ABSTRACT
TANNINS AS ANTI-INFLAMMATORY AGENTS
By Melanie Diane Jeffers
The potential of dietary and herbal polyphenols to act as beneficial antioxidants or anti-inflammatory agents, or as harmful pro-oxidants, is a topic of intense current interest. The first half of this research involved using SDS-PAGE gels to elucidate tannin-protein complexation. The second half of this study attempts to evaluate the anti- inflammatory or pro-inflammatory activity of the tannin pentagalloyl glucose within an ex vivo model system using peripheral blood mononuclear cells from rabbit. The oxidative burst from cells stimulated with phorbol ester was monitored with cytochrome c. The results provide preliminary data that supports the hypothesis that tannins such as pentagalloyl glucose minimize the inflammatory response.
TANNINS AS ANTI-INFLAMMTORY AGENTS
A Thesis
Submitted to the
Faculty of Miami University
in partial fulfillment of
the requirements for the degree of
Masters of Science
Department of Chemistry and Biochemistry
By
Melanie Diane Jeffers
Miami University
Oxford, Ohio
2006
Advisor ______Dr. Ann E. Hagerman Reader ______Dr. John W. Hawes Reader ______Dr. Gary A. Lorigan Reader ______Dr. Kevin W. Kittredge Table of Contents Page List of Figures iv Abbreviations vi 1. Introduction: Tannins as Anti-Inflammatory Agents 1.1. Tannins 1 1.1.1. What is a tannin 1 1.1.2. Human exposure to tannins 2 1.1.3. Tannin-protein complexes 3 1.1.4. Anti-nutritional effects of dietary tannins 4 1.2. Tannins and Reactive Oxygen Species 4 1.2.1. Tannins and the inflammatory system 6 1.2.2. Inflammatory diseases 7 1.2.3. Inflammatory bowel diseases 8 1.3. Hypothesis 10 1.4. Chapter 1 References 12 2. Characterization of soluble protein-tannin complexes using Ribonuclease A and Pentagalloyl glucose (β-1,2,3,4,6-Pentagalloyl-O-D-Glucopyranose) 2.1. Introduction 18 2.2. Material and Methods 19 2.3. Results and Discussion 20 2.4. Chapter 2 References 41 3. Analytical methods for monitoring the inflammatory the response 3.1. Introduction 43 3.1.1. Superoxide as one of the first signaling molecules 43 3.1.2. Rationale for using cytochrome c 44 3.2. Methods 45 3.2.1. Reagents 45 3.2.2. Validation 45 3.3. Results and Discussion 45 3.4. Chapter 3 References 51
ii 4. Investigation of tannins as anti-inflammatory agents 4.1. Introduction 53 4.1.1. Peripheral blood mononuclear cells (PBMCs) 53 4.2. Methods 55 4.2.1. Reagents 55 4.2.2. Isolation of peripheral blood mononuclear cells (PBMC) 56 4.2.3. Incubation and priming of PBMCs 57 4.2.4. PMA stimulation of PBMCs 57 4.2.5. Superoxide production from PBMCs 57 4.2.6. PGG interference with cytochrome c 57 4.2.7. Lucigenin as a probe for superoxide 58 4.3. Results 58 4.4. Discussion 60 4.5. Chapter 4 References 70
iii List of Figures 1. Chapter 1. Figures Page 1.1. Structures of representative tannins. 11 2. Chapter 2. Figures 2.1. Optimal pH for PGG-RNase complex formation. 26 2.2. Optimal RNase:PGG ratio for complex formation. 27 2.3. Effect of excess PGG on complex formation. 28 2.4. SDS PAGE to detect PGG-RNase complexes. 29 2.5. Higher percent acrylamide SDS PAGE to better resolve PGG-RNase complexes. 30 2.6. Gradient SDS-PAGE to detect PGG-RNase complexes. 31 2.7. Effect of dialysis on PGG-RNase complexes. 32 2.8. Effect of β-ME on PGG-RNase complexes. 33 2.9. Order of addition of reagents. 34 2.10. Effect of urea on PGG-RNase complex formation. 35 2.11. Proteolysis with trypsin. 36 2.12. Proteolysis with subtilisin and trypsin. 37 2.13. Proteolysis and Tris/Tricine gels. 38 2.14. Proteolysis with 14C radiolabelled PGG using subtilisin. 39 2.15. Periodate reaction with RNase. 40 3. Chapter 3. Figures 3.1. Cytochrome c spectrum. 48 3.2. Cytochrome c reduction by stimulated monocyte cells. 49 3.3. Relationship between reduced cytochrome c and slope of the spectral peak. 50 4. Chapter 4. Figures 4.1. Superoxide production from rabbit PBMCs in the presence of cytochrome c. 63 4.2. Superoxide production represented as slope of the spectral data. 64 4.3. Superoxide production from cells exposed to tannin. 65
iv 4.4. PGG reduces cytochrome c. 66 4.5. Superoxide production in PGG-treated cells. 67 4.6. Superoxide production from rabbit PBMCs probed with lucigenin. 68 4.7. Superoxide production from cells exposed to tannin using lucigenin as a probe. 69
v Abbreviations
2,2’-azinobis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS•+) bovine serum albumin (BSA) cyclooxygenase (COX-2) dimethyl sulfoxide (DMSO) ethylenediamine-tetraacetic acid (EDTA) epigallocatechin gallate (EGCG) gastrointestinal (GI) Hank’s balanced salt solution (HBSS)
helper T-cells (Th)
hydrogen peroxide (H2O2) hydroxyl radical (OH•) hypochlorous acid (HOCl) inducible nitric oxide synthase (iNOS) inflammatory bowel disease (IBD) interferons (IFN) interleukins (IL) lipopolysaccharide (LPS) β-mercaptoethanol (β-ME) nitric oxide radical (NO•) nitroblue tetrazolium (NBT) β-1,2,3,4,6-pentagalloyl-O-D-glucopyranose (PGG) peripheral blood mononuclear cells (PBMC) phorbol 12-mysristate 13-actetate (PMA)
potassium superoxide (KO2) procyanidin (PC) prostaglandin E2 (PGE2) reactive oxygen species (ROS) ribonuclease A (RNase)
sodium dithionite (Na2S2O4)
vi sodium dodecyl sulfate (SDS) sodium periodate (NaIO4) •- superoxide anion (O2 ) superoxide dismutase (SOD) tumor necrosis factors (TNF)
vii 1. Introduction: Tannins as Anti-Inflammatory Agents 1.1. Tannins 1.1.1. What is a tannin? Tannins are a complex family of naturally occurring polyphenols that most characteristically precipitate gelatin and other proteins (1). Other phenolic compounds also have the ability to bind with protein, but do not form the insoluble, cross-linked complexes typically formed by tannin-protein mixtures (2). Tannins were originally used in the process of tanning or conversion of animal hides into leather by precipitating the proteins, such as collagen, found in the skin of the animal (3). Their name is derived from the Celtic word for oak, a traditional source for tannin-containing extracts for preparing leather (4). Tannins are secondary metabolites produced by plants presumably to provide protection from disease and mammalian herbivores. These polymeric polyphenols, which have molecular weights ranging from 0.5 to 22 kD, have unique chemical and biological activities (5). In addition to binding protein, tannins are also potent metal ion chelators, tightly binding essential metals such as iron and calcium (6, 7). Polymeric polyphenols are also potent radical scavengers and antioxidants (8) and thus have been targeted as good candidates for therapeutic agents against diseases associated with oxidative stress. There are two classes of tannins that are of biological significance to humans; condensed and hydrolyzable tannins. Condensed tannins, or proanthocyanidins, can all be degraded to the pigments known as anthocyanidins. The flavanoid backbones of the condensed tannins are derived from phenylalanine and polyketide biosynthesis (9). The majority of condensed tannins are produced from the flavan-3-ols (+)-catechin (2,3-trans) and (-)-epicatechin (2,3-cis) (10). For example, procyanidin (PC) is a polymer of catechin and epicatechin subunits (Fig.1.1). Hydrolyzable tannins maintain a core polyol, usually glucose, with galloyl groups derived from shikimic acid esterified to the core (9). These tannins can polymerize through oxidative coupling of the galloyl groups resulting in complex structures known as ellagitannins. Pentagalloyl glucose (β-1,2,3,4,6- pentagalloyl-O-D-glucopyranose) or PGG, is one of the simplest gallotannins (Fig.1.1).
1 It consists of a glucose core with five galloyl groups linked to the five alcoholic positions on the pyranose ring by ester bonds (4). PGG is the polyphenol used in the experiments described in this thesis. 1.1.2. Human exposure to tannins Tannins, both condensed and hydrolyzable, are widely dispersed in the plant kingdom, with some plants containing as much as several grams of tannin per kilogram of tissue (2). Fruits such as blueberries and strawberries are some of the most commonly consumed tannin-containing foods in the American diet. Beverages such as red wine, cranberry juice, and apple juice contain approximately 750, 248, and 48 mg tannin per liter, respectively (2). It is estimated that tannin makes up somewhere between 37-56% of green tea constituents and that as much as 10% of black tea solids are made up of unoxidized catechins (11). Consumption of tannin-containing food products varies from region to region. Legumes and beans, staples in many countries, contain large quantities of condensed tannins (12, 13). Sorghum, which contains between 0.13-7.2% condensed tannin, is a common food staple in Asia, Africa, and countries in the Near and Middle East (14). The North American diet, which is relatively poor in grains, legumes and fruits, contains much less tannin. It has been reported that tannin consumption in India ranges from 1.5- 2.5 g/day, depending on the region, and about 1g/day within the USA (2). Tannins are important components of medicinal plants widely used in many cultures, particularly in Asia. For instance, shigyaku-san, used for gastritis and peptic ulcers in Japan, has the root of Paeonia among its many components. The outer skin of this root is also used to treat disorders of the cardiovascular system, such as high blood pressure. PGG is thought to be the biologically significant portion of that root (15). Polyphenolic compounds, when applied topically, may aid in wound healing, burns and inflammation by creating an impenetrable layer with protein or polysaccharide and therefore natural healing can occur underneath (5). Other plants reported to have medicinal properties, and that contain polyphenolic compounds include; Arctostaphylos uva-ursi, used as a diuretic for kidney disorders; Geranium maculatum, used as an anti- inflammatory agent and astringent; and Rubus idaeus or raspberry leaves, used for disorders of the stomach in children (5).
2 1.1.3. Tannin-protein complexes The reversible reaction between protein and tannin was the subject of research dating as early as 1803, when the topic was first described by Sir Humphry Davy (5). Tannins bind with protein to form complexes that may be soluble or insoluble, depending on the reaction conditions. In general, excess tannin tends to promote formation of insoluble complexes, and as protein concentration is increased in comparison to tannin, soluble complexes are more likely to form (16). In the gastrointestinal (GI) tract, protein is present in large excess over tannin; thus, soluble complexes are probably the predominant species. Under non-oxidizing conditions, both soluble and insoluble complexes dissociate upon treatment with sodium dodecyl sulfate (SDS) or other protein denaturants (17). This is consistent with hydrogen bonding and hydrophobic interactions between tannin and protein, as suggested by previous studies (18-20). However, when protein is reacted with oxidized tannin, the complexes generated are SDS-resistant, which is suggestive of covalent stabilization of the complex (17). Complexation is dependent upon the pH of the solution, the chemical nature of the tannin, and the characteristics of the protein (5, 17). In general, precipitation of various proteins by tannin is maximized at or near the isoelectric point of the protein (21). The conformational flexibility of the polyphenol, as well as the protein it is associating with, is an important factor influencing the capacity to complex. As more conformational constraints are placed on either the tannin or the protein, their ability to complex is significantly reduced (5). The solubility of the tannin in water is also a determining factor in the interaction. The less soluble the tannin is, the more likely it is to have strong associations with protein, suggesting the importance of hydrophobic interactions in complexation. Tannin-protein complexes may be very specific for particular proteins including those that have high proline content, a high molecular weight, and an open conformation (22). Saliva contains relatively large quantities of proline-rich proteins and their high affinity for tannins suggests that they may function as tannin-binding proteins (23). Perhaps having an abundance of proline-rich proteins located in the upper gastrointestinal
3 (GI) tract is a protective mechanism that humans and other animals have adapted to defend against the costs of tannin ingestion (24-26). 1.1.4. Anti-nutritional effects of dietary tannins The post-ingestive consequences of tannin consumption have been extensively examined. In general, high levels of dietary tannins limit protein digestibility (27) and reduce uptake of essential metals such as iron (7, 28). Dietary tannin reduces the growth rate of young animals (29). It has been reported that for some herbivorous insects, pupal development, fecundity, and/or feeding habits are all affected by tannins in leaves (30). Most of the direct effects of dietary consumption of tannins are thought to be associated with protein binding in the digestive tract, preventing absorption and utilization of dietary protein. After ingestion, tannins appear to remain in the GI tract without significant absorption. Jiminez-Ramsay et al. (1994) fed experimental animals 14C-labeled condensed tannins and found the majority of the radiolabel in the excreta of the animals (31). However, they demonstrated the presence of some 14C in plasma and other tissues such as the liver. It was unclear whether the label represented intact tannin or metabolites of ingested tannin (31). Similarly, Skopec et al. (2004) fed radiolabeled PGG to rats and found label in the feces, especially when the rats were producing salivary proline-rich proteins (32). Due to their large size, tannins may not be directly absorbed into the bloodstream. Their tendency to complex with proteins and precipitate in aqueous conditions, as well as their polarity, further explains their limited uptake from the GI tract. 1.2. Tannins and Reactive Oxygen Species Despite the negative effects of dietary tannins on protein and metal ion uptake, some dietary tannins may play a role in disease prevention. Several epidemiological studies have demonstrated that diets rich in polyphenolics are associated with diminished risk of diseases such as cancer and heart disease. In particular, the Rotterdam Study, a 5.6 year follow up involving 4,807 subjects older than 55, found a 65% reduction in non- fatal myocardial infarctions for subjects that consumed >33mg flavonols per day (33). The Finnish Mobile Clinic Health Examination Survey, a 28 year follow-up with greater than 10,000 subjects, reported an inverse relationship between consumption of flavonol, flavone, and flavanone containing compounds and occurrence of lung cancer (34). This
4 study also found that a diet rich in these compounds resulted in a reduced risk for asthma and diabetes. These results should be interpreted with caution because the results were based on dietary assessments without the validation of internal biomarkers such as the concentration of flavanoid content within the blood plasma or other target tissues (35). Cancer, cardiovascular diseases, and other diseases are associated with production of reactive oxygen species (ROS). ROS, which can directly damage protein, lipids and nucleic acids, are biologically produced either as a by-product of normal mitochondrial metabolism, or as components of the inflammatory pathway. In general, antioxidants inhibit production of ROS or quench radical species, while anti-inflammatory agents suppress individual biochemical steps in the inflammatory cascade. Although radical quenching can be measured by simple chemical methods using easily detected radical species, anti-inflammatory activities can only be measured in cells or tissues capable of an inflammatory reaction. Reports of both antioxidant and anti-inflammatory activity of tannins suggest a likely mechanism for their role in disease prevention. Some antioxidant activity is associated with radical scavenging. Measures of radical scavenging generally use either the very reactive hydroxyl radical (OH•) or less reactive model radicals such as ABTS•+ (2,2’-azinobis(3-ethylbenzothiazoline-6-sulfonic acid)). Hagerman et al. found phenolics to have an experimental rate constant for scavenging OH• ranging from 1.1 x 109 – 3.1 x 1011 M-1s-1, demonstrating their ability to function as radical scavenging antioxidants (36). Additionally, tannins were found to be more effective scavengers of ABTS•+ than the standard model biological antioxidant, trolox (36). Radical damage in biological systems is due, at least in part, to formation of the hydroxyl radical by reactions similar to the Fenton reaction (Eq.1), and some antioxidants function by inhibiting that reaction. 2+ 3+ • - Fe + H2O2 → Fe + OH + OH (Eq.1) Compounds that chelate metal ions, including some tannins, may minimize OH• production by inhibiting the Fenton reaction (36). Like other redox-active compounds, tannins may have pro-oxidant characteristics that counteract their antioxidant activities. Under conditions that stabilize semiquinone radicals, tannins can oxidatively damage proteins by covalent reaction (37). In some
5 cases, pro-oxidants may initiate the inflammatory pathway, by providing the stimulus for inflammation. For instance, condensed tannins inhibit phagocytosis and induce arachidonic acid production in macrophages, and stimulate chemotactic factors from neutrophils, all of which indicate activation of the inflammatory response (38-41). The occupational inflammatory lung disease, byssinosis, has been linked to air-borne condensed tannin present in grain dust in grain elevators and cotton mills (42). 1.2.1. Tannins and the inflammatory system Tannins could affect the inflammatory response via their radical scavenging activities. There is ample evidence supporting the connection between inflammation and free radical reactions (43-51). Production of superoxide, hydroxyl radical, and other oxidizing species initiates a feedback cycle, with the radical products of inflammatory reactions serving as inducing agents for further inflammation. For example, nitric oxide (NO•), a radical produced by nitric oxide synthase, acts as a second messenger during the process of inflammation. Inducible nitric oxide synthase (iNOS), is produced in response to a variety of inflammatory cytokines as well as lipopolysaccharide (LPS) (52). The iNOS stimulates production of NO• radical which in turn stimulates further inflammation. There have been several promising studies linking the radical scavenging activity of tannins with the inflammatory system, and many of those studies have attempted molecular definition of the active component in the tannin. For example, PGG, which has five galloyl ester groups, has strong anti-tumor activity, via reduced iNOS expression. PGG is more active than other galloylglucoses (53). Gallic acid by itself did not have an effect on the iNOS activity. This implies that the gallate structure, rather than gallic acid moieties themselves, are responsible for the anti-inflammatory and anti- carcinogenic effects observed (53-55). In contrast, Fujiki et al. found that 3-O-tetragalloylquinic acid was the most potent anti-tumor substance tested. However, when this compound was further galloylated, its anti-tumor activity did not improve. Theaflavin-3,3’-O-digallate, with two gallic acid moieties, showed strongest anti-inflammtory and anti-proliferative activity on tumor cells in a separate study (56, 57). Lin and Lin (1997) reported that the galloyl group and the phenolic hydroxyl groups at the 3’ position on EGCG (epigallocatechin gallate), which is a tannin found in green tea, were responsible for its anti-inflammatory
6 properties seen as the inhibition of iNOS in macrophages that have been activated with LPS (58). It has been suggested that the mechanism by which tannins inhibit inflammatory markers may be through oxidation of the tannin and reduction of ROS including free radicals (55). Taken together, these observations further suggest that the high level of galloylation of PGG and the available phenolic groups on these gallic acid groups is what promotes the anti-inflammtory/anti-carcinogenic characteristics associated with tannins (55).
PGG has also been shown to inhibit prostaglandin E2 (PGE2) production whereas
gallic acid only inhibited PGE2 accumulation slightly (55). However, Park et al. demonstrated that EGCG up-regulates cyclooxygenase (COX-2) expression leading to
up-regulated PGE2 production in unstimulated macrophages (59). Thus, free gallic acid; glucose galloyl esters; and flavan-3-ol galloyl esters all have different effects on the prostaglandin-inflammatory response. Despite the fact that many structure-function relationship studies of polymeric polyphenols have been attempted, the important determinants of activity have not been clearly identified, in part because each study focuses on a narrow range of compounds. One of our goals is to establish a method for evaluating the anti-inflammatory activities of a broad range of polyphenols so that we can build on the current understanding of how tannin structure affects its activity in vivo. We have chosen to use an ex vivo model of inflammation for our study. 1.2.2. Inflammatory diseases In healthy individuals, inflammation is a protective response by the body brought about by injury. Inflammation is the body’s attempt to dilute, destroy or wall-off the noxious agent. As a result there are changes in physiology such as swelling, redness, and pain brought about by dilation of the capillaries, increased blood flow, and exudation of plasma fluids into the site of injury. One key inflammatory event is the migration of phagocytic cells to the site of inflammation to destroy the harmful substance. These cells release antigens that then bind to recognition components of the immune system, which results in a coordinated set of events that becomes amplified by the release of pro- inflammatory mediators (60).
7 Macrophages are important components of the normal inflammatory response and of the abnormal response characteristic of inflammatory diseases. Macrophages are activated by and secrete cytokines such as tumor necrosis factors (TNF), interferons (IFN), and interleukins (IL). As macrophages digest microorganisms, they release IL-1,
which stimulates helper T-cells (Th). Macrophages also display antigenic determinants that activate T-cells. Activated T-cells release a host of other interleukins, namely interleukins 2-6, that will in turn activate B-cells to differentiate into cells that release specific antibodies against the microorganism (61). Cytokines are important in mediating the inflammatory response because of their ability to activate monocytes and macrophages. Macrophage activation must be tightly regulated to avoid an inappropriate or extended inflammatory response (62). In a normal inflammatory process, mediators such as suppressor T-cells are released that serve to “shut-down” the inflammatory process once the noxious agent has been contained. When the normal regulatory controls of the inflammatory response do not operate correctly, the consequence may be manifestation of inflammatory diseases such as arthritis, asthma, and inflammatory bowel disease (IBD). The uncertain post-digestive fate of polyphenolics makes it difficult to evaluate whether they play a role in pathologies of peripheral tissues. Low molecular weight non- nutritive antioxidants such as flavanoids are readily absorbed from the GI tract, and it has been suggested that they could supplement endogenous antioxidants, such as superoxide dismutase (SOD) and glutathione (63) or nutritive antioxidants such as vitamins C and E. However, high molecular weight tannins, which are poorly absorbed by the body, may not be particularly effective in the cardiovascular system or other organs. However, free and protein-bound polymeric polyphenols, which have both been found to be efficient radical scavengers (8), may be biologically effective antioxidants that are localized in the GI tract. 1.2.3. Inflammatory bowel diseases The inflammatory diseases are all associated with uncontrolled inflammation and pain. IBD such as Crohn’s disease, ulcerative colitis, gastritis, ulcers, and diverticulitis involve a consistent and re-occurring inflammatory component in the GI tract that results in oxidative damage (64). Within the normal intestinal mucosa, pro-inflammatory
8 cytokines (TNF-α, IFN-γ, IL-1, IL-6, IL-12) and anti-inflammatory cytokines (IL-4, IL- 10, IL-11) are carefully balanced to allow for controlled inflammation (65). It has been suggested that IBD and Crohn’s disease probably result from a failure to down-regulate
the intestinal Th1 inflammatory response (65). Both humoral and cellular components of the immune system may either instigate or propagate inflammatory diseases. During inflammation within the GI tract of IBD patients, phagocytic leukocytes accumulate within the interstitium of the intestines and colon, and release mediators that are involved in the inflammatory process (60). The digestive tract is constantly exposed to ROS because of the reactive oxygen species formed from dietary transition metals, oxidized lipids, and pathogens (66-69). It is the overproduction of these radicals that is suspected to be involved in the pathogenesis of IBD, but it remains unclear whether they initiate or just aggravate the tissue damage in the intestinal lumina (70). Constant exposure to radical species mutates the lining of the blood vessels and the intestinal integrity is compromised, resulting in the possibility of hemorrhagic necrosis (60). Endogenous antioxidants such as SOD, catalase, and glutathione are up-regulated to combat the excess oxidation. SOD is ubiquitous in the cytoplasm of mammalian cells, but has been shown to be deficient within the intestinal cells of patients with IBD (71-73). Approximately 1 million people in the United States are afflicted with inflammatory bowel diseases, and as recently as 1994, these diseases were considered “under studied” by the scientific community (74). Joaquim et al. (2004) have shown, through peroxidized lipid measurements, that colitis is associated with an increased burst of ROS and that treatment with SOD significantly decreased lipid peroxidation measurements, as well as decreased expression of specific adhesion molecules (75). There are several factors affecting the onset of IBD. Genetic, environmental, and microbial factors result in active inflammation of the intestinal mucosa by both immune and non-immune responses. Current therapies focus on suppressing the generalized inflammatory response within the GI tract. The beginning treatment for mild to moderate IBD involves aminosalicylates, but is found to be less effective with Crohn’s disease than with ulcerative colitis (65). Therefore, corticosteroids are often employed with patients that have moderate to severe IBD (65). However, there are serious side affects with these
9 treatments and long term remission is poor. There are a series of other supplementary medications that combat these side effects, but the toxic accumulation and limitations of the current drug therapies pushes the need for new treatment options. Because of the role that cytokines play in the active inflammation of IBD, they are a potential target for biological therapies. Selectively blocking the action of these inflammatory markers and their proposed actions or inducing protective genes, such as endogenous antioxidants, provides a possible new line of defense. We suggest that naturally occurring antioxidants or anti-inflammatory agents found in the diet could provide protection from the disease during its mild initial phases, preventing its development into the more severe and debilitating disease. 1.3. Hypothesis We propose that polymeric polyphenols that bind with protein may remain in the GI tract for over 24 hours after ingestion while continuing to exhibit anti-oxidant and anti-inflammatory activities. This may protect the GI tract from oxidative damage due to uncontrolled inflammatory responses. However, given the variable and sometimes contradictory biological activities noted for dietary tannins, it not currently feasible to make recommendations about which specific polyphenolic-rich foods would best promote gastrointestinal health. One of our long term goals is to develop methods for predicting the biological activities of tannins from knowledge of their molecular structure and chemical properties, so that we can predict both positive and negative activities that ingested tannins may manifest. We plan to screen both antioxidant and pro-oxidant activities, and anti-inflammatory activities, in a cell-based system so that we can start to develop appropriate structure-activity relationships. We propose to use peripheral blood mononuclear cells (PBMCs) to test the ability of tannins to inhibit or stimulate inflammatory cascade. A secondary goal is to asses the lifetime and activities of polyphenolic-protein adducts as well as free polyphenols. If we can generate soluble, stable polyphenol-protein adducts we will apply various methods in attempts to isolate and characterize these complexes.
10 Figure 1.1. Structures of representative tannins. (1) PGG (β-pentagalloyl glucose) as a representative hydrolyzable tannin with a core glucose group and five (G) galloyl groups linked with ester bonds. (2) Sorghum procyanidin (epicatechin16-(4 → 8)-catechin) as a representative condensed tannin containing repetitive flavanoid units. (3) EGCG (epigallocatechin gallate), also a condensed tannin, is a flavan-3-ol gallate ester.
11 1.4. Chapter 1 References
1. Swain, T. and Bate-Smith, E.C. Flavonoid compounds. In Comparative Biochemistry Florkin, M,, Mason H.S., Eds.; Academic Press: Cambridge, UK, 1962, 3, 755-809.
2. Bennick, A. Interaction of plant polyphenols with salivary proteins. Crit. Rev. Oral Biol. Med. 2002, 13, 184-196.
3. Haslam, E. Practical Polyphenolics: from Structure to Molecular Recognition and Physiological Action. Cambridge Univ. Press: Cambridge, UK, 1998, 422.
4. Hagerman, A.E. The Tannin Handbook http://www.users.muohio.edu/hagermae/. Last updated 2002, Last visited 2006.
5. Haslam, E.; Lilley, T.H.; Cai, Y.; Martin, R.; Magnolato, D. Traditional herbal medicines--the role of polyphenols. Planta Med. 1989, 55, 1-8.
6. Mitjavila, S.; Lacombe, C.; Carrera, G.; Derache, R. Tannic acid and oxidized tannic acid on the functional state of rat intestinal epithelium. J. Nutr. 1977, 107, 2113-2121.
7. Chang, M.C.; Bailey, J.W.; Collins, J.L. Dietary tannins from cowpeas and tea transiently alter apparent calcium absorption but not absorption and utilization of protein in rats. J. Nutr. 1994, 124, 283-288.
8. Riedl, K.M. and Hagerman, A.E. Tannin-protein complexes as radical scavengers and radical sinks. J. Agric. Food Chem. 2001, 49, 4917-4923.
9. Haslam, E. and Cai, Y. Plant polyphenols (vegetable tannins): gallic acid metabolism. Nat. Prod. Rep. 1994, 11, 41-66.
10. Xie, D. and Dixon, R.A. Proanthocyanidin biosynthesis--still more questions than answers? Phytochemistry 2005, 66, 2127-2144.
11. Graham, H.N. Green tea composition, consumption, and polyphenol chemistry. Prev. Med. 1992, 21, 334-350.
12. Martin-Tanguy, J.; Guillaume, J.; Kossa, A. Condensed tannins in horse bean seeds: chemical structure and effects on poultry. J. Sci. Food Agric. 1977, 28, 757-765.
13. Karchesy, J. and Hemingway, R.W. Condensed tannins (4-8, 2B-O-7)-linked procyanidins in Archis hypogea L. J. Agric. Food Chem. 1986, 34, 996-970.
14. Subramanian, V.; Butler, L.G.; Jambunathan, R.; Prasada Rao, K.E. Some agronomic and biochemical characters of brown sorghums and their possible role in bird resistance. J. Agric. Food Chem. 1983, 31, 1303-1307.
12 15. Ono, K.; Sawada, T.; Murata, Y.; Saito, E.; Iwasaki, A.; Arakawa, Y.; Kurokawa, K.; Hashimoto, Y. Pentagalloylglucose, an antisecretory component of Paeoniae radix, inhibits gastric H+, K(+)-ATPase. Clin. Chim. Acta 2000, 290, 159-167.
16. Hagerman, A.E. and Robbins, C.T. Implications of soluble tannin-protein complexes for tannin analysis and plant defense mechanisms. J. Chem. Ecol. 1987, 13, 1243-1254.
17. Chen, Y. and Hagerman, A.E. Reaction pH and protein affect the oxidation products of beta-pentagalloyl glucose. Free Radic. Res. 2005, 39, 117-124.
18. Hagerman, A.E. and Butler, L.G. Determination of protein in tannin-protein precipitates. J. Agric. Food Chem. 1980, 28, 944-947.
19. Oh, H.; Hoff, J.E.; Armstrong, G.S.; Haff, L.A. Hydrophobic interaction in tannin- protein complexes. J. Agric. Food Chem. 1980, 28, 394-398.
20. Hagerman, A.E.; Rice, M.E.; Ritchard, N.T. Mechanisms of protein precipitation of two tannins, pentagalloyl glucose and epicatechin 16 (4-8) catechin (procyanin). J. Agric. Food Chem. 1998, 46, 2590-2595.
21. Hagerman, A.E. and Butler, L.G. Protein precipitation method for the quantitative determination of tannins. J. Agric. Food Chem. 1978, 26, 809-812.
22. Hagerman, A.E. and Butler, L.G. The specificity of proanthocyanidin-protein interactions. J. Biol. Chem. 1981, 256, 4494-4497.
23. Butler, L.G.; Rogler, J.C.; Mehansho, H.; Carlson, D.M. Dietary effects of tannins. Prog. Clin. Biol. Res. 1986, 213, 141-157.
24. Rogler, J.C.; Ganduglia, H.R.R.; Elkin, R.G. Effects of nitrogen source and level on the performance of chicks and rats fed low and high tannin sorghum. Nutr. Res. 1985,
25. Clements, S.; Mehansho, H.; Carlson, D.M. Novel multigene families encoding highly repetitive peptide sequences. Sequence analyses of rat and mouse proline-rich protein cDNAs. J. Biol. Chem. 1985, 260, 13471-13477.
26. Mehansho, H.; Hagerman, A.; Clements, S.; Butler, L.; Rogler, J.; Carlson, D.M. Modulation of proline-rich protein biosynthesis in rat parotid glands by sorghums with high tannin levels. Proc. Natl. Acad. Sci. U. S. A. 1983, 80, 3948-3952.
27. Mitaru, B.N.; Reichert, R.D.; Blair, R. The binding of dietary protein by sorghum tannins in the digestive tract of pigs. J. Nutr. 1984, 114, 1787-1796.
28. Brune, M.; Rossander, L.; Hallberg, L. Iron absorption and phenolic compounds: importance of different phenolic structures. Eur. J. Clin. Nutr. 1989, 43, 547-557.
29. Ben Salem, H.; Nefzaoui, A.; Makkar, H.P.S.; Hochlef, H.; Ben Salem, I.; Ben Salem, L. Effect of early experience and adaptation period on voluntary intake, digestion,
13 and growth in Barbarine lambs given tannin-containing (Acacia cyanophylla Lindl. foliage) or tannin-free (oaten hay) diets. Anim. Feed Sci. Technol. 2005, 122, 59-77.
30. Zangerl, A.R. and Berenbaum, M.R. Plant chemistry, insect adaptations to plant chemistry, and host plant utilization patterns. Ecology 1993, 74, 47-54.
31. Jimenez-Ramsey, L.M.; Rogler, J.C.; Housley, T.L.; Butler, L.G.; Elkin, R.G. Absorption and distribution of 14C-labeled condensed tannins and related sorghum phenolics in chickens. J. Agric. Food Chem. 1994, 42, 963-967.
32. Skopec, M.M.; Hagerman, A.E.; Karasov, W.H. Do salivary proline-rich proteins counteract dietary hydrolyzable tannin in laboratory rats? J. Chem. Ecol. 2004, 30, 1679- 1692.
33. Geleijnse, J.M.; Launer, L.J.; Van der Kuip,Deirdre A M.; Hofman, A.; Witteman, J.C.M. Inverse association of tea and flavonoid intakes with incident myocardial infarction: the Rotterdam Study. Am. J. Clin. Nutr. 2002, 75, 880-886.
34. Knekt, P.; Isotupa, S.; Rissanen, H.; Heliovaara, M.; Jarvinen, R.; Hakkinen, S.; Aromaa, A.; Reunanen, A. Quercetin intake and the incidence of cerebrovascular disease. Eur. J. Clin. Nutr. 2000, 54, 415-417.
35. Graf, B.A.; Milbury, P.E.; Blumberg, J.B. Flavonols, flavones, flavanones, and human health: epidemiological evidence. J. Medic. Food 2005, 8, 281-290.
36. Hagerman, A.E.; Riedl, K.M.; Jones, G.A.; Sovik, K.N.; Ritchard, N.T.; Hartzfeld, P.W.; Riechel, T.L. High molecular weight plant polyphenolics (tannins) as biological antioxidants. J. Agric. Food Chem. 1998, 46, 1887-1892.
37. Hagerman, A.E.; Dean, R.T.; Davies, M.J. Radical chemistry of epigallocatechin gallate and its relevance to protein damage. Arch. Biochem. Biophys. 2003, 414, 115-120.
38. Mundie, T.G.; Whitener, C.; Ainsworth, S.K. Byssinosis: Release of prostaglandins, thromboxane, and 5-hydroxytryptamine in bronchopulmonary lavage fluid after inhalation of cotton dust extracts. Am. J. Pathol. 1985, 118, 128-133.
39. Lauque, D.E.; Hempel, S.L.; Schroeder, M.A.; Hyatt, R.E.; Rohrbach, M.S. Evaluation of the contribution of tannin to the acute pulmonary inflammatory response against inhaled cotton mill dust. Am. J. Pathol. 1988, 133, 163-172.
40. Kreofsky, T.J.; Russell, J.A.; Rohrbach, M.S. Inhibition of alveolar macrophage spreading and phagocytosis by cotton bract tannin. A potential mechanism in the pathogenesis of byssinosis. Am. J. Pathol. 1990, 137, 263-274.
41. Rohrbach, M.S.; Kreofsky, T.J.; Vuk-Pavlovic', Z.; Lauque, D. Cotton condensed tannin: a potent modulator of alveolar macrophage host-defense function. Basic Life Sci. 1992, 59, 803-824.
14 42. Cotes, J.E.; Steel, J.; Leathart, G.L. Work-related lung disorders. Blackwell Scientific Publishers: Philadelphia, PA, 1987, 309-319.
43. Cavalieri, E.L. and Rogan, E.G. Radical cations in aromatic hydrocarbon carcinogenesis. Free Radic. Res. Commun. 1990, 11, 77-87.
44. Cerutti, P.A. Prooxidant states and tumor promotion. Science 1985, 227, 375-381.
45. Coryell, V.H. and Stearns, D.M. Molecular analysis of hprt mutations generated in Chinese hamster ovary EM9 cells by uranyl acetate, by hydrogen peroxide, and spontaneously. Mol. Carcinog. 2006, 45, 60-72.
46. Gago-Dominguez, M. and Castelao, J.E. Lipid peroxidation and renal cell carcinoma: further supportive evidence and new mechanistic insights. Free Radic. Biol. Med. 2006, 40, 721-733.
47. Ishii, Y.; Umemura, T.; Kanki, K.; Kuroiwa, Y.; Nishikawa, A.; Ito, R.; Saito, K.; Nakazawa, H.; Hirose, M. Possible involvement of NO-mediated oxidative stress in induction of rat forestomach damage and cell proliferation by combined treatment with catechol and sodium nitrite. Arch. Biochem. Biophys. 2006, 447, 127-135.
48. Kim, H.W.; Murakami, A.; Abe, M.; Ozawa, Y.; Morimitsu, Y.; Williams, M.V.; Ohigashi, H. Suppressive effects of mioga ginger and ginger constituents on reactive oxygen and nitrogen species generation, and the expression of inducible pro- inflammatory genes in macrophages. Antioxidants & Redox Signaling 2005, 7, 1621- 1629.
49. Li, M.; Jang, J.; Surh, Y. Nitric oxide induces apoptosis via AP-1-driven upregulation of COX-2 in rat pheochromocytoma cells. Free Radic. Biol. Med. 2005, 39, 890-899.
50. Yeh, C.; Hou, M.; Tsai, S.; Lin, S.; Hsiao, J.; Huang, J.; Wang, L.; Wu, S.; Hou, L.A.; Ma, H.; Tsai, L. Superoxide anion radical, lipid peroxides and antioxidant status in the blood of patients with breast cancer. Clin. Chim. Acta 2005, 361, 104-111.
51. Toyokuni, S.; Okamoto, K.; Yodoi, J.; Hiai, H. Persistent oxidative stress in cancer. FEBS Lett. 1995, 358, 1-3.
52. Szabo, C. Alterations in nitric oxide production in various forms of circulatory shock. New Horizons, 1995, 3, 2-32.
53. Fujiki, H.; Yoshizawa, S.; Horiuchi, T.; Suganuma, M.; Yatsunami, J.; Nishiwaki, S.; Okabe, S.; Nishiwaki-Matsushima, R.; Okuda, T.; Sugimura, T. Anticarcinogenic effects of (-)-epigallocatechin gallate. Prev. Med. 1992, 21, 503-509.
54. Miyamoto, K.; Kishi, N.; Koshiura, R.; Yoshida, T.; Hatano, T.; Okuda, T. Relationship between the structures and the antitumor activities of tannins. Chem. Pharm. Bull. 1987, 35, 814-822.
15 55. Lee, S.; Lee, I.; Mar, W. Inhibition of inducible nitric oxide synthase and cyclooxygenase-2 activity by 1,2,3,4,6-penta-O-galloyl-beta-D-glucose in murine macrophage cells. Arch. Pharm. Res. 2003, 26, 832-839.
56. Chen, Y.C.; Liang, Y.C.; Lin-Shiau, S.Y.; Ho, C.T.; Lin, J.K. Inhibition of TPA- induced protein kinase C and transcription activator protein-1 binding activities by theaflavin-3,3'-digallate from black tea in NIH3T3 cells. J. Agric. Food Chem. 1999, 47, 1416-1421.
57. Liang, Y.C.; Chen, Y.C.; Lin, Y.L.; Lin-Shiau, S.Y.; Ho, C.T.; Lin, J.K. Suppression of extracellular signals and cell proliferation by the black tea polyphenol, theaflavin-3,3'- digallate. Carcinogenesis 1999, 20, 733-736.
58. Lin, Y.L. and Lin, J.K. (-)-Epigallocatechin-3-gallate blocks the induction of nitric oxide synthase by down-regulating lipopolysaccharide-induced activity of transcription factor nuclear factor-kappaB. Mol. Pharmacol. 1997, 52, 465-472.
59. Park, J.W.; Choi, Y.J.; Suh, S.I.; Kwon, T.K. Involvement of ERK and protein tyrosine phosphatase signaling pathways in EGCG-induced cyclooxygenase-2 expression in Raw 264.7 cells. Biochem. Biophys. Res. Commun. 2001, 286, 721-725.
60. Gallin, J.I.; Goldstein, I.M.; Snyderman, R. Inflammation : basic principles and clinical correlates. Raven Press: New York, N.Y, 1988, 995.
61. Nester, E.W.; Roberts, C.E.; Nester, M.T. Microbiology A human Perspective. Wm. C. Brown: Boston, MA, 1995, 333-401.
62. Gadgil, H.S.; Pabst, K.M.; Giorgianni, F.; Umstot, E.S.; Desiderio, D.M.; Beranova- Giorgianni, S.; Gerling, I.C.; Pabst, M.J. Proteome of monocytes primed with lipopolysaccharide: analysis of the abundant proteins. Proteomics 2003, 3, 1767-1780.
63. Zheng, W.; Sellers, T.A.; Doyle, T.J.; Kushi, L.H.; Potter, J.D.; Folsom, A.R. Retinol, antioxidant vitamins, and cancers of the upper digestive tract in a prospective cohort study of postmenopausal women. Am. J. Epidemiol. 1995, 142, 955-960.
64. Challem, J. The inflammation syndrome : the complete nutritional program to prevent and reverse heart disease, arthritis, diabetes, allergies and asthma. J. Wiley: Hoboken, N.J., 2003.
65. Ardizzone, S. and Bianchi Porro, G. Biologic therapy for inflammatory bowel disease. Drugs 2005, 65, 2253-2286.
66. van der Vliet, A. and Bast, A. Role of reactive oxygen species in intestinal diseases. Free Radic. Biol. Med. 1992, 12, 499-513.
67. Blau, S.; Rubinstein, A.; Bass, P.; Singaram, C.; Kohen, R. Differences in the reducing power along the rat GI tract: lower antioxidant capacity of the colon. Mol. Cell. Biochem. 1999, 194, 185-191.
16 68. Aw, T.Y. Molecular and cellular responses to oxidative stress and changes in oxidation-reduction imbalance in the intestine. Am. J. Clin. Nutr. 1999, 70, 557-565.
69. Srigiridhar, K. and Nair, K.M. Supplementation with alpha-tocopherol or a combination of alpha-tocopherol and ascorbic acid protects the gastrointestinal tract of iron-deficient rats against iron-induced oxidative damage during iron repletion. Br. J. Nutr. 2000, 84, 165-173.
70. Grisham, M.B. and Granger, D.N. Neutrophil-mediated mucosal injury. Role of reactive oxygen metabolites. Dig. Dis. Sci. 1988, 33, 6S-15S.
71. Closa, D.; Bulbena, O.; Rosello-Catafau, J.; Fernandez-Cruz, L.; Gelpi, E. Effect of prostaglandins and superoxide dismutase administration on oxygen free radical production in experimental acute pancreatitis. Inflammation 1993, 17, 563-571.
72. Epperly, M.W.; Sikora, C.A.; DeFilippi, S.J.; Gretton, J.E.; Bar-Sagi, D.; Archer, H.; Carlos, T.; Guo, H.; Greenberger, J.S. Pulmonary irradiation-induced expression of VCAM-I and ICAM-I is decreased by manganese superoxide dismutase- plasmid/liposome (MnSOD-PL) gene therapy. Biol. Blood Marrow Transplantation 2002, 8, 175-187.
73. Xia, B.; Deng, C.S.; Chen, D.J.; Zhou, Y.; Xiao, J.Q. Role of copper zinc superoxide dismutase in the short-term treatment of acetic acid-induced colitis in rats. Acta Gastroenterol. Latinoam. 1996, 26, 227-230.
74. U.S. Department of Health and Human Services Inflammatory Bowel Disease http://www.4woman.gov/faq/ibd.htm. Last updated 2005, Last visited 2006.
75. Segui, J.; Gironella, M.; Sans, M.; Granell, S.; Gil, F.; Gimeno, M.; Coronel, P.; Pique, J.M.; Panes, J. Superoxide dismutase ameliorates TNBS-induced colitis by reducing oxidative stress, adhesion molecule expression, and leukocyte recruitment into the inflamed intestine. J. Leukoc. Biol. 2004, 76, 537-544.
17 2. Characterization of soluble protein-tannin complexes using Ribonuclease A and Pentagalloyl glucose (β-1,2,3,4,6- Pentagalloyl-O-D-Glucopyranose) 2.1. Introduction Tannins bind with protein to form both soluble and insoluble complexes. It is generally accepted that this interaction is reversible, unless processes such as oxidation, changes in pH, and complex formation with metal ions takes place (1). Polyphenol, protein, and soluble complexes are in equilibrium in solution. When tannin is present in excess, soluble complexes are converted to insoluble complexes complicating treatment of the equilibrium expressions. An optimal ratio for precipitation was found for both bovine serum albumin (BSA) and gelatin, but other proteins and/or tannins may exhibit rather different behaviors (2, 3). For instance, salivary proline rich proteins bind with tannin to form insoluble complexes that do not resolublize when additional protein is added (2). It has been suggested that tannins are multivalent ligands that can crosslink with multiple protein molecules to form polymers (4, 5). The protein/polyphenol ratio and the concentrations of the starting reactants are the determining factors for stoichiometries of these complexes. Characterizing this interaction has been the focus of many studies. Logically there are a variety of interactions that could account for tannin-protein complexation including ionic, covalent, and hydrogen bonding as well as hydrophobic interactions. However, it has been shown that tannin-protein complexation does not take place at pH values where tannin would be ionized, suggesting that ionic interactions are not important (6). Hagerman et al. (1978) showed strong evidence supporting hydrogen bonding interactions as the predominant force, but could not exclude hydrophobic forces as a participant (6). Several other studies support hydrophobic interactions between tannin and protein (7, 8). Hagerman et al. (1998) found that PGG precipitated BSA by forming a hydrophobic coat around the protein (9). Conversely, when purified condensed tannin was added to bovine serum albumin it was shown that hydrogen bonding was a major factor in complex formation (9).
18 There is evidence that oxidized forms of tannin interact with protein very differently than native, reduced forms of tannin. When the model protein BSA was added to oxidized PGG, complexes were formed that were resistant to dissociation by SDS detergent. Conversely when complexes were formed with unoxidized PGG and BSA, SDS completely dissociated the complexes. This suggests that covalent stabilization is involved with complex formation in the presence of oxidized PGG, as might be found within the GI tract (10). Our goal with this project was to find, isolate, and characterize soluble covalently linked tannin-protein complexes, for later use in the PBMC inflammatory assay. To do this we choose the model protein ribonuclease A (RNase). The small size of this protein was advantageous for our experiments because it should allow us to see small shifts in molecular weight with electrophoresis or HPLC. We chose to use sodium periodate
(NaIO4) as an oxidant because it had been shown previously in our lab to be an efficient oxidant for PGG (10). In addition, it has been reported that sodium periodate does not oxidize proteins so we predicted it would not affect the RNase component of the reaction mixtures (11) 2.2. Materials and Methods RNase – Type 1-A from bovine pancreas (Sigma R-4875) was dissolved in 10 mM phosphate buffer containing 9.5 mM NaCl, pH 7.4. The PGG, obtained by methanolysis from tannic acid (12) or synthesized using 14C glucose (13), was dissolved in nanopure water and the concentration was measured spectrophotometrically at 280 nm, using the extinction coefficient 5.4x104 M-1cm-1(14). PGG solutions were prepared fresh daily. RNase was combined with PGG in mass ratios ranging from 10:1 to 100:1 [RNase:PGG], and the mixture was vortexed and incubated for 15 min at room temperature. When oxidation was desired, fresh sodium periodate was added to achieve a final concentration of 5.9 mM periodate, and the solutions were allowed to sit for an additional 2-3 min. In general, reactions were adjusted so that the final volume of the mixture with all reactants was 38 µL. Each 38 µL mixture was treated with the appropriate SDS or native sample prep solution before loading about 17 µg protein per lane. The samples were run on 12% native acrylamide gels, Laemmli SDS-PAGE gels
19 with various acrylamide concentrations, or 13% acrylamide Tris/tricine gels (15, 16). The gels were stained with Coomassie blue or were electroblotted to nitrocellulose paper for staining with Ponceau S for protein determination, or with nitroblue tetrazolium (NBT) in glycine at pH 10 (17) or Prussian blue in an attempt to detect PGG. Some gels or blots were imaged with a phosphorimager to detect radioactivity. For proteolysis, we adapted the general outline set forth by Kedishvili et al. (1991) for proteolysis with trypsin (T-1005, Sigma) (18). The enzyme:protein ratio was 1:100 as recommended. In a subsequent experiment, complexes were digested with subtilisin (P-5380, Sigma) before being exposed to typsin. The enzyme:protein ratio was 1:10 mass ratio for both trypsin and subtilisin. 2.3. Results and Discussion Under non-oxidizing conditions, proteins with basic isoelectric points interact best with tannins at more basic pH values. RNase, with a pI of 9.6 (19), was expected to react best with PGG in slightly basic solutions. To test that idea, reaction mixtures were prepared in buffers at pH 2, 4, or 7.4, which correspond to biological pH throughout the GI tract (10). Because we were interested in covalently stabilized complexes, the protein/PGG mixture was treated with the oxidizing agent sodium periodate after incubating the protein with the PGG at the various pH values. The gel shows that there is virtually no difference between RNase without PGG and with PGG at pH values 2 and 4 (Fig. 2.1, Lanes 2-5). However, there is a striking difference in the pattern when RNase is incubated with PGG at pH 7.4. The main protein band disappears and is replaced by a streak, suggesting formation of aggregates (Fig. 2.1, Lane 7). We concluded that pH 7.4 is the optimal pH value to use for characterizing soluble RNase-PGG complexes. When protein and tannin are mixed under non-oxidizing conditions, the complexes that form are usually insoluble if tannin is in excess, and soluble if protein is in excess (3). When we oxidized RNase-PGG complexes with periodate, we saw streaking in the gel, indicating aggregation and complexation, when the RNase:PGG ratio was close to 10:1 (Fig. 2.2, lane 4). Increasing the concentration of PGG by ten times yielded complexes that did not enter the gel, suggesting the formation of insoluble complexes (Fig. 2.2, lane 3). Decreasing the concentration of PGG by ten times, resulted in no detectable complexation (Fig. 2.2, lanes 5 & 6). No complexes were observed
20 when RNase was incubated with PGG in the absence of NaIO4 (Fig. 2.2, lanes 7-9). Because Coomassie blue staining is not specific for protein, also detecting other large hydrophobic materials, we confirmed the results with the gel by transferring samples to nitrocellulose after electrophoresis, and staining with Ponceau S or NBT. The blot stained with Ponceau S confirmed the presence of protein located in the same spots as found with the Coomassie blue staining on the gel (data not shown). The blot with NBT showed the presence of tannin within the streaks that were detected at the 10:1 mass ratio, suggesting the complexes contained tannin as well as protein (data not shown). There was no clear evidence of tannin within the blot for any other mass ratio represented. When a narrower range of concentration ratios was tested, a concentration gradient was observed (Fig. 2.3). At favorable ratios complexes are visible as streaks in the gel (Fig. 2.3 lanes 5-9) and as bands on the edges of the sample wells (not shown). Higher molecular weight complexes apparently formed at higher concentrations of PGG, since neither free protein nor complexes were visible for those samples (Fig. 2.3 lanes 1- 4)
RNase, PGG, and NaIO4 concentrations all had to be increased with SDS-PAGE gels for the complexes to be visible, but on SDS gels there were distinctive horizontal bands that suggested specific complexes rather than aggregates (Fig. 2.4). The expected band for periodate-treated RNase without tannin appeared at around 14kD. However, when tannin was added to the mixture, several additional bands of higher molecular weight were visible, as well as streaking similar to that noted with native gels. Complexes were visible at the top of the resolving gel and within the stacking gel (Fig. 2.4). More of the high molecular weight complex formed at higher PGG concentrations. There are three visible bands in the samples containing PGG and RNase. There is a high mobility band that apparently contains only protein as it is present in all RNase- containing samples. However, in lanes containing protein and tannin there are two additional bands. Those bands are enhanced in the 17% acrylamide gels (Fig. 2.5). In addition, those lanes have streaking and material that remains in the sample wells (Fig. 2.5). In an effort to improve resolution of the complexes, we ran samples on gradient SDS PAGE. The 10-20% gradient gel displayed a very narrow band for RNase without
21 PGG and faint diffuse bands for the complexes that were formed (Fig.2.6A). A shorter gradient with 15-20% acrylamide failed to discriminate higher molecular weight bands any better than the 17% gel (Fig. 2.6B). Dialysis was used to try to remove excess periodate oxidant, and excess free PGG from reaction mixtures. All of the dialyzed samples (Fig. 2.7, Lanes 1-4) have a sharper band for RNase compared with samples that were not dialyzed (Fig. 2.7, Lanes 6-8). Distinct bands with lower mobility are found for the sample containing RNase, PGG, and periodate that was dialyzed overnight (Fig. 2.7 lane 4). However, these samples also had a large amount of insoluble material at the top of the stacking and resolving gels. Less complex formed in samples that were incubated briefly with periodate (Fig. 2.7, lane 8),
or in samples that contained RNase and PGG that were dialyzed before NaIO4 was added (Fig. 2.7, lane 3). It seemed likely that the prolonged incubation time, rather than dialysis, enhanced formation of complexes between PGG and RNAse in an oxidizing environment. To determine whether prolonged oxidation of the system by excess periodate was causing insoluble complexes to form at the expense of soluble PGG-RNase complexes, we used β-mercaptoethanol (β-ME) to reduce unreacted periodate. The β-ME was added after 2 min of oxidation by periodate. Similar low mobility bands are found in the RNase/PGG/periodiate samples in the presence or absence of β-ME (Fig. 2.8). Surprisingly, similar bands were noted in samples containing RNase, PGG and periodate (Fig. 2.8 lane 7) or β-ME (Fig. 2.8 lane 9), but that result was not confirmed in a subsequent experiment (Fig. 2.9 lane 7 vs. lane 10). We did show that addition of β-ME to a mixture of RNase and PGG before addition of periodate prevented the oxidation of complexes to low mobility forms, suggesting that periodate oxidation is an important step in formation of stable complexes (Fig. 2.9 lane 8 vs. lane 9). β-ME is a protein denaturant that acts by reducing disulfides. To determine if protein denaturation by unfolding but not altering redox state of the protein influenced formation of complexes, we added urea to reaction mixtures with RNase, PGG and periodate (Fig. 2.10). We found that urea has little effect on complex formation between PGG and RNase.
22 Our electrophoresis results suggested that periodate oxidation of RNase-PGG mixtures yielded small amounts of higher molecular weight complexes. In order to evaluate sites of reaction between PGG and RNase, we attempted to proteolytically cleave the complexes. Our goal was to compare the peptide products of cleavage of native and modified RNase, and identify site of reaction on the protein. There was no evidence that trypsin cleaved RNase or RNase incubated with PGG (Fig. 2.11). This result is consistent with the resistant nature of RNase reported in the literature (19). However, RNase-PGG complexes formed by periodate oxidation may be partially cleaved by the trypsin treatment, with loss of the main band and increased intensity of higher mobility bands (Fig. 2.11). Exposing RNase as well as the complexes to subtilisin before digestion with typsin resulted in efficient proteolysis (Fig. 2.12). Subtilisin cuts RNase at a single specific site to yield a larger fragment known as RNase S and a small peptide (19). RNase S is more susceptible to trypsin than is native RNase, and several peptide products were obtained. However we did not have sufficient resolution to determine if there were unique peptide products from the RNase-PGG complexes (Fig. 2.12). Tris/Tricine gels often provide better resolution of peptides than the Laemmli SDS PAGE gels that we routinely use. Proteolysis yielded a mixture of peptide products from RNase or RNase plus PGG under non-oxidizing conditions (Fig. 2.13). The low mobility complex found in periodate treated RNase/PGG mixtures is partly resistant to proteolysis (Fig. 2.13). Similar results were obtained when radiolabeled PGG was used and the reaction products were imaged with the phosphorimager (Fig. 2.14). Only the lanes containing RNase, PGG and periodate (Fig. 2.14 lanes 5 & 6, lanes 9 & 10) were visible on the image of the radioactive gel. The isotope was found in the stacking gel and at the interface between stacking resolving gels, and more faintly as a low mobility smear analogous to the smear seen in Coomassie stained PAGE gels. There was no evidence for proteolytic cleavage of the radioactive PGG-RNase containing complex with subtilisin treatment. These proteolysis experiments provide evidence of covalent interaction, as we predicted, but do not provide evidence for the position of interaction or type of bond formed.
23 The RNase control samples run on Laemmli gels had a major band with the expected molecular weight (13.7 kD) of RNase and a minor band with slightly lower mobility (e.g. Fig. 2.9). That lower mobility band, which had an apparent molecular weight of 33.7 kD, was the predominant species in the RNase controls (no PGG, no oxidizing agent) run on Tris/Tricine gels (Fig. 2.13). It has been noted that RNase aggregates spontaneously during storage or purification. Those aggregates can sometimes be dissociated by gentle heating (19). Unfortunately, we found that the contaminants in our commercial preparation of RNase could not be eliminated by heating the protein for 15 min (Fig. 2.15). Surprisingly, we also found that heating in the presence of periodate results in a series of evenly spaced, lower mobility bands (Fig. 2.15). Based on mobility, the bands generated by periodate and heat, were 2.4, 4.4, 6.6 and 8.6 times larger than native RNase, suggesting chemical modification and/or crosslinking to lower mobility forms. This evidence that our oxidizing agent, sodium periodate, significantly reacted with the target protein even in the absence of our polyphenol, PGG, made it impossible to successfully conclude this portion of the work. In summary, we obtained electrophoretic evidence for formation of high molecular weight complexes between the protein RNase and the polyphenol PGG in the presence of the oxidizing agent sodium periodate. These complexes were stable under native and SDS PAGE conditions, suggesting that they contained covalent bonds. When the oxidizing agent was omitted from the PGG-RNase mixture, the bands that we tentatively identified as complexes were not visible on the gels. Addition of a reducing agent, such as β-ME, to periodate before the reaction with protein and polyphenol prevented the reaction, but addition of the reducing agent to the already formed complex did not alter it. The complex may have been more stable to proteolysis than the uncomplexed protein although the results were difficult to interpret due to formation of aggregates of the protein. In future work, we suggest exploration of more proteins and polyphenols. Proteins that are unreactive with periodate, and proteins that do not have strong tendency to spontaneously aggregate, would be preferable substrates. It is possible that polyphenols that oxidize more readily than PGG would be better reactants. The
24 electrophoretic method of detection of the complexes could perhaps be augmented by HPLC using a size exclusion column to detect complexes.
25 Figure 2.1. Optimal pH for PGG-RNase complex formation. PGG (1.6 µg) was incubated with 16 µg RNase A for 15 min at pH 2.0, 4.0 or 7.4. Sodium periodate was added to the mixtures to achieve a final concentration of 2.95 mM and the samples were incubated for about 2 min. The samples were run on 12% PAGE. Lanes contain RNase alone at pH 2.0 (1); RNase + NaIO4 at pH 2.0 (2), pH 4 (4), or pH 7.4 (6); RNase + PGG
+ NaIO4 at pH 2.0 (3), pH 4 (5), or pH 7.4 (7). (Page 42, Book 1).
26 Figure 2.2. Optimal RNase:PGG ratio for complex formation. RNase (16 µg) was incubated with various amounts of PGG for 15 min before adding periodate to some samples to achieve a final concentration of 2.95 mM in a 38 µl final reaction volume. The samples were run on a native gel (shown below). Lanes contain RNAse (1); RNase
+ NaIO4 (2), RNase + NaIO4 with 16 µg PGG (3), 1.6 µg PGG (4), 0.16 µg PGG (5), and
0.016 µg (6); RNase without NaIO4 with 1.6 µg PGG (7), 0.16 µg PGG (8), and 0.016 µg PGG (9). The arrow indicates the smeared aggregates visible when the mass ratio of RNase to PGG is 10:1. (Page 45, Book 1).
27 Figure 2.3. Effect of excess PGG on complex formation. RNase (16 µg) was incubated with various amounts of PGG for 15 min before adding periodate to the samples to achieve a final concentration of 2.95 mM. The samples were run on a native gel. Mass ratios of RNase:PGG were 5:1 (1), 6:1 (2), 7:1 (3), 8:1 (4), 10:1 (5), 13:1 (6), 20:1 (7), 40:1 (8), 100:1 (9). (Page 56, Book 1).
28 Figure 2.4. SDS PAGE to detect PGG-RNase complexes. RNase (48 µg) was incubated with PGG (8.4 - 1.2 µg) for 15 min before adding NaIO4 to achieve a final concentration of 8.9 mM. Samples were run on 15% SDS PAGE. Lanes contain RNase with no added PGG (1), RNase with 8.4 µg PGG (2), with 7.1 µg PGG (3), with 5.8 µg PGG (4), with 4.7 µg PGG (5), with 3.5 µg PGG (6), with 2.4 µg PGG (7), and RNase with 1.2 µg PGG (8). Lane (9) contained a molecular weight marker. The arrow indicates the position of the high molecular weight aggregates that do not enter the gel. (Page 82, Book 1).
29 Figure 2.5. Higher percent acrylamide SDS PAGE to better resolve PGG-RNase complexes. RNase (32 µg) was incubated with PGG (5.2 – 0.7 µg) for 15 min before adding periodate to achieve a final concentration of 5.9 mM. Samples were run on a 17% acrylamide gel. The reaction samples including oxidized RNase (1), oxidized PGG (2), and RNase with 5.2 µg PGG (3), RNase with 3.6µg PGG (4), RNase with 2.9 µg PGG (5), RNase with 2.2 µg PGG (6), RNase with 1.5 µg PGG (7), RNase with 0.7 µg PGG (8). Molecular weight marker is in lane (9). (Page 2, Book 2)
30 Figure 2.6. Gradient SDS-PAGE to detect PGG-RNase complexes. RNase (32 µg) was incubated with various amounts of PGG for 15 min before adding periodate to all samples. The samples were run on a 10-20% acrylamide (A) or a 15-20% acrylamide (B) gradient SDS gel. Lanes contain RNase (1), 5.2 µg PGG (2), and RNase with 5.2 µg PGG (3), RNase with 3.6 µg PGG (4), RNase with 2.92 µg PGG (5), RNase with 2.2 µg PGG (6), RNase with 1.5 µg PGG (7), and RNase with 0.73 µg PGG (8), with a molecular weight marker (9). Lane 5 gel A was not loaded correctly. (Page 18&20, Book 2)
A.
B.
31 Figure 2.7. Effect of dialysis on PGG-RNase complexes. RNase (3.8 mg) was incubated with PGG (0.53 mg) in 4.7 mL buffer for 15 min. Periodate was added, to achieve a final concentration of 5.9 mM, to some samples before dialysis and to some other samples after dialysis. The samples were placed in 12 kD MWCO dialysis tubing and dialyzed overnight at 4oC against pH7.4 buffer. The samples were run on SDS PAGE. Lanes contain dialyzed samples of RNase (1); RNase + PGG (2); RNase + PGG with
NaIO4 added after dialysis (3); RNase + PGG with NaIO4 added before dialysis (4). The remaining samples were not dialyzed, and consist of RNase (6), RNase + PGG (7), and
RNase + PGG + NaIO4 (8). Lanes 5 and 9 were empty and lane 10 contained the colored marker. (Page 67, Book 2).
32 Figure 2.8. Effect of β-ME on PGG-RNase complexes. RNase (32 µg) was incubated with various amounts of PGG for 15 min before adding periodate to achieve a final concentration of 5.9 mM. β-ME was diluted with water (1:1, v:v) and 4 µL was added 2
min after the addition of NaIO4. After 2 min, the samples were run on 17% SDS-PAGE
gels. Lanes contain RNase and NaIO4 without (1) or with (2) β-ME; PGG and NaIO4 without (3), or with (4) β-ME; RNase, PGG (6 µg) and NaIO4 without (5) or with (6) β-
ME; RNase, PGG (3.4 µg) and NaIO4 without (7) or with (8) β-ME; RNase, PGG (3.4 µg) and β-ME (9), and colored molecular weight markers (10). (Page 52, Book 2).
33 Figure 2.9. Order of addition of reagents. RNase (32 µg) was incubated with 3.4 µg PGG for 15 min before adding periodate followed or preceded by β-ME (4 µL of 1:1 dilution). The samples were run on 17% SDS-PAGE gels. Protein lanes (1-5) contain
RNase (2) or RNase with NaIO4 (1), with β-ME (3), with NaIO4 followed by β-ME (4), or with β-ME followed by NaIO4 (5). PGG lanes include RNase with PGG (6), or RNase with PGG followed by β-ME (7), followed first by NaIO4 and then by β-ME (8), followed first by β-ME and then by NaIO4 (9), or followed by NaIO4 (10). The main RNase band (13.7 kD) is indicated with an arrow on the left, and lower mobility bands representing complex are indicated by arrows on the right. (Page 53, Book 2).
34 Figure 2.10. Effect of urea on PGG-RNase complex formation. RNase (32 µg) was incubated with 3.4 µg PGG for 15 min before adding periodate followed by 30 µL of 8 M urea and β-ME (4 µL of 1:1 dilution). The samples were run on 17% SDS-PAGE gels.
Lanes contain RNase with NaIO4 followed by urea (1) or urea and β-ME (4); RNase with
urea (2) followed by β-ME (3), or followed by NaIO4 and β-ME (5); RNase with PGG
followed by urea (6), and then by β-ME (7), or NaIO4 and β-ME (9); RNase with PGG
followed by NaIO4 and then by urea (10) or by both urea and β-ME (8). (Page 59, Book 2).
35 Figure 2.11. Proteolysis with trypsin. RNase (96 µg) was incubated for 15 min with 10.4 µg of PGG in 114 µL buffer. Periodate was added to some samples, followed after 2 min by DTT to achieve a final concentration of 0.1 mM. Trypsin (0.3 µg) was added immediately after the DTT, resulting in a final reaction mixture of 125 µL. Aliquots of 40 µL of each of the reaction mixtures were removed at 0, 60, and 120 min and stopped with 25 µL of SDS prep solution containing β-ME. The samples were run on a 17% SDS PAGE gel. Lanes contain samples taken at the indicated times from reactions containing RNase at 0 min (1), 60 min (2), and 120 min (3); RNase with PGG at 0 min (4), 60 min
(5), and 120 min (6); and RNase with PGG and NaIO4 at 0 min (7), 60 min (8), and 120 min (9), and colored molecular weight marker (10). (Page 73, Book 2).
36 Figure 2.12. Proteolysis with subtilisin and trypsin. RNase (96 µg) was incubated for 15 min with 10.4 µg of PGG in 114 µL buffer. Periodate was added to some samples, followed after 2 min by DTT to achieve a final concentration of 0.1 mM. A 40 µL aliquot was removed from of each sample and 25 µL of SDS prep solution containing β- ME was added before boiling for 5 min. Immediately after the removal of the first aliquot, 3.2 µg subtilisin was added to the remaining 74 µL of each reaction mixture. After 20 min at room temperature a second aliquot was removed at and prepared for electrophoresis as above. Trypsin (3.2 µg) was then added and after 20 min incubation, the last aliquot was removed for treatment as above. The samples were run on a 17% SDS PAGE gel. Lanes contain RNase before treatment (1), after subtilisin (2), after trypsin (3); RNase with PGG before treatment (4), after subtilisin (5), after trypsin (6);
RNase with PGG and NaIO4 before treatment (7), after subtilisin (8), after trypsin (9), and colored molecular weight marker (10). (Page 80, Book 2).
37 Figure 2.13. Proteolysis and Tris/Tricine gels. RNase (96 µg) was incubated for 15 min with 10.4 µg of PGG in 114 µL buffer. Periodate was added to some samples, followed after 2 min by DTT to achieve a final concentration of 0.1 mM. A 40 µL aliquot was removed from of each sample and 25 µL of SDS prep solution containing β-ME was added before boiling for 5 min. Immediately after the removal of the first aliquot, 32 µg subtilisin was added to the remaining 74 µL of each reaction mixture. After 20 min at room temperature a second aliquot was removed at and prepared for electrophoresis as above. Trypsin (32 µg) was then added and after 20 min incubation, the last aliquot was removed for treatment as above. The samples were run on a Tris/tricine gel. Lanes contain RNase before treatment (1), after subtilisin (2), after trypsin (3); RNase with
PGG after subtilisin (4), after trypsin (5); RNase with PGG and NaIO4 before treatment (6), after subtilisin (7), after trypsin (8). The 13.7 kD RNase band and the 33.7 kD aggregates are indicated by arrows on the left. (Page 86, Book 2; molecular weights from Page 3, Book 3).
38 Figure 2.14. Proteolysis with 14C radiolabelled PGG using subtilisin. RNase (128 µg) was incubated with 6.3 µg of radiolabelled PGG (0.27 µCi) for 15 min before adding periodate to some samples. Some samples were quenched with β-ME before the addition of subtilisin (0.8 µg) to result in a total solution volume of 100 µL. A 45 µL aliquot was immediately removed and stopped with 13 µL of stopping solution and then heated for 10 min at 95°C. After cooling, 3 µL of 0.1 M NaOH was added to neutralize the solution before the addition of 10 µL of sample prep solution containing β-ME and then heating again for 10 min at 95°C. Lanes contained RNase exposed to subtilisin for 0 min (1) and 60 min (2), RNase quenched with β-ME, then exposed to subtilisin for 0 min (3) and 60 min (4), RNase with radiolabelled PGG, with the addition of periodate before exposure to subtilisin for 0 min (5) and 60 min (6), RNase with the addition of periodate before exposure to subtilisin for 0 min (7) and 60 min (8), and RNase with radiolabelled PGG, with the addition of periodate and then quenched with β-ME, before exposure to subtilisin for 0 min (9) and 60 min (10). Arrows on the right indicate the bottom of the sample well, the top of the resolving gel, and the poorly resolved low mobility material. (Page 38, Book 3)
39 Figure 2.15. Periodate reaction with RNase. RNase (64 µg) was dissolved in water or pH 7.4 buffer and periodate was added before samples were heated at 60oC for 15 min and run on Tris/tricine gel. Lanes contain samples that were not heated (lanes 1-4) or samples that were heated (lanes 5-8). Lanes contain RNase dissolved in water (1 and 5) with added periodate (3 and 7); RNase dissolved in buffer (2 and 6), with added periodate (4 and 8). (P.99, Book 2).
40 2.4. Chapter 2 References
1. Bennick, A. Interaction of plant polyphenols with salivary proteins. Crit. Rev. Oral Biol. Med. 2002, 13, 184-196.
2. Luck, G.; Liao, H.; Murray, N.J.; Grimmer, H.R.; Warminski, E.E.; Williamson, M.P.; Lilley, T.H.; Haslam, E. Polyphenols, astringency and proline-rich proteins. Phytochemistry 1994, 37, 357-371.
3. Hagerman, A.E. and Robbins, C.T. Implications of soluble tannin-protein complexes for tannin analysis and plant defense mechanisms. J. Chem. Ecol. 1987, 13, 1243-1254.
4. Chen, Y. and Hagerman, A.E. Quantitative examination of oxidized polyphenol- protein complexes. J. Agric. Food Chem. 2004, 52, 6061-6067.
5. McManus, J.P.; Davis, K.G.; Beart, J.E.; Gaffney, S.H.; Lilley, T.H.; Haslam, E. Polyphenol interactions. Part 1. Introduction: some observations on the reversible complexation of polyphenols with proteins and polysaccharides. J. Chem. Soc. Perk. Trans. 2 1985, 1429-1438.
6. Hagerman, A.E. and Butler, L.G. Protein precipitation method for the quantitative determination of tannins. J. Agric. Food Chem. 1978, 26, 809-812.
7. Oh, H.; Hoff, J.E.; Armstrong, G.S.; Haff, L.A. Hydrophobic Interaction in Tannin- Protein Complexes. J. Agric. Food Chem. 1980, 28, 394-398.
8. Murray, N.J.; Williamson, M.P.; Lilley, T.H.; Haslam, E. Study of the interaction between salivary proline-rich proteins and a polyphenol by 1H-NMR spectroscopy. Eur. J. Biochem. 1994, 219, 923-935.
9. Hagerman, A.E.; Rice, M.E.; Ritchard, N.T. Mechanisms of protein precipitation of two tannins, pentagalloyl glucose and epicatechin 16(4-8) catechin (procyanin). J. Agric. Food Chem. 1998, 46, 2590-2595.
10. Chen, Y. and Hagerman, A.E. Reaction pH and protein affect the oxidation products of beta-pentagalloyl glucose. Free Radic. Res. 2005, 39, 117-124.
11. Burdine, L.; Gillette, T.G.; Lin, H.; Kodadek, T. Periodate-triggered cross-linking of DOPA-containing peptide-protein complexes. J. Am. Chem. Soc. 2004, 126, 11442- 11443.
12. Chen, Y. and Hagerman, A.E. Characterization of soluble non-covalent complexes between bovine serum albumin and beta-1,2,3,4,6-penta-O-galloyl-D-glucopyranose by MALDI-TOF MS. J. Agric. Food Chem. 2004, 52, 4008-4011.
41 13. Chen, Y.; Hagerman, A.E.; Minto, R.E. Preparation of 1,2,3,4,6-Penta-O-galloyl-[U- 14 C]-D-glucopyranose. J. Label Compound Radiopharm. 2003, 46, 99-105.
14. Chen, Y. and Hagerman, A.E. Characterizing soluble non-covalent complexes between bovine serum albumin and β-1,2,3,4,6-penta-O-galloyl-D-glucopyranose by MALDI-TOF mass spectrometry. J. Agric. Food Chem. 2004, 52, 4008-4011.
15. Baker Laboratory Tris-Tricine Protein/Peptide Separation Gels http://depts.washington.edu/~bakerpg/protocols/tris_tricine_gels.html. Last updated 1999, Last visited 2006.
16. The Protein Purification Facility Tricine Gels Protocol. http://www.Is.huji.ac.il/~purification/Protocols/Tricine.html Last updated 2005,Last visited 2005.
17. Paz, M.A.; Fluckiger, R.; Boak, A.; Kagan, H.M.; Gallop, P.M. Specific detection of quinoproteins by redox-cycling staining. J. Biol. Chem. 1991, 266, 689-692.
18. Kedishvili, N.Y.; Popov, K.M.; Harris, R.A. The effect of ligand binding on the proteolytic pattern of methylmalonate semialdehyde dehydrogenase. Arch. Biochem. Biophys. 1991, 290, 21-26.
19. Boyer, P.D.; Krebs, E.G.; Sigman, D.S. The Enzymes. 3rd edition; Academic Press: New York, NY 1970.
42 3. Analytical methods for monitoring the inflammatory response
3.1. Introduction
3.1.1. Superoxide as one of the first signaling molecules One of the first events in the inflammatory cascade is the overproduction of reactive oxygen species (ROS), an event referred to as the respiratory burst. ROS represent all metabolites of oxygen as well as some other radicals, including superoxide •- • anion (O2 ), hydroxyl radical (OH ), hydrogen peroxide (H2O2), and hypochlorous acid (HOCl). There are several pathways that result in the production of ROS. For example, ROS are released when immune complexes bind to phagocytes, reactions that are crucial for the elimination of invading pathogens (1). However, because ROS perpetuate the inflammatory response by recruiting more phagocytic leucocytes to the area, they can also lead to oxidative damage of the tissue surrounding the inflamed area. The enzyme NADPH-oxidase univalently reduces molecular oxygen to superoxide anion during the respiratory burst. The NADPH-oxidase system remains unassembled and dormant in resting, or unstimulated cells. When cells become inflamed or are stimulated, soluble components of the cell, known as cytosolic factors, are relocated to the plasma membrane of the cells and the enzyme is assembled (2). The intact system produces superoxide as follows: •- + + NADPH + 2O2 → 2 O2 + NADP + H (Eq.1) Superoxide anion itself is not highly reactive, but the reaction products of superoxide including hydroxyl radical result in peroxidation of lipids (3), oxidation of proteins and enzymes (4), and base modifications and chain breakage in nucleic acids (5, 6). Presumably these chemical oxidations lead to the association between overproduction of superoxide and several human diseases and aging (7, 8). In an effort to develop an analytical ex vivo assay for determining the effects of tannins on the inflammatory response, we chose to monitor the initial respiratory burst. We hoped that measuring superoxide directly would provide us with a convenient marker for inflammatory activity of cells after treatment with various tannins. Cytochrome c
43 provides a convenient analytical tool for measuring superoxide, since cytochrome undergoes a characteristic spectral shift upon reduction by superoxide or other reducing agents.
3.1.2. Rationale for using cytochrome c Cytochrome c is a heme protein responsible for the transfer of electrons to oxygen via cytochrome c oxidase in the electron transport chain. The centrally located iron carries the electrons in the following electron transfer reaction: Fe3+ + e- → Fe2+ (Eq.2) The heme group absorbs strongly in the visible range of the electromagnetic spectrum resulting in the intense color of cytochrome c. When oxidized cytochrome c accepts electrons from a reducing agent, its color changes from orange to pink, signifying a change from oxidized to reduced cytochrome c. The color change is readily detected spectrophotometrically (9). In addition, cytochrome c is non-toxic to the cells and doesn’t interfere with their normal functioning. It is inexpensive and readily available. All these features make reduction of cytochrome c an attractive quantitative probe for superoxide production by cells (10-12). In order to validate the superoxide assay, we used chemical reductants to quantitate the spectral changes that accompany reduction of cytochrome c. Cytochrome c
can be reduced by sodium dithionite (Na2S2O4), via two different mechanisms. The first 2- mechanism involves S2O4 as the reducing species: 2- 3+ 2+ - S2O4 + Fe → Fe + S2O4 (Eq.3) - 3+ 2+ S2O4 + Fe → Fe + S2O4 (Eq.4) •- The second mechanism involves SO2 as the reducing species:
•- S2O4 ≡ 2SO2 (Eq.5)
•- 3+ 2+ SO2 + Fe → Fe + SO2 (Eq.6) 2- Lambeth and Palmer (2002) presented evidence that suggests that both S2O4 and •- SO2 act as the reducing species for the reduction of ferricytochrome c with excess dithionite (9). Reduction of cytochrome c by superoxide was confirmed by using potassium superoxide (KO2) as a superoxide source:
44 3+ •- 2+ Fe + O2 → Fe + O2 (Eq.7)
To use KO2 as a reagent, it must be dissolved in an aprotic solvent such as DMSO
(dimethyl sulfoxide). In protic solvents, at slightly acidic pH, the reaction of KO2 to yield hydrogen peroxide is rapid: + •- 2H + 2O2 → H2O2 + O2 (Eq.8) Hydrogen peroxide is unable to reduce cytochrome c, so it is important to have the stock solution of un-reacted potassium superoxide that can be added to reaction
mixtures that already contain cytochrome c. We used suspensions of KO2 in DMSO for our in vitro experiments. Our initial experiments were focused on establishing controlled methods for determining superoxide production using cytochrome c. We then extended those results to measuring superoxide production in rabbit monocytes. 3.2. Methods 3.2.1. Reagents Oxidized (Fe3+) horse heart cytochrome c (Sigma C-2506) was dissolved in autoclaved PBS buffer pH 7.8 at a concentration of 10 mg/mL. It was then aliquoted into microcentrifuge tubes for individual use and stored at -20°C. Potassium superoxide (Aldrich, 278904) was mixed with DMSO (Pharmco, CAS# 67-68-5) and sonicated for approximately 5 min to obtain a suspension with an approximate concentration of 0.5 mg/mL. Dithionite was dissolved in water in a concentrated solution. 3.2.2. Validation Up to five aliquots (10 µL each) of concentrated dithionite were added to 0.6 mg of cytochrome c dissolved in 1000 µL of buffer. The reaction mixture was mixed by inversion between each addition of dithionite within a one milliliter cuvette. The visible absorbance spectrum was recorded after each aliquot was added.
KO2 was added to wells of a 24 well plate containing 300 µg cytochrome c and RPMI 1640 media in a final volume of 500 µL. It was dispensed in increments ranging from 0 to 125 µL. The spectra were recorded immediately with the microplate reader. 3.3. Results and Discussion Cytochrome c (Fe3+) is reduced with an accompanying change in absorbance centered at 550 nm (Fig. 3.1). Chemical reducing agents such as dithionite or superoxide
45 yield clean spectral data, which can be used in conjunction with the extinction coefficient 3 -1 -1 for cytochrome c to quantitate the reduction (∆E550=21x10 M cm ) (13). Preliminary experiments with cell cultures indicated that there are spectral changes in the cell background, which make it difficult to measure superoxide production simply by monitoring the change in absorbance at 550 nm. Frequently, cells that were not stimulated to produce superoxide had higher absorbance at 550 nm than cells that produced superoxide (Fig. 3.2). Examination of the spectra demonstrated that the change in shape of the cytochrome c spectrum in the region of the 550 nm peak was quite distinct in the superoxide-producing cells, despite the high backgrounds characteristic of the control cells. Spectral shape can be quantitated by measurements of slope, so we used dithionite to incrementally reduce a sample of cytochrome c, and examined the slope of the peak in the 545-550 nm region of the spectrum. The slope of the absorbance spectrum between 544 nm and 546 nm was sensitive to reduction. We obtained a linear relationship between slope and the amount of cytochrome c reduced, monitored directly at 550 nm (Fig. 3.3). There is a 1:1 stoichiometric relationship between moles cytochrome c reduced and moles superoxide scavenged (Eq.7), so in cell experiments we used the slope measurement at 544-546 nm to determine amount of superoxide produced (slope = 0.0031 nmol superoxide - 0.0008) (Fig. 3.3). With this method, we could detect as little as 1 nmol superoxide.
We used KO2 to confirm that superoxide was an appropriate reducing agent for cytochrome c. We did record the characteristic shift in cytochrome c spectrum when
KO2 was added (data not shown). One shortcoming in our experiments was the limited
solubility of KO2 in DMSO, making us work with a suspension rather than a solution of known concentration. Valentine and Curtis (14) suggest using 0.30M 18-crown-6 ether
in DMSO as a stable solvent for 0.15M KO2, and in future experiments requiring potassium superoxide we would recommend using crown ether to improve solubility. In conclusion, we choose to monitor superoxide production as an indicator of inflammation, using cytochrome c as the detection probe, for several reasons. Because superoxide production is one of the first events to take place upon stimulation of inflammation, the total assay time is relatively short resulting in more experiments with less time needed to perform them. Cytochrome c makes an excellent probe because it is
46 inexpensive, readily available, spectroscopically active, and easily converted from moles of reduced cytochrome c to moles of superoxide produced because of the 1:1 stoichiometry. Lastly, the cytochrome c reduction assay is a well established method that •- has been shown to have great sensitivity in detecting O2 production in mammalian cells.
47 Figure 3.1. Cytochrome c spectrum. Spectrum of oxidized cytochrome c (□) and dithionite reduced cytochrome c (■).
1.2
1
0.8
0.6
absorbance 0.4
0.2
0 540 545 550 555 560 wavelength (nm)
48 Figure 3.2. Cytochrome c reduction by stimulated monocyte cells. Typical unstimulated control cells (□) and phorbol ester-stimulated cells (■). Cells were collected and primed in the presence of LPS. The control cells were not induced to produce superoxide, while phorbol ester was used as described in Chapter 4 to induce the respiratory burst in the stimulated cells. All cell samples contained cytochrome c as a probe to detect superoxide production. After stimulation, cells were removed by centrifugation and the spectrum of the supernatants was obtained using a plate reader. (Page 21, Book 4).
0.32
0.3
0.28
0.26 absorbance
0.24
0.22 540 545 550 555 560 wavelength (nm)
49 Figure 3.3. Relationship between reduced cytochrome c and slope of the spectral peak. Cytochrome c was reduced by addition of successive aliquots of dithionite, and the absorbance spectrum between 540-560 nm was recorded. The absorbance at 550 nm was used in combination with the extinction coefficient to determine how much cytochrome c was reduced with each addition of dithionite. The slope of the spectral peak calculated between 544-546 nm is plotted as a function of reduced cytochrome c.
0.02
0.015
0.01
y = 0.0031x - 0.0008 R2 = 0.9852 Slope (544-546nm)Slope 0.005
0 0246810 Cytochrome C reduced (nmoles)
50 3.4. Chapter 3 References
1. Karupiah, G.; Hunt, N.H.; King, N.J.; Chaudhri, G. NADPH oxidase, Nramp1 and nitric oxide synthase 2 in the host antimicrobial response. Rev. Immunogenet. 2000, 2, 387-415.
2. Caldefie-Chezet, F.; Walrand, S.; Moinard, C.; Tridon, A.; Chassagne, J.; Vasson, M. Is the neutrophil reactive oxygen species production measured by luminol and lucigenin chemiluminescence intra or extracellular? Comparison with DCFH-DA flow cytometry and cytochrome c reduction. Clin. Chim. Acta 2002, 319, 9-17.
3. Noguchi, N.; Nakada, A.; Itoh, Y.; Watanabe, A.; Niki, E. Formation of active oxygen species and lipid peroxidation induced by hypochlorite. Arch. Biochem. Biophys. 2002, 397, 440-447.
4. Chelomin, V.P.; Zakhartsev, M.V.; Kurilenko, A.V.; Belcheva, N.N. An in vitro study of the effect of reactive oxygen species on subcellular distribution of deposited cadmium in digestive gland of mussel Crenomytilus grayanus. Aquatic Toxicol. 2005, 73, 181-189.
5. Takeuchi, T.; Nakajima, M.; Morimoto, K. Relationship between the intracellular reactive oxygen species and the induction of oxidative DNA damage in human neutrophil-like cells. Carcinogenesis 1996, 17, 1543-1548.
6. Halliwell, B. Oxygen and nitrogen are pro-carcinogens. Damage to DNA by reactive oxygen, chlorine and nitrogen species: measurement, mechanism and the effects of nutrition. Mutat. Res. 1999, 443, 37-52.
7. Raha, S. and Robinson, B.H. Mitochondria, oxygen free radicals, disease and aging. Trends Biochem. Sci. 2000, 25, 502-508.
8. Droge, W. Free radicals in the physiological control of cell function. Physiol. Rev. 2002, 82, 47-95.
9. Lambeth, D.O. and Palmer, G. The kinetics and mechanism of reduction of electron transfer proteins and other compounds of biological interest by dithionite. J. Biol. Chem. 1973, 248, 6095-6103.
10. Megyeri, P.; Pabst, K.M.; Pabst, M.J. Serine protease inhibitors block priming of monocytes for enhanced release of superoxide. Immunology 1995, 86, 629-635.
11. Saito, H.; Salmon, J.A.; Moncada, S. Influence of cholesterol feeding on the production of eicosanoids, tissue plasminogen activator and superoxide anion (O2-) by rabbit blood monocytes. Atherosclerosis 1986, 61, 141-148.
51 12. Aida, Y. and Pabst, M.J. Neutrophil responses to lipopolysaccharide. Effect of adherence on triggering and priming of the respiratory burst. J. Immunol. 1991, 146, 1271-1276.
13. Lokesh, B.R. and Cunningham, M.L. Further studies on the formation of oxygen radicals by potassium superoxide in aqueous medium for biochemical investigations. Toxicol. Lett. 1986, 34, 75-84.
14. Valentine, J.S. and Curtis, A.B. A convenient preparation of solutions of superoxide aniom and the reaction of superoxide anion with a copper (II) complex. J. Am. Chem. Soc. 1975, 97, 224-226.
52 4. Investigation of tannins as anti-inflammatory agents 4.1. Introduction 4.1.1. Peripheral blood mononuclear cells (PBMCs) A source of cells was required to allow us to develop an ex vivo system for assessing the anti-inflammatory activities of various tannins and tannin-protein complexes. Systems that have been used in other labs include mammalian cell lines of monocytes or similar cells, cells derived from mouse spleen, or mammalian PBMCs. We chose to use PBMCs because they can conveniently be collected without sacrificing animals for every experiment, and because they are normal, rather than transformed cells. Blood is a convenient, readily available source for PBMCs. We choose to use rabbit blood as our cell source because it is relatively inexpensive to house and maintain a rabbit. Assuming that the rabbit is in a healthy, physiologically invariant state, experimental results obtained from each experiment can be compared within reason. In contrast, human blood donors are not always conveniently available, can be physiologically variable, and can only be used with permission of the Human Safety Committee. The inflammatory response of PBMCs has been characterized in detail. Generally PBMCs contain between 25-35% monocytes/macrophages, 65-75% lymphocytes, and less than 1% neutrophils, eosinophils, and basophils (1). All of these cells originate from hematopoietic stem cells and are essential elements in the immune response. They can be found in circulating blood/lymphatic system, lungs, spleen, tonsils and the small intestine (2). Hematopoietic stem cells develop into leukocytes, platelets, and erythrocytes. Because cells other than monocytes/macrophages may produce cytokines and other inflammatory stimuli, it is important to include the entire white blood cell population rather than just enriched subpopulations. Using the whole population provides the most flexibility for measuring a variety of inflammatory markers with the same preparation at the same time and it is more biologically relevant to include all leukocytes in the measurements of inflammatory markers. PBMCs have been found to be more sensitive and give superior results to whole blood, monocytic cell lines, and purified monocytes (3, 4).
53 Once cells have been obtained, they must be primed and activated for use in the assay. Priming is enhancement of cellular functions such as increased production of cellular enzymes, glycoproteins, and production of ROS in order to enhance phagocytosis. Priming of cells such as neutrophils and monocytes/macrophages takes place when the cells are incubated with LPS, as well as in response to cytokine release by other cells within the mononuclear collection (5). Within monocytes/macrophages there is an increase in SOD and catalase in response to LPS priming, which implies protection of these cells against their own ROS products that are so necessary for killing invading pathogens. LPS has been shown to up-regulate complement receptors and adhesion molecules to allow for adherence to the Teflon or serum-coated glass (6, 7). Gadgil et al. (2003) found 16 cellular proteins and enzymes, including SOD and catalase, to be up- regulated in response to LPS-priming (8). After priming, the cells must be stimulated by agents such as phorbol 12- myristate 13-acetate (PMA) to initiate an inflammatory response. PMA activates protein kinase C, which in turn stimulates the NADPH-oxidase enzyme system that is responsible for superoxide production. PMA is a synthetic analog of diacylglycerol (9- 11). Adherence is an essential part of the immune response and it has been suggested that it contributes to superoxide production in vitro (12). Pabst et al. (1982) tested this hypothesis by incubating mononuclear cells cultures in dishes treated with 0.1% poly(2- hydroxyethyl methacrylate) in ethanol, which discourages adherence (12). It was found that the cells in untreated dishes produced 35 % more superoxide anion in the presence of cytochrome c. In vivo, monocytes migrate from the blood stream into the inflamed tissue to differentiate into macrophages and adherence is a necessary part of this process. Microscopic observation of monocytes primed with LPS, reveals that they have developed extensive pseudopods suggesting adherence. Pabst also found that the time that the cell cultures were in contact with LPS influenced that amount of superoxide production (12). The shorter the incubation period, up to 48 hours, the less superoxide was produced. Based on the literature, we chose to isolate PBMCs from rabbit blood and to
culture them for up to several days in a CO2-enriched atmosphere in tissue culture plates
54 that encourage adherence. We chose to prime the cells with LPS, and to stimulate them with PMA so that we could measure inflammation by monitoring superoxide production in the presence cytochrome c. We planned to add tannin to the system at various times to evaluate its affect on the inflammatory response.
4.2. Methods
4.2.1. Reagents
All glassware was sterilized by baking in a dry oven for 24 hours. Na2EDTA (Sigma ED2P) was dissolved in water at 0.1 mM and filter sterilized with vacuum filtration using Corning disposable bottle top filters with 0.22 µm polyethersulfone membranes. To obtain Histopaque with a density of 1.10 g/mL, Histopaque 1.083 g/mL and Histopaque 1.119 g/mL were mixed in equal volumes. All Histopaque was purchased from Sigma already sterilized. Solutions were mixed under the sterile hood and stored in a sterilized glass bottle at 4°C. RPMI 1640 media, without L-glutamine, phenol red, or sodium bicarbonate, was purchased through Caisson Laboratories (RPMI- 013P). The media was reconstituted with nano-pure water and supplemented with 2000 mg/L of sodium bicarbonate, 2 mM L-glutamine, 100 units/mL of penicillin, 100 µg/ml of streptomycin, and 10% fetal bovine serum (Invitrogen, 10082-139). The media was filter sterilized and poured into a one-liter media bottle with a screw top and stored at 4°C. Hank’s balanced salt solution (HBSS) without calcium, magnesium, or sodium bicarbonate was purchased through Sigma (H4385) as a 10x stock. HBSS was then diluted ten times, supplemented with sodium bicarbonate (0.35 g/L) and filter sterilized in the same manner as the RPMI 1640 media. It was stored at room temperature under the sterile hood. Trypan blue (Sigma T-6146) was dissolved at 4% by weight in PBS buffer pH 7.8 with 0.1 mM EDTA, and stored at room temperature under the sterile hood. LPS from E.coli, (Sigma L-2880), was made by dissolving LPS in PBS buffer pH of 7.8, supplemented with 0.1 mM EDTA, at a concentration of 0.25 mg/mL and then diluted to 1x10-4 mg/mL for use with the cells. LPS was sterile filtered and frozen aliquots were stored for later use. PMA (Sigma P-8139) was prepared in DMSO, at 0.16 µmole/mL. DMSO was used in place of PMA for unstimulated control cells. Oxidized (Fe3+) horse
55 heart cytochrome c (Sigma C-2506) was dissolved in autoclaved PBS buffer pH 7.8 at a concentration of 10 mg/mL. It was then aliquoted into microcentrifuge tubes for individual use and stored at -20°C. Lucigenin was dissolved in DMSO and buffer to achieve a stock solution of 8.5 mg/mL. Lucigenin was first dissolved in DMSO at half of the final solution volume and then diluted to the final volume with PBS pH 7.8 buffer. 4.2.2. Isolation of peripheral blood mononuclear cells (PBMC) To obtain PBMCs, approximately 50 mL of rabbit blood was collected in
Na2EDTA supplemented falcon tubes (50 mg Na2EDTA in a 50 mL tube) with a 21gauge butterfly needle. The blood was then diluted with sterile-filtered HBSS without calcium and magnesium at a ratio of 1:2, (v:v, blood:HBSS), and split into several falcon tubes allowing room for the Histopaque. Using a 10.2 cm long needle, Histopaque (density = 1.10 g/mL) was slowly layered underneath the diluted blood at a ratio of 4:3 (v:v, blood:Histopaque). Before use, the Histopaque was warmed in a hot water bath to approximately 27°C. Samples were centrifuged in a swinging-bucket centrifuge (IEC Centra-8) at room temperature for 30 min at 700 x g. The top layer, plasma, was carefully removed and discarded. Using a fine tipped transfer pipette, the next layer, containing the mononuclear cells, was removed and transferred to an autoclaved 50 mL falcon tube. The cell suspension was then centrifuged for 30 min at room temperature and 700 x g, to pellet the cells. The supernatant was removed and the cell pellet was resuspended in approximately 24 mL of HBSS and 18 mL of Histopaque was layered underneath the cell suspension to remove the red blood cells that contaminated the sample. The mononuclear layer was removed and suspended in 12 mL HBSS. The samples were centrifuged for 30 min at room temperature and 700 x g to pellet the PBMCs. The cells were then washed two times with HBSS by resuspending the cells in HBSS, and then centrifuging the cell suspension at 700 x g, room temperature, for 20-30 min or until the supernatant was clear. Finally, they were resuspended in RPMI 1640 media at a ratio of one volume of media to two volumes of blood collected. Media was warmed in a hot water bath to approximately 37°C, prior to use. Cells were then counted on a hemocytometer and viability was determined with trypan blue staining. Cells were diluted 1:20 with the 4% trypan blue stain and viable cells were identified by their exclusion of the blue stain.
56 4.2.3. Incubation and priming of PBMCs The cells were placed into a 24 well plate to achieve a final concentration of 5.0- 6.0x106 cells/mL of RPMI 1640 complete media. The suspension was dispensed such that the final volume within each well was 500 µL after all of the solutions were added. Typically this was done with about 457 µL of cells and additional media. The remaining volume was made up of LPS, Tannin, PMA, and cytochrome c. The plates were sealed in
CO2-producing pouches (Becton, Dickinson and Company 260662) and allowed to
incubate for 24 hours at 37°C. The bags maintained the cells at about 3-12% CO2 which helps induce adherence. An identical set of cell-free samples was always placed in one row on the plate to serve as a control. After 24 hours of incubation, all cells were primed with LPS by adding 5 µl of 0.1 ng/µl LPS to each well. In some samples, 5 µL of PGG dissolved in PBS buffer, pH 7.8, at a final concentration range of 1 nM - 1 mM per well was added immediately after LPS. Cells were incubated for another 24 hours at 37°C in
the CO2-producing bags. 4.2.4 PMA stimulation of PBMCs Primed cells were then stimulated with PMA in the presence of cytochrome c, which was used to detect superoxide production. Cytochrome c (10 mg/ml) was warmed in a hot water bath to 37°C before adding 30 µL to each well to achieve a final concentration of 48.5 µM. After adding cytochrome c, 3 µL of PMA (0.16 µmol/mL in DMSO) was added to each well to achieve a final concentration of 1 µM. The order of addition is important to ensure effective trapping of all of the superoxide that is produced. A set of LPS-primed cells that were not stimulated with PMA served as the unstimulated
control cells. All samples were incubated for 60 min at 37°C in the CO2-producing bags. 4.2.5. Superoxide production from PBMCs The medium was transferred from each well into a 0.65 µL microfuge tube and centrifuged for 10 min at 700 x g to sediment any non-adhering cells. For spectrophotometric readings, 250 µL of supernatant was transferred to a 96 well plate. Absorbance readings were taken at 2 nm intervals from 540 nm to 650 nm using a BIO- Tek Synergy HT microplate reader. 4.2.6. PGG interference with cytochrome c
57 Up to 30 µL of PGG (0.041µg/µL) was added in one microliter increments to 0.6 mg of cytochrome c dissolved in 1000 µL of buffer and read on a spectrophotometer. The reaction mixture was mixed by inversion, within a cuvette, between each addition of PGG. The visible absorbance spectrum was recorded after each aliquot was added. 4.2.7. Lucigenin as a probe for superoxide In an attempt to measure superoxide production with lucigenin instead of cytochrome c, primed cells were stimulated with PMA in the presence of lucigenin which was an alternate probe to detect superoxide production. Lucigenin was dispensed in the wells at 30 µL to achieve a final concentration of 1 µM, with the final volume remaining at 500 µL for the luminescent experiments. After adding lucigenin, 3 µL of PMA (0.16 µmol/mL in DMSO) was added to each well to achieve a final concentration of 1 µM. A set of LPS-primed cells that were not stimulated with PMA served as the unstimulated control cells. All samples were incubated for 60 min at 37°C in the BIO-Tek Synergy HT microplate reader. The wells within the plate were read directly as they incubated within the plate reader at 37°C. Readings were taken every minute for 60 min at a sensitivity of 100, and luminescence was detected by a photo multiplier tube within the luminometer of the BIO- Tek Synergy HT microplate reader. The optics were read from the bottom position through clear plates, using the Lum/E filter set. 4.3. Results Cytochrome c is reduced by the superoxide produced by the stimulated cells (Fig. 4.1A). There is a very low level of superoxide produced by the unstimulated control cells (Fig. 4.1A) Cytochrome c is not reduced by either PMA or DMSO as shown by the featureless spectra of cell-free controls (Fig. 4.1B). Although it was clear that superoxide was produced when the cells were stimulated by superoxide, it was not always easy to directly quantitate superoxide from the spectral data. Direct absorbance measurements at 550 nm did not always provide sufficient information to evaluate superoxide production. For example, in a typical experiment the stimulated cells clearly reduced cytochrome c but had an A550 of only 0.26, while the unstimulated controls did not reduce cytochrome c but had an absorbance of 0.30 (Chapter 3, Fig.3.2). We demonstrated that the slope of the reduced cytochrome c
58 absorbance peak was directly related to the amount of reduced cytochrome c (Chapter 3, Fig. 3.3) and used that method to measure cytochrome c production in the rest of our cell experiments. Using the slope of the reduced cytochrome c spectrum to quantitate superoxide production, we found that PMA-stimulated cells produced about 3 times more superoxide than the unstimulated control cells (Fig. 4.2). As expected, cytochrome c was not reduced in the cell-free control wells (Fig. 4.2) To evaluate the effects of PGG on the inflammatory response, various amounts of PGG were added to cells during the priming stage. The PGG-treated cells reduce more cytochrome c than the cells that were not exposed to the tannin (Fig. 4.3). We noted that PGG-containing cell-free control samples reduced a substantial amount of cytochrome c, while cell-free controls that did not contain PGG did not reduce cytochrome c (Fig. 4.3). We speculated that cytochrome c could be reduced by strong reductants such as polyphenolics. We confirmed that PGG itself reduces cytochrome c with control experiments (Fig.4.4). Cell-free control data could be used to correct for the interference by PGG in the measurement of superoxide by cytochrome c reduction. The amount of superoxide produced by cells was calculated by subtracting out the background reduction due to PGG, obtained from cell-free samples exposed to the same amount of PGG. The result was converted to nmoles of superoxide using the calibration curve obtained with dithionite (Chapter 3, Fig. 3.3). PGG appeared to suppress superoxide production in PBMCs in a concentration dependent manner (Fig.4.5), such that 1µM PGG suppresses superoxide production by more than 50%. Due to the finding that PGG and other phenolics interfere with the cytochrome c assay for superoxide by reducing cytochrome c (Fig. 4.4), we began preliminary experiments to investigate the utility of luminol and lucigenin for monitoring the superoxide burst in our cell cultures. The luminescence of luminol or lucigenin, is another highly sensitive and widely used probe for measuring superoxide production by mammalian cells. Because luminol permeates the cell membrane, its chemiluminescence is representative of both intracellular and extracellular reactive oxygen species. When luminol reacts with an oxidizing species such as superoxide, it is oxidized to the
59 electronically excited aminophthalate anion. As the ion relaxes to the ground state, it releases photons which can be detected by a photo multiplier tube within the luminometer (13). Luminol chemiluminescence has long been used to detect superoxide production in the enzymatic reaction of xanthine oxidase (14). Lucigenin measures the superoxide anion directly and has been shown to yield results that are closely correlated to the cytochrome c reduction assay (9). This is because lucigenin does not permeate cells and hence only measures extracellular •- superoxide production. The radical form of lucigenin that reacts with O2 is univalently reduced rather than oxidized as with luminol. This reduced form of lucigenin then reacts with another superoxide radical to yield dioxetane, which then decomposes to into two molecules of acridone. One of the two molecules is electronically excited and releases a photon when it returns to the ground state (15). The chemiluminescence is then detected by a luminometer. In an effort to overcome the interference that we encountered when PGG was used in the cytochrome c assay for superoxide, we attempted to measure superoxide production by chemiluminescence of lucigenin (Fig. 4.6). In our preliminary experiments we were able to measure the respiratory burst using lucigenin. We did observe a slightly higher rate of chemiluminescence in stimulated cells than in unstimulated control cells (Fig. 4.6) but had problems with sensitivity that were not resolved. When the area under the curve was calculated for the chemiluminescence data, and background luminescence from unstimulated control cells was subtracted out, PGG seems to suppress superoxide production from cells at high concentrations of tannin (Fig. 4.7). Unfortunately, we were unable to optimize the system to give sensitive results and did not pursue it in further experiments. 4.4. Discusion Our goal was to develop an ex vivo assay that allowed us to screen tannins and tannin-protein complexes for their anti- as well as pro-oxidant properties. We chose to isolate rabbit PBMCs for our assay. In order to isolate rabbit PBMCs, we adapted methods used for human and rabbit blood (1, 16-18). Rabbit blood is denser than human blood and therefore requires a different density gradient for isolating PBMCs. Feldman and Mogelesky found a 1:1 (v/v) mixture of Histopaque 1.083 g/mL and Histopaque
60 1.119 g/mL (Histopaque 1.10 g/mL) to provide superior results in contrast to other commonly used density gradients such as Ficoll-Paque, Lymphopaque, or Histopaque of other densities. Histopaque 1.10 g/mL was found to yield about 95% purity of mononuclear cells, which was the highest percent purity found within that study. Furthermore cell viability was found to be greater than 95% (19). The reagents, concentrations, and times of exposure for priming and stimulating PBMCs to produce superoxide were adapted from the Pabst lab (8, 12, 17). We chose to use LPS to prime the cells for enhance superoxide production and PMA to stimulate the cells. It has been shown that LPS primes macrophages for enhanced superoxide production upon stimulation with PMA, and that this response can be effective at even extremely low concentrations, 0.1ng/mL, making LPS almost as biologically sensitive as ricin (12, 20-22). Pabst et al. (1982) suggest that this small amount of LPS required for priming suggests a mechanism relevant to natural cell systems (12). Trace amounts of endotoxin leave the GI tract before being inactivated by serum lipoproteins, which help to keep circulating monocytes primed for superoxide response to injury or noxious agents (12). Pabst found monocytes that were treated with LPS developed extensive pseudopods, suggesting that LPS helps in adherence. Megyeri suggests that perhaps LPS works by activating a protease that cleaves an inhibitor of priming (17). Monocytes are extremely effective at killing bacteria based on their ability to generate superoxide and that as they exist in the bloodstream, they have a high superoxide generating capacity (12). We found the 48 hour time frame for the assay to be the best time for maximal superoxide production and this is consistent with the literature (8, 12, 17). The time of addition of PGG was chosen to agree with other studies that use inhibitors to stop the macrophage response (17). Despite the fact that this assay holds a lot of potential benefits there are some short comings as well. The cell isolation technique is difficult and sophisticated and may result in accidental priming of monocytes (12). To circumvent this, training is important as well as sterile conditions that result in endotoxin free reagents and supplies. Red blood cells inevitably contaminate the sample, which can result in interference with spectroscopic results. Running through another density gradient or employing a
61 hypotonic lysis procedure can reduce the amount of contaminating red blood cells, but it reduces, sometimes significantly, the amount of PBMCs. The cell concentration was difficult to determine and even when the concentrations were the same from one experiment to the next, the resultant superoxide production was variable. And lastly, our
lab would need to acquire a CO2 incubator for extended use of this protocol. The CO2 producing bags work fine in the short term, but are expensive and variable. Our system has been shown to produce superoxide from rabbit PBMCs that have been primed with LPS and stimulated with PMA, as suggested by cytochrome c and preliminary data from lucigenin. Unfortunately, tannins such as PGG also reduce cytochrome c, as does superoxide, interfering with the results obtained from the cytochrome c reduction assay. However, when calculations are preformed to subtract out the background reduction, created from PGG, from the total reduction, it appears as though tannin does decrease the total amount of superoxide produced from stimulated cells. This was supported by preliminary data with lucigenin and PGG. Further work on finding methods for monitoring inflammation are needed to allow us to effectively use our ex vivo system.
62 Figure 4.1. Superoxide production from rabbit PBMCs in the presence of cytochrome c. (A) Rabbit PBMCs were primed with LPS for 24 hr. Cytochrome c was added to quantitate superoxide production in unstimulated control cells (□) and in cells that were stimulated with PMA (■) (B) Cell-free controls were wells that contained media but no cells. (Page 26, Book 4) A. 0.69
0.59
0.49
absorbance
0.39
0.29 540 545 550 555 560 wavelength (nm)
B. 0.69
0.59
0.49
absorbance 0.39
0.29 540 545 550 555 560 wavelength (nm)
63 Figure 4.2. Superoxide production represented as slope of the spectral data. Slope data is calculated from the spectra of unstimulated control cells and PMA stimulated cells (□) and samples without PBMCs (■). The bars represent the means of triplicate values, with the error bars representing the standard deviations. (Page 26, Book 4).
0.035
0.03
0.025
0.02
0.015
Slope (544-546) 0.01
0.005
0 Unstimulated Stimulated control
64 Figure 4.3. Superoxide production from cells exposed to tannin. Cells were primed with LPS without PGG or in the presence of 1nM PGG, 0.5µM PGG, or 1µM PGG. The amount of cytochrome c reduced by cell-free controls (■) is compared to cytochrome c reduction by the PMA-stimulated cells (□).
0.045 0.04 0.035 0.03 0.025 0.02 0.015 Slope (545-546) 0.01 0.005 0 0nM 1nM 0.5µM 1µM PGG
65 Figure 4.4. PGG reduces cytochrome c. PGG (0.04 µg/µL) was added in one microliter increments to 0.6 mg of cytochrome c dissolved in 1000 µL of buffer. The amount of cytochrome c reduced was calculated from the extinction coefficient at 550 nm.
35
30
25
20
15
10
5 Cytochrome C reduced (nmoles) C Cytochrome 0 00.511.5 PGG (ug)
66 Figure 4.5. Superoxide production in PGG-treated cells. Superoxide production calculated from the difference between the slopes for cell-free and cell-containing samples with various amounts of PGG.
14
12
10
8
6
Superoxide (nmoles) 4
2
0 0 0.2 0.4 0.6 0.8 1 PGG (uM)
67 Figure 4.6. Superoxide production from rabbit PBMCs probed with lucigenin. Chemiluminescence of lucigenin from PMA stimulated cells (■) and unstimulated control cells (□). (Page 75, Book 4).
35
30
25
20
15
Chemiluminescence 10
5
0 0 102030405060 Time (min)
68 Figure 4.7. Superoxide production from cells exposed to tannin using lucigenin as a probe. Calculated area under the curve for lucigenin, PMA stimulated cells and unstimulated control cells. Calculated raw data (A) and data adjusted by subtracting out the unstimulated control cell background from each sample (B). (Page 75, Book 4).
A. 1400 1200
1000 800
Area 600
400
200
0 A n n in n M nni n P a anni a DMSO T T d /low Tannin A/high MA/high T M MA/me P P P PMA No
275 B.
225
175
125
Area
75
25
-25 O A n in in S M ni n n nnin nn P a a DM Ta T T h gh w i ig /lo /h h A A M P PMA/ PMA/med Tan No PM
69 4.5 Chapter 4 References
1. Tracy, P.B.; Robinson, R.A.; Worfolk, L.A.; Allen, D.H. Procoagulant activities expressed by peripheral blood mononuclear cells. Meth. Enzymol. 1993, 222, 281-299.
2. Nester, E.W.; Roberts, C.E.; Nester, M.T. Microbiology: A Human Perspective. William C. Brown: Boston, MA 1995; 333-401.
3. Poole, S.; Mistry, Y.; Ball, C.; Gaines Das, R.E.; Opie, L.P.; Tucker, G.; Patel, M. A rapid 'one-plate' in vitro test for pyrogens. J. Immunol. Methods 2003, 274, 209-220.
4. Taktak, Y.S.; Selkirk, S.; Bristow, A.F.; Carpenter, A.; Ball, C.; Rafferty, B.; Poole, S. Assay of pyrogens by interleukin-6 release from monocytic cell lines. J. Pharm. Pharmacol. 1991, 43, 578-582.
5. Aida, Y. and Pabst, M.J. Priming of neutrophils by lipopolysaccharide for enhanced release of superoxide. Requirement for plasma but not for tumor necrosis factor-alpha. J. Immunol. 1990, 145, 3017-3025.
6. Aida, Y. and Pabst, M.J. Neutrophil responses to lipopolysaccharide. Effect of adherence on triggering and priming of the respiratory burst. J. Immunol. 1991, 146, 1271-1276.
7. Marra, M.N.; Wilde, C.G.; Griffith, J.E.; Snable, J.L.; Scott, R.W. Bactericidal/permeability-increasing protein has endotoxin-neutralizing activity. J. Immunol. 1990, 144, 662-666.
8. Gadgil, H.S.; Pabst, K.M.; Giorgianni, F.; Umstot, E.S.; Desiderio, D.M.; Beranova- Giorgianni, S.; Gerling, I.C.; Pabst, M.J. Proteome of monocytes primed with lipopolysaccharide: analysis of the abundant proteins. Proteomics 2003, 3, 1767-1780.
9. Caldefie-Chezet, F.; Walrand, S.; Moinard, C.; Tridon, A.; Chassagne, J.; Vasson, M. Is the neutrophil reactive oxygen species production measured by luminol and lucigenin chemiluminescence intra or extracellular? Comparison with DCFH-DA flow cytometry and cytochrome c reduction. Clin. Chim. Acta 2002, 319, 9-17.
10. Dallegri, F.; Patrone, F.; Ballestrero, A.; Frumento, G.; Sacchetti, C. Chemotactic peptide enhancement of PMA triggered monocyte cytotoxicity. Clin. Exp. Immunol. 1984, 57, 717-721.
11. Karupiah, G.; Hunt, N.H.; King, N.J.; Chaudhri, G. NADPH oxidase, Nramp1 and nitric oxide synthase 2 in the host antimicrobial response. Rev. Immunogenet. 2000, 2, 387-415.
70 12. Pabst, M.J.; Hedegaard, H.B.; Johnston, R.B.J. Cultured human monocytes require exposure to bacterial products to maintain an optimal oxygen radical response. J. Immunol. 1982, 128, 123-128.
13. Allen, R.C. and Loose, L.D. Phagocytic activation of a luminol-dependent chemiluminescence in rabbit alveolar and peritoneal macrophages. Biochem. Biophys. Res. Commun. 1976, 69, 245-252.
14. Totter, J.R.; deDucros, E.C.; and Riveiro, C. Use of chemiluminescent compounds as possible indicators of radical production during xanthine oxidase action. J. Biol. Chem. 1960, 235, 1839-1842.
15. Faulkner, K. and Fridovich, I. Luminol and lucigenin as detectors for superoxide. Free Radic. Biol. Med. 1993, 15, 447-451.
16. Isaguliants, M.G.; Kadoshnikov, Y.P.; Kalinina, T.I.; Smirnov, V.D.; Wahren, B. Specificity of humoral and cellular immune response against recombinant particles of nucleocapsid protein of human hepatitis B virus in rabbits. Biochemistry.Moscow 1998, 63, 551-558.
17. Megyeri, P.; Pabst, K.M.; Pabst, M.J. Serine protease inhibitors block priming of monocytes for enhanced release of superoxide. Immunology 1995, 86, 629-635.
18. Paulnock, D.M. Macrophages:a practical approach. Oxford University Press: Oxford, United Kingdom, 2000.
19. Feldman, D.L. and Mogelesky, T.C. Use of Histopaque for isolating mononuclear cells from rabbit blood. J. Immunol. Methods 1987, 102, 243-249.
20. Olsnes, S.; Refsnes, K.; Pihl, A. Mechanism of action of the toxic lectins abrin and ricin. Nature 1974, 249, 627-631.
21. Speer, C.P.; Pabst, M.J.; Hedegaard, H.B.; Rest, R.F.; Johnston, R.B.,Jr Enhanced release of oxygen metabolites by monocyte-derived macrophages exposed to proteolytic enzymes: activity of neutrophil elastase and cathepsin G. J. Immunol. 1984, 133, 2151- 2156.
22. Johnston, R.B., Jr.; Chadwick, D.A.; Cohn, Z.A. Priming of macrophages for enhanced oxidative metabolism by exposure to proteolytic enzymes. J. Exp. Med. 1981, 153, 1678-1683.
71