nrf2 and anti-inflammatory properties, and a cross-talk between NF-κB and Nrf2 has been suggested (23–25). Thus, we tested the ability of the combination extract to induce Nrf2 activity through EpRE sites. In HepG2 cells, the combination extract potently induced Nfr2 activity at a concentration of 3 mg/mL coffee and walnut plus 0.3 mg/mL thyme, oregano, and clove, whereas one tenth of this concentration did not induce Nrf2 activity (Fig. 5A). Furthermore, we monitored the Nrf2 activity in EpRE-luciferase mice over the same time span as for the NF-κB mice. A single dose of the combination extract or control vehicle was administered by oral gavage to mice at time 0 hour and luminescence was measured at 0, 2, 4, and 6 hours. The mice receiving the extract had significantly increased Nrf2 activity at 2 and 4 hours, and higher total activity (calculated by AUC), compared with the mice receiving control vehicle only (Fig. 5B).

Inhibitors of NF-κB carry great promise for both prevention and therapy in cancers and chronic inflammation. An extract made from coffee, walnut, thyme, oregano, and clove inhibited NF-κB activation in U937-κB cells in a synergistic fashion compared with the effects of the single extracts. By using noninvasive molecular imaging of transgenic NF-κB luciferase reporter mice, we showed that the same combined extract efficiently inhibit LPS-induced NF-κB activation in vivo. Furthermore, a large set of downstream target genes were found to have significantly lower expression levels in the livers of mice receiving the extract, compared with in the LPS-treated mice. NF-κB may also be influenced by the activation of other transcription factors, such as Nrf2, which is orchestrating the endogenous antioxidant defense and phase II enzymes through binding to EpRE sites. In this study, we also found induction of Nrf2 activity by the combination extract both in vitro and in vivo.

As recently shown, NF-κB is a direct link between inflammation and cancer (1–3), and aberrant activation of NF-κB has been found in several types of human cancers (8). Inflammation coordinated through NF-κB activation will most often promote tumor growth (13) and inflammation-promoted tumor growth can be turned to tumor regression by inhibition of NF-κB in cancerous cells (2). NF-κB is essential for survival and cytokine production of monocytes and macrophages (26) and further NF-κB activation in macrophages is associated with cancer development. The cytokine profiles of monocytes and macrophages are also closely linked to tumor growth (9, 10). In the liver, the connections between hepatitis infection, inflammation or fibrosis, and development of hepatocellular carcinoma is well established (27, 28). Kupffer cells (resident liver macrophages) display increased NF-κB activation in response to liver injury, resulting in production and secretion of proinflammatory cytokines such as TNF-α and IL-6 (29). These cytokines may in turn promote fibrosis, tumor growth, and metastasis (27, 30); however, knocking out NF-κB activation in myeloid cells has been shown to reduce both cytokine production and hepatocellular carcinoma (10). In this study, 18 of 20 genes displaying a significantly different liver expression in the LPS + extract–treated mice compared with mice receiving LPS only are known NF-κB target genes (ref. 12 and references herein). These genes include a range of inflammatory cytokines/chemokines (including TNF-α and IL-6), immune receptors (Cd80, Cd40, Il2ra, and Ptprc), growth factor (Csf1), cell adhesion molecule (Vcam1), enzymes (Hmox1 and Nos2), and mitogen (Edn1), plus two NF-κB proteins (Nfkb1 and Nfkb2). The combination of chemokines, Vcam1, Nos2, and Csf1 are important in leukocyte recruitment, infiltration, and differentiation in tissues (reviewed in refs. 31, 32). Thus, the reduced expression of genes coding for these molecules, in combination with the observed decrease in LPS-induced NF-κB activation in the monocytic cell line as well as in the liver (and intestine) following administration of plant extracts, suggest that these plant extracts may contribute to chemoprevention by altering the microenvironment to a less inflammatory, less tumor-promoting milieu. In addition, to the extent that inflammation is an important factor in disease conditions other than cancer, the consumption of these foods should be regarded as part of a generally healthy diet.

The inhibition of NF-κB activity in liver, testis, and epididymis indicate that compounds from the extract have been absorbed and are bioactive after absorption. In the intestine, the extract may produce its effect both by acting in the intestinal lumen and after absorption into intestinal cells. Further work should be done to identify the bioactive compounds as well as the mechanisms of action, in all four organs. The intestine and liver are the main organs for metabolism of phytochemicals. There is limited knowledge on the tissue distribution of phytochemicals present in our extract. However, quercetin, resveratrol, and epigallocatechin gallate have previously been detected in the intestine, liver, and testes of mice and rats (33–35). Consistent with these results, we would expect the highest concentrations of phytochemicals from our extract in these organs and thus also a greater chance of detectible effects on a major signaling pathway such as NF-κB.

Induction of Nrf2 may affect the organ specificity of NF-κB responses as there are lines of evidence suggesting a cross-talk between the NF-κB and Nrf2 signaling pathways (23–25), and this cross-talk is most likely to be evident in organs with the highest Nrf2-activity. A wide range of phytochemicals are known to activate Nrf2, leading to induced expression of phase 2 enzymes and endogenous antioxidants (collectively called cytoprotective proteins; ref. 36), and many of these phytochemicals are also known to inhibit NF-κB (25). We also found our combination extract to induce Nrf2 activity, both in vitro and in vivo in transgenic EpRE reporter mice. Like NF-κB, Nrf2 has been implicated in carcinogenesis and inflammation both in liver and the gastrointestinal tract (37–39), and an induction of cytoprotective proteins is an alternative approach to reducing the tissue damage caused by carcinogens or excessive inflammatory response.

Phytochemicals are mainly metabolized in the intestine and liver. Thus, other organs will be exposed to metabolites or conjugates of the ingested compounds, possibly resulting in organ-specific NF-κB responses. In the spleen and possibly in the heart, the LPS-induced NF-κB activation was further increased in the mice receiving extract. It should be noted, however, that the absolute NF-κB activity (adjusted for protein contents of the samples) in the liver is 8- and 4-fold higher than in spleen and heart, respectively. In combination with the large size of the liver, this implies that the liver is contributing considerably more than any other organs to the overall NF-κB activity in the mice. We speculate that it would be an advantage to maintain a high NF-κB activity in the spleen during an infection to ensure the activity of adaptive immune responses and that this activation of NF-κB might be through the alternative NF-κB pathway, which is important in both development and maintenance of the spleen (11).

In the search for mechanisms behind the protective effects of plant-based diets, reductionistic approaches studying isolated compounds have dominated. Such an approach might fit well into pharmacologic research, but in doing so, we might overlook additive, synergistic, or antagonistic effects that arise from combinations of compounds found in plant foods and diets. Thus, we have chosen to make extracts of foods the focus of this study. In the present work, we found synergistic inhibitory effects on LPS-induced NF-κB activation with the combination of five dietary plant extracts, compared with the sum of effects of the five individual extracts. Keeping the concentration for the single extracts constant, the inhibitory synergistic effect on NF-κB activation by the combined extract persisted even at half the concentration. At a level of one fifth of the concentration, the added concentration of extracted foods in the combined extract is comparable with each individual extract tested alone, meaning that the effect of the combined extract is not simply an effect of a higher dose of food. This underlines the potential of synergistic effects being created by a variety of foods. Such effects would not be detectable in a more reductionistic model.

The most recent report from the World Cancer Research Fond/American Institute for Cancer Research (17) concludes that coffee is unlikely to have an effect on cancers of the pancreas or kidneys. At present, there is limited evidence on the effects of coffee intake on other cancer sites; however, some epidemiologic studies suggest preventive effects of coffee in the liver (40, 41). In a large epidemiologic study, we observed that intake of coffee was inversely associated with reduced death attributed to oxidative stress and inflammatory diseases (42). Oxidative stress and inflammation are both closely linked to NF-κB activation and are involved in cancer initiation and development (43, 44). In this perspective, inhibition of NF-κB might contribute to the inverse relationship between coffee intake and disease, with the liver as a main target organ.

Walnuts, clove, thyme, and oregano are all rich in redox-active phytochemicals (45, 46); however, none of these dietary plants have been extensively studied with respect to chemopreventive properties. The observed NF-κB inhibition by our combination extract may, at least partly, be attributed to rosmarinic acid (47) found in thyme and oregano, or eugenol (48) found in clove, which are all identified as inhibitors of NF-κB; however, thymol, a major phytochemical in thyme, did not modulate NF-κB activity (20). Thyme and oregano essential oils in combination decreased the levels of IL-1β and IL-6, as well as inflammation-related tissue damage in a model of colitis (49), both of which may be a result of NF-κB inhibition in the colon. In addition, thyme may induce the level of endogenous cytoprotective proteins in the liver (50). Walnuts are also a rich source of essential fatty acids with known anti-inflammatory properties (51).

Approaches that focus on foods rather than single compounds should be valuable in future research to unravel the chemopreventive effects of foods. To our knowledge, this is the first report to indicate synergistic effects of foods on NF-κB activation; however, more research is warranted to confirm and to expand the yet limited knowledge about these synergistic effects. Our experiments show that ordinary dietary plants may be potent modulators of the NF-κB signaling pathway both in vitro and in vivo, and thus support a role of chemoprevention through dietary means.

R. Blomhoff has interests in Cgene AS, which was established by Birkeland Innovation (the Technology transfer office at the University of Oslo), and holds the commercial rights to the transgenic luciferase reporter mice. D.R. Jacobs, Jr. is an unpaid member of the Scientific Advisory Council of the California Walnut Commission. The other authors disclosed no potential conflicts of interest.

The U937-κB cells were generously donated by Christian Drevon, Department of Nutrition, University of Oslo. We thank Siv K. Bøhn, Sverre Holm, Ingrid F. Gjelstad, Hanne Reiersen, Lene Gilen, and Kari Holte for the technical assistance and Cgene for providing the IVIS100 imaging system and the reporter mice.

Grant Support: Grants from the Norwegian Cancer Society (I. Paur, L.M. Austenaa, and R. Blomhoff) and the Johan Throne Holst Foundation (R. Blomhoff and D.R. Jacobs).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.