Genistein's effects on lytic activity and intracellular tyrosine phosphorylation of natural killer cells

iv


INTRODUCTION
The lower incidence of breast, colon, and prostate cancer in Asian countries brought attention to the differences in diet between Asian and Western cultures, particularly the consumption of soy-containing foods (36). The anti-cancer effects are thought to be due to isoflavones in soy, principally genistein. Use of in vivo (animal) models, demonstrate that soy products containing genistein in the diet significantly reduce the risk of cancer. Use of in vitro models, demonstrate that genistein inhibits the proliferation of human tumor cell lines (5). Previous in vitro studies in our lab demonstrated enhancement of lytic activity of harvested human NK cells against human tumor cells in the presence of low concentrations of genistein (< 1 µM). These effects may be related to regulation of NK cells that are involved in killing tumor cells and virally transformed cells.
Tyrosine phosphorylation has been found to be an early and necessary signal for lytic activity in natural killer (NK) cells (20). Genistein, an isoflavone, is a specific inhibitor of protein tyrosine kinases (PTKs) and can be used as a tool for understanding the role PTKs play in the activation of cytotoxicity (2). Previous studies have shown that genistein (10 µg/ml or 37 µM) inhibits the src protein tyrosine kinase p56 1 ck in "NK-rich" cells exposed to IL-2 with a resultant decrease in cytotoxic activity against K562 cells (39). Thus, biological studies have shown that genistein can enhance or inhibit NK cytotoxicity depending on the assay conditions used.
The purpose of this research was to ask if low concentrations of genistein had a direct effect on natural killer cell lytic activity and if the observed cytobiological effects of genistein were related to changes in signal transduction, particularly protein tyrosine kinases (PTKs). Initial studies used NK cells harvested from peripheral blood leukocytes. When it was observed that genistein had an effect on lytic activity, a pure NK cell line was to study the direct results of genistein on NK cells. The NK3.3 cell line was used as a model system so that a pure population of natural killer ceUs would define direct results of genistein on those cells. Intracellular levels of phosphorylated tyros:ine residues after treatment with low doses of genistein were examined by flow cytometry (21,24). Lytic activity was assayed in parallel by a chromium release assay (12). Together these studies attempt to define a relationship between enhanced lytic function and intracellular tyrosine phosphorylation due to genistein.

LITERATURE REVIEW
The immune system is divided into two branches. There is acquired (adaptive) immunity that specifically recognizes and eliminates that which is foreign. The second branch is innate (natural) immunity, which is the basic resistance to disease one is born with. Contrary to some beliefs, the innate immune system is not some left over relic from evolution nor is it the "containment crew" until adaptive immunity "kicks in". Rather adaptive immunity owes its existence to innate immunity. Innate immunity is composed of anatomic barriers, phagocytic cells, complement, and natural killer cells. The last three components of the innate immune system generate signals that activate and even direct the type of effector response of acquired immunity (35,46). Neutrophils and macrophages are two key phagocytic cells of innate immunity.
Inflammation is the hallmark of an innate immune response and it also activates neutrophils and macrophages (46). Neutrophils are the most numerous of granulocytes found in the blood. Macrophages are derived from monocytes circulating in the blood that migrate into tissue and become tissue specific macrophages (28). Both cell types recognize conserved structures found on microbes called PAMPS for pathogen-associated molecular patterns and components of the complement system (35,46). In addition, both have complement receptors and Fe receptors in order to recognize organisms covered with C3b, an active fragment of complement component C3 or antibody, a process called opsonization. Neutrophils and macrophages engulf microbes. The granules of neutrophils contain lytic enzymes and bactericidal substances that fuse with a phagosome to digest infectious material (28). Neutrophil granules also induce the release of biological response modifiers, cytokines and chemokines, by macrophages (46). Macrophages, after internalizing infectious material, destroy them by the creation of a phagolysosome that contains hydrogen peroxide, peroxidase, lysozyme, and other hydrolytic enzymes, which digest the material. The digestive fragments (antigens) are brought to the surface of the macrophage where they can be presented to T-cells (28,47). Macrophages binding to infectious material also stimulate the release of cytokines and chemokines.
Cytokines released by phagocytic cells are able to activate the complement cascade. Complement is a group of serum proteins involved in inflammation, activation of macrophages, and participates in an enzymatic cascade that culminates in lysis of cell membranes (47). Complement and phagocytic cells have a symbiotic relationship. Complement activation induces chemotaxis and opsonization, which generates the participation of macrophages. Macrophages produce cytokines that stimulate the liver to produce acute phase proteins, Creactive protein and mannan-binding lectin (MBP) that activate complement by different pathways. C-reactive protein does this by binding to and activating complement component one, much like the initiation of the classic pathway by antibody binding. MBP recognizes and binds to mannose found on bacterial cell walls to activate complement by the lectin pathway. This reaction substitutes for the activation of the first component of complement.
Phagocytic cells and complement are efficient for dealing with extracellular bacterial infections but they cannot reach pathogens (viruses as an example) that reside inside of cells. Natural killer cells have the intrinsic ability to lyse virally infected cells and tumor cells particularly hemopoietic tumors. NK cells play a role in innate immunity as effector cells and a source of cytokines that modulate other immune cell activities (45).

Natural Killer Cells
Natural killer (NK) cells can be found in earthworms, sharks, common fish, amphibians, reptiles, birds, and mammals. In humans, NK cells are the third major class of lymphocytes and comprise 10 -15% of the lymphocyte population in the blood, liver, and especially in the spleen (red pulp), but are rarely found in lymph nodes. NK cells are thymic independent, do not undergo immunoglobulin (lg) or T cell receptor (TCR) gene rearrangement, have no classic antigen binding receptors, and no CD3. NK cells have non-specific cytotoxicity and do not require antigenic stimulation or MHC class restricted presentation of antigen. The developmental lineage of NK cells is unknown, however there are two theories as to their origins.
One is that NK cells develop from a distinct lineage from that of T-cells (since SCIO mice who have no T-cells still have NK cells) or second, both NK·and T cells originate from the same early progenitor cell before rearrangement of TCR genes since NK cells do share some markers with T-cells. Morphologically, NK cells are referred to as large granular lymphoctyes (LGL) due to having a high cytoplasm to nucleus ratio, a kidney shaped nucleus, and large azurophilic granules (28). NK cells can be activated (to kill targets) by cytokines or cell-to-cell contact (44). Definitive cell surface markers for NK cells are CD16 and CD56. CD16 is a FcyRIII receptor found as a Type I transmembrane protein on NK cells. CD16 belongs to the lg superfamily and mediates binding to lgG coated particles. CD56 also is a Type I transmembrane protein, member of the lg superfamily, and is an isoform of N-CAM (neural cell adhesion molecule) called a hemophilic adhesion molecule. CO2 (LFA-2) is another marker found on the NK cell surface. CO2 is also found on T-cells and is sometimes called the E-rossette receptor. It is a Type I transmembrane protein of the lg superfamily, an adhesion molecule whose ligand is CD58 that is found on many cell types. In the case of T-cells, CO2 serves as an accessory molecule to promote binding of cytotoxic T-cells to their targets (23). CD57, also known as HNK-1, is an oligosaccharide present on NK cells with no known function. NK cells express the and y signal transducing subunits of the IL-2 receptor but not the a subunit required for high affinity binding of IL-2 to the receptor. When expressed, IL-2Ra reduces the amount of IL-2 needed for activation and growth. Since there is no IL-2Ra expressed after stimulation with IL-2, NK cells, unlike T cells when activated with IL-2, cannot reduce the amount of IL-2 needed to maintain activation (1).
When stimulated by high concentrations of IL-2, NK cells proliferate and lytic activity is enhanced. This enhancement is a consequence of IL-2 activation to transform NK cells to lymphokine activated killer (LAK) cells. LAK cells have enhanced cytotoxic ability and a broader range of target specificity. This means that LAK cells, unlike NK cells, can kill a wider variety of tumor cells and even normal cell types like epithelial cells.
There are two pathways that NK cells utilize for cytolytic activity: antibody dependent cellular cytotoxicity (ADCC) by CD16 and direct cell-mediated cytotoxicity. With ADCC, the target cell is coated with specific lgG to which the low affinity Fey receptor, CD16, binds. This binding engages the pathway to lytic activity while also activating NK cells to synthesize and secrete cytokines such as TNF and IFN-y. Since CD16 is a low affinity receptor, it can only bind to aggregate but not monomeric lgG (1).
Antibody independent lysis is not as well understood as that for ADCC since no known definitive activating receptor for NK cells or ligand on target cells has been identified. There are several theories as to how this pathway may be activated. One is that direct cell-mediated cell lysis is triggered by an unidentified surface NK receptor to an unknown structure on a malignant or virally infected cell (1,20).
Another is that target cells lack expression of a normal protective molecule. The identification of the target molecule not being encoded in MHC class I or that lack of expression of MHC class I induces lysis. The latter theory is supported by the fact that virally infected cells and tumor cells have decreased MHC I expression.
There is also the two-receptor model theory, which postulates regulation of NK lytic activity through inhibition of lysis. That is, the NK cell receives signals from two different receptors. One receptor binds to ligands on the target cell that are possibly carbohydrates, this is the signal to kill. This signal is prevented by a second receptor that binds to certain alleles of MHC class I. This negative signal overrides the positive signal. Low MHC class I expression would result in a reduction of the negative signal so that the positive signal would be greater and result in killing (1,28,44).
Though no known activating receptor has been characterized for NK cells this has not been the case for inhibitory receptors found on NKcells. Inhibitory receptors have been found on macrophages, mast cells, and killer cells (13 (56). In addition there is also a homodimer of p70 called p140 that can recognize HLA-A alleles (13). Human KIR recognition of polymorphic determinates of HLA molecules has been found to be limited to the carboxy-terminal of the a1 domain of MHC class I (56). A second group of human inhibitory receptors are Type II transmembrane proteins with C type lectin extracellular domains. One type of this receptor, CD94/ NKG2, exists as a hetrodimer that also binds MHC class I (13,25,29). As CD94/ NKG2 has C-lectin domains, it may be able to bind to a common carbohydrate in different MHC class I alleles and would enable this receptor to have a broader binding capacity allowing it to recognize remaining HLA-B and HLA-A alleles not recognized by the Type I Kl Rs. Both classes of inhibitory receptors have in common that they recognize and bind to MHC class I molecules but also contain two immunoreceptor tyrosine inhibitory motifs (ITIMs) separated by 26-28 amino acids in their long cytoplasmic tails (18,37).
It is hypothesized that NK inhibitory receptors prevent cell activation in certain conditions, times, or locations to protect sensitive areas. It is suggested that inhibitory receptors prevent harmful activity of NK cells while the cells mature and migrate where their presence is required (13). However, binding of KIRs does not lead to cell death or long-lived anergy. The identification of KIRs provides evidence to validate the two-receptor theory for activation.
NK inhibitory receptors override activation signals to prevent lysis. When KIRs bind MHC class I ligand, downstream signaling events responsible for generating inositiol-1, 4,5-trisphosphate (IP3) and increasing intracellular calcium are shut down and there is a partial disruption of tyrosine phosphorylation (56). Upon engagement of KIR to MHC class I, tyrosine residues in the ITIMs are phosphorylated by a src family member (lck and lyn, but not fyn) (7,14). This phosphorylation in turn recruits and activates the tyrosine phosphatase, SHP-1 (14,15). SHP-1 appears to dephosphorylate an early substrate that is found in the activation pathway. A possible candidate for that substrate could be the adapter protein pp36 that associates with phospholipase C gamma (PLCy) or another adapter protein called Grb2. Grb2 and pp36 control PLCy enzymatic activity (55). For inhibition by KIRs, there must also be co-ligation with an activating receptor. This would suggest that KIRs and activating receptors would need to be in close proximity to each other (13).
There are a group of KIRs with short cytoplasmic tails (39 amino acids long) that have no ITIMs, and have a charged lysine in the transmembrane domain indicating they have the ability to associate with signal transducing molecules. These receptors are known as killer activating receptors (KARs) and have been proposed to be activating receptors for NK cells (18).

Mechanism of Target Cell Lysis
The process of target cell lysis by NK cells is the same whether initiated by ADCC or the Ab-independent pathway and is similar to that used by cytotoxic T lymphocytes. This process includes granule exocytosis and initiation of DNA fragmentation and apoptosis. As they contain large granules at all times in the cytoplasm, NK cells are always constitutively cytotoxic unlike T-cells which must be stimulated before granules appear. The cytotoxic process is initiated by the recognition of target through cell-to-cell contact. The contact occurs through CD16 on an NK cell binding to lgG on a target cell or the binding of the unknown receptor and ligand for Ab-independent lysis. Once adherence takes place, a signal is sent to activate the lytic pathway. Degranulation of the NK cell occurs at the site of contact.
NK granules contain perforin, granzymes (neutral serine proteases), proteoglycan, and cytotoxins. These substances deliver the "lethal hit" to the target cell. Perforin is a monomeric pore forming protein that requires the presence of calcium so it can bind to the target membrane and form a pore. After entering the target cell, granzymes bind to substrates that are involved with apoptosis or can be transported to the nucleus where they can directly bind and activate death substrates. Cytokines such as TNF-a enter the target to further damage the target cell integrity. So unlike the "true" lysis seen with complement, NK cells activate an endogenous apoptotic pathway with granzymes leading to cell death. Proteoglycan (chondroitin sulphate A) protects the NK cell from self-destruction. After an NK cell delivers its "lethal" hit, it can be "recycled" to lyse other targets, become inactive, or undergo apoptosis (1,28,31,44,47,50).

Natural Killer Cell Functions
The two major roles of NK cells are immune surveillance against tumor cells and virally infected cells and the release of cytokines (specifically IFN-y) to activate macrophages and T lymphocytes (44). NK cells can inhibit proliferation of tumor cells and blood borne metastasizing tumor cells by their characteristic "natural" killing (42). Tumor cells have decreased MHC class I expression especially those tumors mediated by oncogenic viruses. While reduction of MHC class I compromises cytotoxic T lymphocyte activity, it seems to be one requirement for recognition and lysis by NK cells (1,28). Tumor cells may also express specific surface proteins that NK cells recognize (1). NK cells can lyse tumor cells coated with antibody by ADCC and this killing is unaffected by the presence or absence of MHC class I. Chediak-Higashi syndrome and the animal model of this disease called beige mouse readily demonstrates the importance of NK cells in tumor immunity. In Chediak-Higashi patients, NK cell lytic activity is impaired due to an abnormality in the granules that contain granzymes. Patients experience recurrent infections and an aggressive but nonmalignant infiltration of organs by lymphoid cells (1,28).
NK cells have other functions as they serve to control infections caused by viruses, parasites, fungi, and bacteria. NK cell produced cytokines are of more importance than NK lytic activity in controlling bacterial infections (53). Bacterial infections cause inflammatory signals that attract NK cells. NK cells can destroy extracelluar bacteria via ADCC by binding to opsonizing lgG bound to bacteria.
Macrophages containing intracellular bacteria are stimulated to produce IL-12, which activates NK cells. In response to this activation, NK cells produce cytokines (IFN-y) to recruit other immune effectors involved with inflammatory responses (1,22,42). and hematopoiesis (IL-3, GM-CSF). NK cells have roles in autoimmune disorders and graft vs. host disease (GVHD). When antibody binds to self-antigen as in autoimmune diseases, CD16 on NK cells binds to lgG, which activates lysis by ADCC. GVHD occurs When grafted effector cells mount a rejection response to the host resulting in epithelial cell necrosis. NK cells cannot recognize allotype antigens but transplanted T helper cells do and are stimulated to produce IL-2. In response to IL-2, NK cells differentiate into LAKs that are able to lyse host epithelial cells (1).

NK3.3 Cells
Most studies of NK cells have been performed using partially purified cells from humans or animals since the method of harvesting NK cells by density gradient centrifugation with Ficoll-Hypaque does not yield a pure population. However, research on NK cell biology using partially purified NK cells was unable to rule out cytotoxicity due to other immune cells, eliminate possible cross-talk between other cell types, or solve the problem of collecting an adequate quantity of pure NK cells. As NK3.3 cells are constitutively active due to IL-2 dependence, they were used to show that IL-2 activated NK cells natural cytotoxicity is reversibly inhibited by extracellular nucleotides with increasing negative charge in a manner similar to nonstimulated NK cells. However, when using the adenine nucleotide analog, 5'p(fluorosulfonyl)benzoyladenosine (5' FSBA), cytotoxicity was found to be irreversibly inhibited. This information can be used to discover the role that nucleotide-binding proteins have on extracellular regulation of cytotoxicity of NK cells (19). The natural killer tumor recognition (NK-TR) protein was found to be necessary in the activation of direct cell contact lysis by NK3.3 ceHs (40). Glucocorticoid suppression of NK cytotoxicity was investigated using NK3. In parallel, KIR cross-linking by GL 183 was examined for increased lipid kinase activity of Pl3-K. In these assays, p85a was immunoprecipated from NK3.3 cells that were cross-linked with GL 183 and p85a was assayed for its ability to transfer 32 P to Pl4-phosphate. It was observed that Pl3-K activity paralleled the recruitment of p85a to KIR cytoplasmic tails (34).
Pl3-K is upstream of other signaling intermediates. It was observed that another consequence of KIR cross-linking by GL 183 was the activation of AKT by Pl3-K.
AKT is a serine/threonine kinase credited with anti-apoptotic activity. This result suggested that Kl Rs could also convey positive signals for growth and survival.
Activation of AKT would protect NK cells from apoptosis. This means that not only could KIR participate in inhibitory signaling but also positive signaling (34).
NK3.3 cells have been used to examine cell cycle progression in NK cells.

Biological Response Modifiers
Biological response modifiers are substances made from living organisms that can produce an action or change a condition. Cytokines are one example of a biological response modifier. Cytokines are low molecular weight soluble proteins made and secreted by cells that can influence the function of other cells expressing specific surface receptors. Cytokine receptor binding triggers signal transduction pathways to a.lter gene expression (28,44). Lymphocytes are an important source of cytokines as well as macrophages. Cytokines have functions in the immune system but also have roles in other systems such as developmental biology (23). The term cytokine is used for a large group of diverse molecules. However, cytokines can be categorized into four principal groups by their actions. The first group are those cytokines that mediate innate immunity and include the type I interferons (IFNs) that are the first line of defense for viral infections as well the cytokines which mediate Cytokines are able mediate a broad range of activities in the immune system. NK cells are regulated by specific cytokines and also secrete specific cytokines in response to a stimulus.
The most potent activators of NK cells are IL-2, IL-12, and INF-y. IL-12 was formerly know as NK stimulating factor and is a protective cytokine in infections involving intracellular pathogens. IL-12 activates, induces proliferation, and enhances the secretion of IFN-y and cytotoxicity of NK cells (23,47). IL-12 works best in synergy with IL-2 to activate NK cells (44). IL-2 was previously known as T cell growth factor (47). The major function of IL-2 is to induce immune responses. IL-2 is made primarily by T helper cells after stimulation by MHC class II and antigen stimulation. IL-2's effect on NK cells is to increase lytic activity, induce clonal expansion, and generate differentiation to LAKs (23,44). Interferon gamma (IFN-y) is best known as a major activating factor for macrophages as it increases their metabolism, phagocytic and killing activities, especially those involving intracellular pathogens. IFN-y also increases MHC class I expression and induces MHC class II expression. IFN-y is able to induce differentiation of Beel.I$ and NK cells and affect T cell development. It can increase NK cell adherence to target cells and activate and increase the rate of target lysis (23,44).
To a lesser degree the following cytokines are also known to activate NK cells: IL-1, IL-15, and Type I interferons (a and P) (28). IL-15 has similar effects to IL-2 as it increases proliferation and effector function of NK ceUs. IL-1 is a T cell activator and enhances NK cell activity (23,28). Type I interferons (a and p) are secreted by virally infected cells and stimulate NK cells. This activity of NK cells by IFN-a and IFN-P is the first line of defense to viral and some bacterial infections until T cytotoxic cells differentiate into cytotoxic T lymphocytes (28).
Recent studies with chemokines (chemotactic cytokines) have shown that they too can increase cytotoxicity and direct the migration of NK cells to sites of inflammation and tumor cells (32). Chemokines are low molecular weight heparin binding polypeptides that are released at inflammatory sites and initiate the migration of immune cells. Chemokines are divided into families by the location of cysteine residues in the NH2 terminus: CXC (a), CC (P), C (y), and CX 3 C (8).
Chemokines also mediate expression and/or adhesiveness of leukocyte integrins and are specific for cell types. Chemokine binding to its receptor triggers an activating signal that is mediated by heterotrimeric G-proteins that are associated with the receptor. The signal induced by chemokine binding causes a conformational change in the integrin molecules on the cell membrane. This increases the affinity of intergins for adhesion molecules. Cells will migrate in the direction of high concentrations of chemokines (28,47).
NK cells do not normally recirculate between blood and lymphoid tissues but they can after the administration of biological response modifiers such as chemokines.
LAK cells do not migrate towards tumor sites (32). Chemokines induce chemotaxis of NK cells and mobilization of their intracellular calcium, which is a common second messenger in signaling events. The CC (P) chemokines in particular induce granule exocytosis of NK cells therefore activating and enhancing their cytotoxicity.
Research with chemokines and NK cells suggests that chemokine activated NK cells are bettered suited for tumor therapy than IL-2 activated LAK cells. NK cells have better migratory behavior (extravasate into tissues) than LAK cells which do not respond to chemokines and do not migrate to tumor sites. Moreover, to maintain LAK cells in vivo, a continuous infusion of IL-2 is required which can be toxic in high doses; this is not the case with chemokines. Chemokines attract and activate NK cells at sites of viral infection to lyse virally infected cells. Chemokines also have the ability to inhibit repl.ication of HIV-1 infected cells by binding to chemokine coreceptors. Taken together these actions suggest that chemokines can be used as tools to study signal transduction pathways in NK cells (32).
NK cells must be activated to synthesize and secrete cytokines. The dominant cytokine that NK cells are known to produce is IFN-y. IFN-y is also called immune or type II interferon and is a cytokine that regulates immune-mediated inflammation.
Like type I interferons, IFN-y is antiviral and antiproliferative. However, IFN-y binds to a different cell surface receptor and also has immunoregulatory functions affecting primarily macrophage activation ( 1).
Activated NK cells also produce tumor necrosis factor-a (TNF-a) and granulocyte-macrophage colony stimulating factor (GM-CSF) (55). TNF-a also referred to as cachetin, is a prominent inflammatory cytokine with a wide range of activities. Early studies of this cytokine showed that it could kill tumor cells in tissue.
TNF-a is an endogenous pyrogen, induces leukocytosis, enhances endothelial adhesiveness, and enhances the production of other cytokines. GM-CSF is a cytokine that is important for the development of myefoid lineage-specific bone marrow stem cells. GM-CSF sends strong signals to induce TH 1 differentiation and response (23).
Activated NK cells secrete the following interleukins: IL-1, IL-2, IL-3, IL-4, and IL-12 (28,42,47). The primary function of IL-1 is that of a mediator cytokine of a host inflammatory response in innate immunity (1). IL-1 stimulates T and B cells, travels to the brain to induce fever and augment corticosteroid release, and induces the production of acute phase proteins in the liver. An example of an inflammatory response that IL-1 participates in is the production of prostaglandins and degradative enzymes (collagenase) that are involved in cartilage and bone destruction (47).
With the production of IL-2 and IL-12, one should note that cytokines, like hormones, demonstrate autocrine action, meaning that the cell that produces the cytokine can also be the target of the cytokine. These two cytokines when working together are potent effectors of NK cell and CTL cytotoxic activity. IL-12 stimulates naive T cells to TH 1 response while IL-2 contributes to the growth and NK cells can be stimulated by and produce cytokines to biologically modify the response of the immune system. However, biological response modifiers are perhaps not the only substances to effect on NK cell activity. This research will examine genistein, a phytochemical, which can modify the cytolytic response of NK cells against target cells.

Genistein
Phytochemicals are plant derived bioactive non-nutrients (48). One group within the phytochemicals is the phytoestrogens that have nonsteroidal structure that can act as estrogen receptor agonist or antagonists (48). Two main classes of phytoestrogens are lignans and isoflavones (49). The chemical structure of isoflavones is similar to mammalian estrogen due to the presence of a phenolic ring (49). This phenolic ring is a condition required for compounds that bind the estrogen receptor (ER) (30).
lsoflavones are found in high concentrations in soybeans and soy-derived proteins. The major isoflavones in soybeans are the isoflavone glycoside conjugates genistin and daidzin, and the isoflavone aglycones genistein and daidzein (58).
lsoflavones have been discovered to have a broad range of hormonal and nonhormonal activities that aid in the prevention of hormonal-dependent diseases such as cancer (49). The isoflavone thought to be responsible for these activities is genistein. Genistein is believed to be responsible for anti-estrogenic and anticarcinogenic activity in hormone dependent diseases (59,67). However, for the most part, little is known about the mechanisms by which genistein performs these activities.
Genistein has been shown to be a specific inhibitor of tyrosine kinases. Akiyama, et al, observed genistein to inhibit epidermal growth factor (EGF) stimulated tyrosine phosphorylation of A431 cells. Further, it was seen that genistein competed with ATP and inhibition by genistein binding to enzyme-substrate complexes. This finding may explain why those eel.ls lacking estrogen receptor are affected by genistein (2).
Genistein is believed to stimulate proliferation through estrogen receptor mediated pathways while its anti-proliferative actions are thought to be due to inhibition of tyrosine kinases (59). Two research studies have examined the antiestrogenic effects of genistein on MCF-7 cells, a human breast cancer cell line that is positive for ER (59,67). Both studies examined genistein's effects on ER binding, pS2 mRNA expression, pS2 protein, and cell proliferation. Wang, et al, (59)  The biphasic responses related to concentration were also noted. These studies suggested that genistein acts as an estrogen agonist at physiologic concentrations.
However, at increased genistein concentrations, when proliferation was inhibited, the pS2 protein was present suggesting an ER-independent cellular mechanism(s) may inhibit the ER-dependent proliferation (67).

Signal Transduction during NK Cell Activation
In cell biology, signal transduction is the process by which a cell converts an extracellular signal to produce intracellular signals to cause physiological responses. to be responsible for this activity (9,11,31). In particular, the src-family kinase Lek has been linked to this phosphorylation activity since it has been found to physically associate with FcyRIII complex (17). Phosphorylation of ITAMs by Lek enables syk family kinases (ZAP-70 and Syk) to bind to them (11,31). However, FcR complex signaling can occur in the absence of Lek, Fyn (another src-family kinase) also thought to couple with the receptor, and ZAP-70 (38,57,60). It has also been demonstrated that Syk activation (Lek independent) alone is sufficient to generate a cytotoxic response while ZAP-70 (Lek dependent) requires co-stimulation by a src family kinase (26). These findings prompted research looking into Syk kinase as the signaling element used to tyrosine phosphorylate ITAMs and initiate downstream signaling events. Through the use of pharmacological inhibition of Syk kinase activity (piceatannol, ~ 6 µg/ml), expression of dominant-negative kinase inactive Syk, and tumor cells transfected with MHC class I, it was determined that Syk kinase was an early and central signaling protein in ADCC and natural cytotoxicity (10).
Once the FcR complex gathers its activated kinases, it can activate a number of downstream signaling molecules. One of these substrates is phospholipase C-y (PLC-y) that is an enzyme that hydrolyzes phosphoinositides. The products from that reaction, inositol trisphosphate (IP3) and sn-1, 2-diacylglycerol (DAG), mediate the . increase in free intracellular calcium and the activation of protein kinase C (PKC) (9). The increased concentration of free intracellular calcium is required for the release of granules in the delivery of the "lethal" hit (31 ).

Another downstream effector of FcR signaling is phosphatidylinositol 3-kinase (Pl-3K). It was once thought that cytotoxic activities initiated through FcR ligation
were PKC dependent since PKG-activating phorbol esters and calcium ionophores could induce NK cytotoxicity. However, it was shown that ADCC still occurred even in the presence of PKC inhibitors. It was found that natural lysis was PKC dependent while FcR lysis was Pl-3K dependent (8).
Not only does the Fe-induced tyrosine kinase cascade activate downstream enzymes but it also activates small molecular weight GTP-binding proteins (G proteins). In the case of Ras, a G protein involved with mitogenic signaling pathways, it becomes activated by adapter proteins that have been phosphorylated by proximal PTKs induced by ligation of FcR. The adapter proteins p36 and She are tyrosine phosphorylated by these PTKs and then binds to another adapter protein, Grb2, which binds to Sos, the guanine nucleotide exchange factor for Ras. Sos stimulates inactive Ras to give bound GDP for GTP to become activated (9, 11 ).
Ras activation is believed to regulate transcriptional events that mediate NK cell effector functions following receptor ligation (31). The guanine nucleotide exchange factor (GEF) Vav is tyrosine phosphorylated after FcR ligation after which it activates its downstream effector Rac1 (a small G protein). It has been shown that over expression of Vav enhances NK cell cytotoxicity and that a mutation that prevents Vav from mediating GTP for GDP on Rac1 will abolish this enhancement. Dominantnegative Rac1 has also been shown to inhibit NK cell mediated lysis. This all suggests that Rac1 is a downstream effector involved with regulation of cytotoxicity.
This event is also observed with natural killing of target cells (6).
Signal transduction studies dealing with natural cytotoxicity have been hindered by a clear molecular characterization of a triggering receptor. Natural cytotoxicity like ADCC utilizes protein tyrosine kinase activation, PLC-y release of phosphoinositides, and an increase in free intracellular calcium (20,41,51 ). Because natural cytotoxicity shares these early and requisite events with ADCC there was speculation that a receptor with ITAM subunits could be involved. Research has shown that if this is the case it would not be due to the sandy since NK cells lacking these units can perform natural cytotoxicity (52). And like ADCC, Lek, Fyn, or ZAP-70 are not required for natural cytotoxicity (38,57,60). However, a central role for Syk tyrosine kinase was discovered in natural cytotoxicity (10). Natural cytotoxicity was found to be PKC dependent, Pl-3K independent (8). Vav and its effector target Rac1 were found to have a role in regulating natural cytotoxicity (6). There is evidence that natural cytotoxicity can occur in the absence of PLC-dependent calcium signaling (43,66).

Summary
Innate immunity is the first line of defense of the immune system. Natural killer cells play a role in innate immunity as effector cells. Tyrosine phosphorylation has been implicated as a signal for lytic activity of NK cells (20). In vitro studies with isoflavones, genistein and daidzein, have been shown to enhance NK lytic activity of peripheral blood leukocytes. This study was designed to ask if low concentrations of genistein had a direct effect on natural killer cell lytic activity and if this effect was related to tyrosine phosphorylation. A pure NK cell line, NK3.3, a chromium release assay, and flow cytometry were used to study the direct effects of geniste,in on NK cells.

NK3.3 Cell Stimulation and Inhibition
Human recombinant IL-2 was used in flow cytometry and chromium release assay at concentrations of 2.5, 5, 10, and 15 International Units (IU). Rested NK3.3 cells were incubated with IL-2 for 30 minutes at 37° C, 5% CO2 for the chromium release assay and for 10 or 30 minutes at 37° C, 5% CO2 for flow cytometry. IL-2 was used to restimulate NK 3.3 cells to respond to a natural ligand.

Chromium .Release Assay
The lytic ability of the NK 3.

Fl.ow Cytometry Analysis
Analysis was performed on a XL-MCL (Coulter Corp., Miami, FL). Green fluorescence from FITC was collected through a 525 nm bandpass filter. Samples were gated on forward scatter and side scatter (linear). Ten thousand gated events in a bit map were analyzed for each sample. Data was stored in a listmode file for reanalysis using Elite Software (Coulter Corp).

Studies with "NK-rich" populations
Chromium release assays were performed with NK cells harvested from peripheral blood leukocytes stimulated with genistein and daidzein. Figure 1 shows the average percent cytotoxicity of 8 chromium release assays with baseline set at 100% to standardize results. Baseline cytotoxicity for "NK-rich" samples varied between 13 to 58.8%. Lytic activity was 40% higher than that observed at baseline cytotoxicity. Lytic activity was significantly enhanced at concentrations 0.1 µM to 1.0 µM of daidzein and genistein (p< .001). Lytic activity was significantly suppressed at concentrations 2.5 µM to 10.0 µM. The amount of suppression due to these concentrations was found to be significantly greater with genistein (p < .003). We hypothesized that lytic enhancement was linked to tyrosine phosphorylation. As genistein was known to affect PTK's, we choose to study genistein in subsequent experiments. To see if enhancement of lytic activity was due to specific effects on NK cells, research with genistein was conducted on a NK cell line, NK3.3 cells (27).

Activated NK3.3 cells used to establish chromium release assay parameters
Activity of tumor killing by NK cells can be mea$ured using a chromium release assay. However, the initial rate of lysis for NK3.3 is more rapid than that of freshly determined that cytotoxic activity was also dependent on culture conditions before the assay and that the effector to target (E:T) cell ratio had no effect on cytotoxicity.
As culture conditions were found to have an effect on NK3.3 lytic activity

Genistein's effects on lytic activity of activated and rested NK3.3 cells
To determine the direct effects of genistein on natural killer cells, activated and

Quantitation of Intracellular Phosphotyrosine
Activation of NK cell cytotoxicity is dependent on tyrosine phosphorylation (20).    down (as seen in figure 6) rather than separating into distinct populations.

Intracellular phosphotyrosine levels of activated NK3.3 cells with genistein
To determine if there was a correlation between the concentrations of genistein that enhanced lytic activity and intracellular phosphotyrosine levels, flow cytometry analysis was performed on activated NK3.3 cells after a 10-minute incubation with genistein or genistein plus vanadate (50 µM). Incubation in the presence of vanadate was done to amplify any changes in tyrosine phosphorylation due to incubation with the low doses of genistein. (Vanadate at 50µM was used as it was found to be the concentration for optimal amplification as shown in figure 7). Figure 8   that the amount of suppression due to genistein was the same in samples with or without vanadate. In addition, this data supports the premise that vanadate can amplify changes in tyrosine phosphorylation without affecting the trend of the effect.

Intracellular phosphotyrosine levels of rested NK3.3 cells with genistein
To see if there was a difference between activated and rested intracellular phosphotyrosine levels, rested NK3.3 cells were incubated with genistein or genistein plus vanadate (50 µM) for 10 minutes prior to staining and flow cytometric analysis. Figure 9 shows that in the case of rested NK3.

DISCUSSION
Our studies with NK cells isolated from human peripheral blood showed enhancement of lytic activity in the presence of low doses of genistein and daidzein.
This enhancement was observed specifically at concentrations of 0.1 µM to 0.5 µM with lytic activity 40% higher than baseline cytotoxicity. This is the first time that daidzein has been shown to enhance lytic activity like genistein.
Genistein was studied because a great deal of literature has been written about it being a PTK inhibitor. We were interested to see if this enhancement of lytic activity was linked to tyrosine phosphorylation. Experiments using rested NK3.3 cells, which were designed to mimic the physiological state of peripheral NK cells harvested from blood, not only showed enhancement of cytotoxicity, but also had a corresponding enhancement of intracellular phosphotyrosine residues. Genistein at concentrations of 0.25 to 1 0 µM enhanced cytotoxicity by ~ 13% as determined by chromium release assays. Again this concentration range differs from that observed with "NK-rich" population studies.
Genistein was seen to enhance intracellular phosphotyrosine levels of rested NK3.3 cells across all concentrations. This enhancement was amplified by the presence of vanadate.
Taking all results together, the enhancement of NK3.3 cytotoxicity by genistein appears to be acting through a pathway independent of tyrosine phosphorylation.
Baseline intracellular phosphotyrosine for rested NK3.3 cells is lower than that observed for activated NK3.3 cells. The effect of genistein on NK3.3 cell PTK activity is clearly dependent on the activation state of the cell.
Though rested NK3.3 cells could have lytic activity restored by IL-2, genistein was found to suppress this effect. The increase in cytotoxicity of rested NK3.3 cells due to genistein was less than that when stimulated with IL-2. In addition, incubation of rested NK3.3 cells with IL-2 and genistein demonstrated less cytotoxicity than incubation with IL-2 alone. This would suggest that genistein enhancement is not through the use of IL-2 receptors and/or that it is using a pathway separate from IL-2 activation. Alternatively, different receptors maybe expressed that genistein can act upon depending on the activation state of the cell.
Genistein at low doses on NK3.3 cells showed an enhancement of lytic activity over a broader concentration range than that of "NK-rich" populations. However, this lytic activity was not as great as that seen with the "NK-rich" populations, suggesting that genistein is also acting upon other cell types in the population, which could lead to cross talk between those cells and NK cells.