BACTERIAL PATHOGEN-ASSOCIATED MOLECULAR PATTERNS UPREGULATE HUMAN GLUCOCORTICOID RECEPTOR EXPRESSION IN PERIPHERAL BLOOD MONONUCLEAR CELLS.

ABSTRACT: It is well known that bacterial components (pathogen-associated molecular patterns [PAMPs]) induce a proinflammatory response through pattern recognition receptor signaling. What is not known, however, is how the inflammatory response is downregulated. We hypothesize that bacterial products initiate compensatory anti-inflammatory responses by inducing expression of the human glucocorticoid receptor (hGR). Peripheral blood mononuclear cells (PBMCs) were isolated from leukocytes concentrated from single human donors (Leukopaks). PBMCs were treated with a gram-negative bacterial component, LPS, or gram-positive bacterial components, lipoteichoic acid (LTA) or peptidoglycan (PGN), for 1, 3, or 13 h. Protein expression of hGR was evaluated by Western blot analysis. RNA was extracted from similarly treated cells for reverse transcription-polymerase chain reaction analysis of hGR and cytokine expression. At 13 h after LPS treatment, there was an increase in the reference hGR protein (hGRα) expressed within some but not all PBMCs isolated from Leukopaks. There was also a dose-dependent increase in hGRα expression with increasing concentrations of PGN (10 and 50 µg/mL). LTA, however, did not affect hGRα expression. PGN also increased the mRNA expression of an hGR splice variant, hGR-B(54). The mRNA expression changes for the inflammatory cytokines were Leukopak specific. We found that cell wall components of both gram-positive and gram-negative bacteria can increase the expression of hGRα. Although these PAMPs augment the inflammatory response, it seems that there is a simultaneous upregulation of hGRα expression. Because binding of cortisol to hGRα typically induces anti-inflammatory proteins, the same PAMPs that induce an inflammatory response seem to also initiate a negative feedback system by inducing hGRα expression in PBMCs.

A Novel Human Glucocorticoid Receptor Variant, G459V, is Hyperactive in Response to Steroids.

ABSTRACT: A potential cause of the variable response to injury and sepsis is the variability of a patient's human glucocorticoid receptor (hGR) profile. To identify hGR variants, blood samples were collected on admission and biweekly thereafter from hospitalized patients who sustained at least a 20% total body surface area burn injury. A hyperactive G1376T single-nucleotide polymorphism (SNP) isoform was identified. This SNP led to a single amino acid change of glutamine to valine at site 459, "G459V," in the DNA-binding domain. The isoform's activity was tested in a reporter assay after treatment with steroids, the hGR antagonist RU486 (mifepristone) alone, or RU486 followed by steroids. When treated with hydrocortisone, the hGR G459V isoform had a hyperactive response; its activity was over 30 times greater than the reference hGRα. Unexpectedly, G459V had significantly increased activity when treated with the hGR antagonist RU486. With the combination of both RU486 and hydrocortisone, G459V activity was repressed, but greater than that of RU486 alone. Finally, when hGRα was cotransfected with G459V to simulate isoform interaction, the activity was closer to that of the hGRα profile than the G459V isoform. The unique activity of the G459V isoform shows that some variants of hGR have the potential to alter a person's response to stress and steroid treatment and may be a factor as to why mitigating the clinical response to sepsis and other stressors has been so elusive.

Uncovering a multitude of human glucocorticoid receptor variants: an expansive survey of a single gene.

BACKGROUND: Glucocorticoids are commonly used in the clinical setting for their potent anti-inflammatory effects; however, significant variations in response to treatment have been demonstrated. Although the underlying mechanisms have yet to be fully understood, this variable response may be a result of alterations in human glucocorticoid receptor (hGR) expression and function. In addition to hGRα, the biologically active isoform, a screening of current databases and publications revealed five alternative splice isoforms and hundreds of variants that have been reported to date. Many of these changes in the hGR-coding gene, NR3C1, have been linked to pathophysiology. However, many studies focus on evaluating hGR expression in vitro or detecting previously reported variants. RESULTS: In this study, blood from healthy volunteers, burn and asthma patients, as well as from peripheral blood mononuclear cells isolated from leukoreduced donor whole blood, were screened for NR3C1 isoforms. We identified more than 1500 variants, including an additional 21 unique splice isoforms which contain 15 new cryptic exons. A dynamic database, named the Universal hGR (UhGR), was created to annotate and visualize the variants. CONCLUSION: This identification of naturally occurring and stress-induced hGR isoforms, as well as the establishment of an hGR-specific database, may reveal new patterns or suggest areas of interest that will lead to the improved understanding of the human stress response system.

Lipopolysaccharide Stress Induces Cryptic Exon Splice Variants of the Human Glucocorticoid Receptor.

Glucocorticoids are widely used in the treatment of numerous inflammatory conditions, including sepsis. Unfortunately, patient response to glucocorticoid therapy can be inconsistent. Variations in the human glucocorticoid receptor (hGR) may contribute to the differential patient response. We screened for hGR variants in the buffy coats of burn patients and peripheral blood mononuclear cells (PBMCs) treated with lipopolysaccharide. Three novel splice variants containing cryptic exons were upregulated in the PBMCs after lipopolysaccharide exposure at 3 and 13 h with the greatest observed expression at 3 h. Luciferase assays revealed that two of the isoforms had no significant activity in comparison with the reference hGR when stimulated with hydrocortisone. The third isoform had an augmented response that was greater than the reference hGR at a high cortisol dose. This shows that PBMCs are able to produce variant hGR isoforms in response to stress. Furthermore, lipopolysaccharide stress appears to induce these hGR variants, potentially by influencing mRNA splicing. In the future, identifying hGR expression profiles may be a key component in individually tailoring a patient's treatment to sepsis and injury.

A novel human glucocorticoid receptor SNP results in increased transactivation potential.

Glucocorticoids are one of the most widely used therapeutics in the treatment of a variety of inflammatory disorders. However, it is known that there are variable patient responses to glucocorticoid treatment; there are responders and non-responders, or those that need higher dosages. Polymorphisms in the glucocorticoid receptor (GR) have been implicated in this variability. In this study, ninety-seven volunteers were surveyed for polymorphisms in the human GR-alpha (hGRα), the accepted biologically active reference isoform. One isoform identified in our survey, named hGR DL-2, had four single nucleotide polymorphisms (SNPs), one synonymous and three non-synonymous, and a four base pair deletion resulting in a frame shift and early termination to produce a 743 amino acid putative protein. hGR DL-2 had a decrease in transactivation potential of more than 90%. Upon further analysis of the individual SNPs and deletion, one SNP, A829G, which results in a lysine to glutamic acid amino acid change at position 277, was found to increase the transactivation potential of hGR more than eight times the full-length reference. Furthermore, the hGRα-A829G isoform had a differential hyperactive response to various exogenous steroids. Increasing our knowledge as to how various SNPs affect hGR activity may help in understanding the unpredictable patient response to steroid treatment, and is a step towards personalizing patient care.

Highly Variable Genomic Landscape of Endogenous Retroviruses in the C57BL/6J Inbred Strain, Depending on Individual Mouse, Gender, Organ Type, and Organ Location.

Transposable repetitive elements, named the "TREome," represent ~40% of the mouse genome. We postulate that the germ line genome undergoes temporal and spatial diversification into somatic genomes in conjunction with the TREome activity. C57BL/6J inbred mice were subjected to genomic landscape analyses using a TREome probe from murine leukemia virus-type endogenous retroviruses (MLV-ERVs). None shared the same MLV-ERV landscape within each comparison group: (1) sperm and 18 tissues from one mouse, (2) six brain compartments from two females, (3) spleen and thymus samples from four age groups, (4) three spatial tissue sets from two females, and (5) kidney and liver samples from three females and three males. Interestingly, males had more genomic MLV-ERV copies than females; moreover, only in the males, the kidneys had higher MLV-ERV copies than the livers. Perhaps, the mouse-, gender-, and tissue/cell-dependent MLV-ERV landscapes are linked to the individual-specific and dynamic phenotypes of the C57BL/6J inbred population.

A Tri-Nucleotide Pattern in a 3' UTR Segment Affects The Activity of a Human Glucocorticoid Receptor Isoform.

We previously identified a truncated human glucocorticoid receptor (hGR) isoform of 118 amino acids, hGR-S1(-349A), that despite lacking the major functional domains, was more hyperactive after glucocorticoid treatment than the full-length receptor. Furthermore, its 3' untranslated region (UTR) was required. To dissect the underlying mechanisms for hyperactivity, a series of hGR isoforms with consecutive deletions in the 3' UTR were created to test their transactivation potential using reporter assays. The hGR-S1(-349A) isoform retaining 1303 bp of 3' UTR displayed unusually high activity with or without glucocorticoid stimulation. Unexpectedly, a complete loss of significant activity was observed with isoforms retaining 1293 bp or 1263 bp of 3' UTR. Analysis of the 20 bp region neighboring the 1293 bp site showed a pattern: 3'UTR termination at every third base pair in this region resulted in a loss of transactivation potential while the other sites retained hyperactivity with or without glucocorticoid stimulation. Variations in the activity of an hGR isoform, due to changes in the 3' UTR sequence configuration, may provide an important link in explaining inconsistent responses to glucocorticoid treatment in individuals and ultimately enable tailored, patient-specific care. Furthermore, understanding the mechanisms underlying the cyclic hyperactivity/loss of activity phenomenon may be a step toward identifying a novel mechanism of gene regulation.

Inherently variable responses to glucocorticoid stress among endogenous retroviruses isolated from 23 mouse strains.

Active participation of endogenous retroviruses (ERVs) in disease processes has been exemplified by the finding that the HERV (human ERV)-W envelope protein is involved in the pathogenesis of multiple sclerosis, an autoimmune disease. We also demonstrated that injury-elicited stressors alter the expression of murine ERVs (MuERVs), both murine leukemia virus-type and mouse mammary tumor virus (MMTV)-type (MMTV-MuERV). In this study, to evaluate MMTV-MuERVs' responses to stress (e.g., injury, infection)-elicited systemic glucocorticoid (GC) levels, we examined the GC-stress response of 64 MMTV-MuERV promoters isolated from the genomes of 23 mouse strains. All 64 promoters responded to treatment with a synthetic GC, dexamethasone (DEX), at a wide range from a 0.6- to 85.7-fold increase in reporter activity compared to no treatment. An analysis of the 10 lowest and 10 highest DEX responders revealed specific promoter elements exclusively present in either the three lowest or the two highest responders. Each promoter had a unique profile of transcription regulatory elements and the glucocorticoid response element (GRE) was identified in all promoters with the number of GREs ranging from 2 to 7. The three lowest DEX responders were the only promoters with two GREs. The findings from this study suggest that certain MMTV-MuERVs are more responsive to stress-elicited systemic GC elevation compared to the others. The mouse strain-specific genomic MMTV-MuERV profiles and individual MMTV-MuERVs' differential responses to GC-stress might explain, at least in part, the variable inflammatory responses to injury and/or infection, often observed among different mouse strains. This article is part of a Special Issue entitled: Immune and Metabolic Alterations in Trauma and Sepsis edited by Dr. Raghavan Raju.

Genomic landscapes of endogenous retroviruses unveil intricate genetics of conventional and genetically-engineered laboratory mouse strains.

Laboratory strains of mice, both conventional and genetically engineered, have been introduced as critical components of a broad range of studies investigating normal and disease biology. Currently, the genetic identity of laboratory mice is primarily confirmed by surveying polymorphisms in selected sets of "conventional" genes and/or microsatellites in the absence of a single completely sequenced mouse genome. First, we examined variations in the genomic landscapes of transposable repetitive elements, named the TREome, in conventional and genetically engineered mouse strains using murine leukemia virus-type endogenous retroviruses (MLV-ERVs) as a probe. A survey of the genomes from 56 conventional strains revealed strain-specific TREome landscapes, and certain families (e.g., C57BL) of strains were discernible with defined patterns. Interestingly, the TREome landscapes of C3H/HeJ (toll-like receptor-4 [TLR4] mutant) inbred mice were different from its control C3H/HeOuJ (TLR4 wild-type) strain. In addition, a CD14 knock-out strain had a distinct TREome landscape compared to its control/backcross C57BL/6J strain. Second, an examination of superantigen (SAg, a "TREome gene") coding sequences of mouse mammary tumor virus-type ERVs in the genomes of the 46 conventional strains revealed a high diversity, suggesting a potential role of SAgs in strain-specific immune phenotypes. The findings from this study indicate that unexplored and intricate genomic variations exist in laboratory mouse strains, both conventional and genetically engineered. The TREome-based high-resolution genetics surveillance system for laboratory mice would contribute to efficient study design with quality control and accurate data interpretation. This genetics system can be easily adapted to other species ranging from plants to humans.

Repetitive element signature-based visualization, distance computation, and classification of 1766 microbial genomes.

The genomes of living organisms are populated with pleomorphic repetitive elements (REs) of varying densities. Our hypothesis that genomic RE landscapes are species/strain/individual-specific was implemented into the Genome Signature Imaging system to visualize and compute the RE-based signatures of any genome. Following the occurrence profiling of 5-nucleotide REs/words, the information from top-50 frequency words was transformed into a genome-specific signature and visualized as Genome Signature Images (GSIs), using a CMYK scheme. An algorithm for computing distances among GSIs was formulated using the GSIs' variables (word identity, frequency, and frequency order). The utility of the GSI-distance computation system was demonstrated with control genomes. GSI-based computation of genome-relatedness among 1766 microbes (117 archaea and 1649 bacteria) identified their clustering patterns; although the majority paralleled the established classification, some did not. The Genome Signature Imaging system, with its visualization and distance computation functions, enables genome-scale evolutionary studies involving numerous genomes with varying sizes.

Temporal and spatial rearrangements of a repetitive element array on C57BL/6J mouse genome.

Repetitive elements (REs) make up the vast majority of the mammalian genomes. We identified species-specific genomic libraries of RE arrays. The non-random configurations of RE arrays suggest their functions. We tested whether RE arrays undergo age- and tissue/cell-specific rearrangements. An RE array of C57BL/6J mice, containing tandem repeats of a mosaic of transposable REs, was selected to examine rearrangements in different ages and tissues. There were marked changes in the array configuration in the genomes of the skin and brain in all mice of six weeks and older, whereas the heart and liver had alterations at 29weeks. The temporal variations were confirmed by identifying putative rearrangement junctions. Temporal and spatial rearrangements of certain RE arrays may contribute to the acquired characteristics of the genome information system.

Hyperactive Human Glucocorticoid Receptor Isoforms and their Implications for the Stress Response.

Glucocorticoids are indispensable therapeutic agents in diseases of inflammation, but their effectiveness in treating advanced septic shock has been inconsistent. Our understanding of the mechanisms causing this variability to steroid therapy remains limited. Previous studies in our laboratory have implicated human glucocorticoid receptor (hGR) polymorphisms as one of the likely reasons for this variability. We examined the effect of two single-nucleotide polymorphisms (SNPs) on the transactivation potential of the hGR in the absence and presence of exogenous steroids. An isoform containing a novel naturally occurring human SNP, T1463C, was found to have a hyperactive response with treatment of all three steroids examined while maintaining low activity in the absence of steroids, relative to reference hGR. In comparison, another hGR isoform with the A2297G SNP, previously identified in our laboratory, demonstrated hyperactive transactivational response in the absence of steroids; however, it had a significant increase in activity after treatment with only one of the glucocorticoids (hydrocortisone) tested. These results offer a possible explanation for the clinical variability seen among individuals in response to stress or shock.

Divergent and dynamic activity of endogenous retroviruses in burn patients and their inflammatory potential.

Genes constitute ~3% of the human genome, whereas human endogenous retroviruses (HERVs) represent ~8%. We examined post-burn HERV expression in patients' blood cells, and the inflammatory potentials of the burn-associated HERVs were evaluated. Buffy coat cells, collected at various time points from 11 patients, were screened for the expression of eight HERV families, and we identified their divergent expression profiles depending on patient, HERV, and time point. The population of expressed HERV sequences was patient-specific, suggesting HERVs' inherent genomic polymorphisms and/or differential expression potentials depending on characteristics of patients and courses of injury response. Some HERVs were shared among the patients, while the others were divergent. Interestingly, one burn-associated HERV gag gene from a patient's genome induced IL-6, IL-1ß, Ptgs-2, and iNOS. These findings demonstrate that injury stressors initiate divergent HERV responses depending on patient, HERV, and disease course and implicate HERVs as genetic elements contributing to polymorphic injury pathophysiology.

Injury-elicited stressors alter endogenous retrovirus expression in lymphocytes depending on cell type and source lymphoid organ.

BACKGROUND: Murine leukemia virus-type endogenous retroviruses (MuLV-ERVs) constitute ~10% of the mouse genome and are associated with various pathophysiologic processes. In this study, we examined whether MuLV-ERVs' response to burn-elicited stressors is specific for certain lymphocyte populations and/or locations of lymphoid organ. RESULTS: B- and T-cells, which were sorted from nine lymphoid organs of C57BL/6J mice after burn, were subjected to MuLV-ERV expression analyses. Overall, the post-burn MuLV-ERV expression pattern was dependent on lymphocyte type, time after injury, location of lymphoid organ, and MuLV-ERV type. For instance, the MuLV-ERV expression in T-cells from the thymus and three cervical lymph nodes decreased at 3 hours post-burn while the expression of some MuLV-ERVs was augmented in B-cells derived from the mesenteric lymph node. The MuLV-ERV U3 sequences population of the burn-24 hours group was less diverse in comparison to the no burn and burn-3 hours groups. In addition, it was apparent that at the 24 hours time point, the U3 populations of B-cells from both no burn and burn groups were less heterogeneous than the T-cells' U3 populations. Using the U3 sequences, some of which were isolated only from specific experimental groups (B- vs. T-cells; no burn vs. burn), as probes, 51 putative MuLV-ERVs, including 16 full-length proviruses, were mapped followed by characterization of their biologic properties. CONCLUSION: MuLV-ERVs' response to burn-elicited stressors may be differentially controlled depending on lymphocyte type, location of lymphoid organ, MuLV-ERV type, and stress duration.

Single nucleotide polymorphisms and type of steroid impact the functional response of the human glucocorticoid receptor.

BACKGROUND: Clinical trials evaluating the use of steroids in septic shock have shown variable outcomes. Our previous studies have implicated human glucocorticoid receptor (hGR) polymorphisms as a possible cause of altered steroid response. To further evaluate this variability, we hypothesized that hGR polymorphisms along with type of steroid influence the functional response. METHODS: Total RNA was isolated from healthy human blood samples and surveyed for the hGR gene. The National Center for Biotechnology Information hGRα sequence was used as a reference, and two unique single nucleotide polymorphisms (SNPs) (A214G and T962C) were selected for evaluation. Functional response was measured using a luciferase reporting assay after transfecting hGR isoforms into tsA201 cells and stimulation with graded concentrations of hydrocortisone (HYD), methylprednisolone (MPS), and dexamethasone (DEX). RESULTS: Each isoform had a unique dose-response curve with the optimal activity depending on concentration and type of steroid. The presence of either SNP A214G or T962C resulted in a decreased response when compared with hGRα when stimulated with HYD (P < 0.01). The same decreased response occurred for the SNPs with DEX stimulation, but at a much lower concentration range than HYD (P < 0.01). However, in the presence of MPS, SNP A214G resulted in greater activity when compared with hGRα (P < 0.01), whereas the presence of T962C resulted in activity equivalent to hGRα. CONCLUSIONS: SNPs, type of steroid, and concentration range impact the functional response of the hGR. A greater understanding of hGR polymorphisms and steroid response may further elucidate mechanisms explaining the variable response seen with patient treatment.

Enhanced steroid response of a human glucocorticoid receptor splice variant.

Glucocorticoids remain a recommended therapy in advanced septic shock despite the often unpredictable response, and our understanding of the mechanisms regulating the steroid and stress response remains limited. Since the initial sequencing of the human glucocorticoid receptor α and ß gene (hGRα and hGRß), only three additional splice variants have been identified--all of which have been postulated to contribute to steroid resistance. During a survey of 97 healthy humans' blood, we identified two novel hGR splice isoforms (hGR-S1 and hGR-S1(-349A) retaining intron H between exons 8 and 9. Human GR-S1(-349A) contained a base deletion causing an early termination and a truncated protein of 118 amino acids, whereas hGR-S1 had an early termination occurring within intron H and resulted in a 745-amino acid protein. Both isoforms had decreased transactivation potentials compared with hGRα when tested in the absence of exogenous steroids. However, after treating with exogenous steroids, dose-response studies showed hGR-S1(-349A) had a substantial augmentation in activity at higher concentrations of hydrocortisone and methylprednisolone when compared with hGRα, whereas hGR-S1 did not. Removal of the 3' untranslated region (3'UTR) of the hGR-S1(-349A) mRNA sequence resulted in a loss of the augmented response. The isoform hGR-S1(-349A) augments the response to steroids, and this significant response appears to be critically regulated by the 3'UTR. The identification and evaluation of these unique hGR isoforms helps further the understanding of the complex genetic regulation of the stress and steroid response.

Novel hyperactive glucocorticoid receptor isoform identified within a human population.

Glucocorticoids serve as important therapeutic agents in diseases of inflammation, but clinical use, especially in advanced septic shock, remains controversial because of the unpredictable response. Prior studies correlate human glucocorticoid receptor (hGR) isoforms with a decreased response to steroid therapy. Further analysis of additional hGR isoforms may improve the understanding of the steroid response. Ninety-seven human volunteers' blood samples were surveyed for hGR isoforms. An isoform matching National Center for Biotechnology Informatics (NCBI) hGRα (hGR NCBI) served as a reference. Two isoforms were of particular interest-one isoform had three nonsynonymous single-nucleotide polymorphisms (SNPs) (hGR NS-1), and the second had a single-nucleotide deletion (hGR DL-1) resulting in a truncated protein. Transactivation potentials were measured using a luciferase reporter assay. Human glucocorticoid receptor NS-1 had activity more than twice of hGR NCBI, whereas hGR DL-1 demonstrated less than 10% of the activity of hGR NCBI. Cotransfection of two isoforms revealed that the presence of hGR NS-1 increased transactivation potential, whereas hGR DL-1 decreased activity. Synthetic constructs isolating individual and paired SNPs of hGR NS-1 were created to identify the SNP responsible for hyperactivity. Transactivation studies revealed a SNP within the ligand-binding domain exerted the greatest influence over hyperactivity. In evaluating the response to hydrocortisone, hGR NCBI and hGR NS-1 displayed an increased dose-dependent response, but hGR NS-1 had a response more than twice hGR NCBI. Characterization of the novel hyperactive hGR NS-1 provides insight into a possible mechanism underlying the unpredictable response to steroid treatment.

Cosegregation of CD14 locus and polymorphic alleles of glucocorticoid receptor and protocadherins into CD14 knockout mouse genome.

Targeted mutagenesis technology has been used to investigate the biological characteristics of genes. We incidentally observed a discrepancy in the size of the glucocorticoid receptor (GCR) between CD14 knockout (KO) mice and backcross controls. The CD14 KO mice were generated using 129S4/SvJae-derived J1 embryonic stem cells, C57BL/ 6J donor blastocysts, and C57BL/6J backcross strain. In this study, the extent of genomic heterogeneity of the CD14 KO mice, potentially affecting their phenotype, was characterized in comparison to C57BL/6J controls. There were polymorphic alleles of the GCR gene in CD14 KO and C57BL/6J mice: a tandem repeat of 8 CAGs and 17 CAGs in the transactivation domain, respectively. The subsequent finding of eight CAGs in all 129 substrains examined and colocalization of CD14 and GCR genes on the same contig on chromosome 18 suggested that the GCR allele in the J1 embryonic stem cell genome cosegregated with the targeted CD14 locus. Interestingly, all three clusters of the protocadherin family, central genes in determining neuronal networks, were mapped between the CD14 and GCR loci. Further analyses revealed numerous nonsynonymous coding single-nucleotide polymorphisms within the protocadherin family between CD14 KO and C57BL/6J mice. In addition, heterogeneous profiles of endogenous retroviruses, which constitute approximately 10% of the genome, were observed between them. These findings suggest that cosegregation of genes flanking the targeted locus leads to a substantial level of genetic heterogeneity in CD14 KO mice compared with their backcross controls. Phenotypic changes observed in some KO mice may not be as definitive as expected.

Conjugated linoleic acid alters matrix metalloproteinases of metastatic mouse mammary tumor cells.

Conjugated linoleic acid (CLA) is a group of linoleic acid derivatives that has been implicated in animal studies to reduce a number of components of mammary tumorigenesis. Previously, we showed that CLA could alter the latency and metastasis of the highly metastatic transplantable line 4526 mouse mammary tumor. Several possible mechanisms have been proposed for the actions of CLA, but here we assessed how CLA may act to alter the expression and activity of matrix-modifying proteins within tumors from line 4526. In vitro, highly metastatic mouse mammary tumor cells had significantly decreased invasiveness after treatment with CLA, an indication that matrix-modifying proteins may have been altered. Using these same highly metastatic cells, primary tumors were grown in mice of separate groups fed 0, 0.1, 0.5, and 1% CLA (wt:wt) and evaluated for their levels and activities of matrix-modifying enzymes, enzyme inhibitors, and enzyme activators. The addition of CLA to the diet increased steady-state levels of messenger RNA (mRNA) of the matrix metalloproteinases (MMP) -2 and -9 in primary tumors removed from mice. However, western analysis revealed that although relative levels of the proform of MMP-9 were consistent with the mRNA observations, MMP-2 proform levels were actually decreased by dietary CLA. The activity of MMP-2 was barely detectable, but gelatin zymography and an in vitro activity assay showed that MMP-9 activity was significantly decreased by CLA. The steady-state mRNA and protein levels of tissue inhibitors of metalloproteinase-1 (TIMP-1) and TIMP-2, natural inhibitors of MMP, were increased at higher dietary CLA levels relative to low or no CLA. Suppression of MMP activity, therefore, may be 1 pathway through which CLA reduces tumor invasion and spread.

Beef tallow increases the potency of conjugated linoleic acid in the reduction of mouse mammary tumor metastasis.

Animal studies consistently show that dietary conjugated linoleic acid (CLA) reduces mammary tumorigenesis including metastasis. Relatively low concentrations of CLA are required for those effects, and a threshold level exists above which there is no added reduction. We reasoned that the concentration of CLA required to effectively alter mammary tumor metastasis may be dependent on the type of dietary fat because select fatty acids can enhance or suppress normal or malignant cell growth and metastasis. For this study, the diets (a total of 12 different groups) differed in fatty acid composition but not in energy from fat (40%). In experiments involving spontaneous metastasis, mice were fed for 11 wk; in experiments in which mice were injected i.v. with tumor cells, they were fed for 7 wk. Mice were then assessed for the effect of CLA concentration on mammary tumorigenesis. Mammary tumor growth was not altered, but metastasis was significantly decreased when beef tallow (BT) replaced half of a defined vegetable fat blend (VFB). That blend reflects the typical fat content of a Western diet. In addition, that same VFB:BT diet lowered the concentration of CLA required to significantly decrease mammary tumor metastasis from 0.1% of the diet to 0.05%. A diet in which corn oil replaced half of the VFB did not lower the threshold from 0.1 to 0.05%. In vitro, the main fatty acid in vegetable oil, linoleic acid, reduced the efficacy of CLA toxicity on mammary tumor cells in culture. Alternatively, fatty acids normally found in BT, such as oleic, stearic, and palmitic acids, either did not change or enhanced the cytolytic effects of CLA isomers on mouse mammary tumor cells in culture. These data provide evidence that dietary BT, itself with negligible levels of CLA, may increase the efficacy of dietary CLA in reducing mammary tumorigenesis.