Phospholipases

Phospholipase A₂ (PLA₂) in lysosomes or bound to cell membranes catalyzes the breaking of the sn-2 bond, facilitating release of AA or other polyunsaturated fatty acids (see Figure 24-2). The enzyme is crucial to regulation of eicosanoid synthesis because it is in the nonesterified state that the polyunsaturated precursors enter into the cascades, leading to eicosanoid formation. Only a scant amount of oxidation at carbon 15 of AA occurs catalytically when the fatty acid is still covalently bound as part of a phospholipid.¹ More than 25 different PLA₂ isoforms have been characterized and grouped based on primary structure, localization, and Ca²⁺ requirements.² Cytosolic PLA₂ group IV is the major catalyst of arachidonate release that leads to PG and LT production. The structure of cPLA₂ reveals a monomeric cytosolic protein with a molecular size of 85 kD. Potent and selective inhibitors of cPLA₂ have been developed.³ Lysophospholipids “left over” after the action of PLA₂ are direct precursors of platelet-activating factor (PAF), a potent mediator of inflammation, which is generated by acylation (addition of fatty acid) in the open sn-2 position of the lysophospholipid. One member of the PLA₂ superfamily, lipoprotein PLA₂, may serve as a marker for cardiovascular risk.⁴

Four distinct types of PLA₂ activities hydrolyze fatty acids esterified at the sn-2 position: Secretory PLA₂ (sPLA₂) has small disulfide cross-linked proteins that require Ca²⁺ in mM concentrations for optimal activity. Cytoplasmic PLA₂ (cPLA₂) has larger proteins requiring Ca²⁺ in µM concentrations, which are AA-selective and can deacylate diacylphospholipids completely, preventing accumulation of potentially toxic lysophospholipids. Ca²⁺-independent PLA₂ (iPLA₂) exhibits specificity for plasmalogen substrates. Finally, PAF acetylhydrolase (PAF-PLA₂) has a series of isozymes specific for short chains.⁴

Under basal conditions, AA is liberated by iPLA₂ and then reincorporated into cell membranes (reacylated). It is unavailable for appreciable eicosanoid biosynthesis. The acylase enzymes competitively inhibit the cyclooxygenase (COX) isoenzymes. After receptor activation and cell stimulation, intracellular Ca²⁺ levels increase and the Ca²⁺-dependent cPLA₂ liberates arachidonate at a rate that exceeds the rate of reacylation, leading to arachidonate metabolism by COX isoenzymes and by lipoxygenases (LO). Initial formation of PGE₂ seems to be due to preferential coupling among cPLA₂, COX-1, and cytosolic PGH-PGE isomerase. More intense inflammation leads to participation of secreted sPLA₂ and amplification of PGE₂ biosynthesis by inducible COX-2. It is an oversimplification to regard availability of AA as the sole rate-limiting step in cellular eicosanoid biosynthesis. The coordinated action of phospholipases and the restricted expression and altered activation of COX isoenzymes are also important.⁵

Phospholipase C (PLC) hydrolyzes the polar head group (e.g., inositol, choline) from phospholipids to yield diacylglycerol (DAG) and the polar head group. Direct protein isolation and molecular-cloning studies have revealed multiple PLC isozymes in mammalian tissues. Phosphatidylinositol-PLC occurs in cytosolic (cPLC) and secreted (sPLC) forms and can be divided into three major classes (PLC-β, PLC-γ, and PLC-δ) based on substrate specificity. PLC with specificity for phosphatidylinositol and phosphorylated phosphatidylinositol is a key component of phosphatidylinositol-mediated signaling pathways. DAG is an activator of protein kinase C (PKC), and rapid production of this lipid by phosphatidylinositol-PLC hydrolysis of the phosphorylated phosphatidylinositol pool is a primary step in signaling. Further AA is made available by the sequential actions of diglyceride lipase and monoglyceride lipase.⁶ PLC with activity on phosphatidylcholine has also been identified. Peripheral blood monocytes from patients with rheumatoid arthritis (RA) exhibit greater PLA₂ and PLC activity than cells from healthy volunteers. PLA₂ concentrations did not correlate with disease activity, but the greatest increases in PLC enzyme activity were seen in cells from patients with the most severe, persistent, proliferative disease, not in cells from patients with the most active disease at the time the cells were studied.⁷ In accord with the need for balance in regulation of the inflammatory response, sPLA₂, numbering nine members in humans, has both inflammatory and anti-inflammatory actions. Thus, group V sPLA₂ counters the inflammatory activity of group IIA sPLA₂ by regulating cysteinyl leukotriene synthesis and promoting immune complex clearance in a murine model of autoimmune erosive inflammatory arthritis. It is of interest that concentrations of IIA sPLA₂ are far greater than those of V sPLA₂ in synovial fluid from patients with RA.⁸ In keeping with the fact that nothing in biology and the pathogenesis of disease is simple is the observation that treatment of RA patients with sPLA(2)-IIA inhibitors has only transient benefit and that sPLA(2)-IIA has activity (induction of COX-2) independent of its enzyme activity.⁹

Hydrolysis of phospholipids by phospholipase D (PLD) produces phosphatidic acid (PA) and the respective polar head groups. The capacity of cells to interconvert PA and DAG through the action of specific cellular phosphatases and kinases (Figure 24-3) suggests that AA release from DAG and a variety of intracellular signaling and protein-trafficking events may be regulated by PLD activity. The targets of phosphorylation within the PLD₂ molecule that are key to its regulation have been mapped. PLD may be activated after or independent of PLC activation.¹⁰

Figure 24-3 Reactions catalyzed by phospholipases C and D, illustrating interconversion of diacylglycerol (DAG) and phosphatidic acid (PA).

The tetraenoic precursor (AA) is the most abundant of the three precursor fatty acids in cells of individuals who eat usual Western-style diets. Metabolites of AA constitute the “2” series (dienoic) PGs (two double bonds in the molecule), and the metabolic pathway has acquired the familiar name “arachidonic acid AA cascade.” However, diets enriched in eicosapentaenoic or gamma-linolenic acid, the other eicosanoid precursors, lead to formation of different eicosanoids. Figure 24-4 illustrates the COX and 5-LO lipoxygenase pathways of the cascade.

Figure 24-4 Cyclooxygenase pathway of arachidonic acid metabolism. MDA, malondialdehyde; PG, prostaglandins; TX, thromboxane.

Cyclooxygenase Pathway

The first step in the biosynthesis of the “prostanoids” (e.g., PGs, TXs, prostacyclin) is catalyzed by the bifunctional PG endoperoxide synthase isozyme (PGHS)-1 (COX-1) and PGHS-2 (COX-2). COX is a homodimeric enzyme that integrates into only a single leaflet of the lipid bilayer of the cell membrane. The COX active site is a long hydrophobic channel. Aspirin and most other nonsteroidal anti-inflammatory drugs (NSAIDs) exclude arachidonate from the upper portion of the channel.¹¹ To form the characteristic five-carbon ring structure (TXs contain a six-member ring), the precursor fatty acids must have double bonds at carbons 8, 11, and 14 (numbering from the carboxyl group). When a molecule of oxygen is inserted across carbons 9 and 11, ring closure occurs enzymatically across C8 and C12, creating the unstable PG endoperoxide PGG. Subsequent peroxidation yields PGH with formation of the cyclopentane ring. PGH serves as the common precursor for PGs, prostacyclin, and TXs that are formed under the influence of terminal synthases (see Figure 24-4). In addition to the activity of phospholipases, regulation of PG synthesis also occurs at the level of PGHS gene expression. PGHS levels are increased by interleukin (IL)-1, platelet-derived growth factor, and epidermal growth factor, agents that increase PG synthesis.

Cell membranes constitute the source of substrate AA and the site of action of eicosanoid-forming enzymes. PG synthesis can also form at lipid bodies, which are non–membrane-bound, lipid-rich cytoplasmic inclusions that develop in cells associated with inflammation. Lipid bodies isolated from human monocytes express PGHS activity, are reservoirs of arachidonyl phospholipids, and can function as domains of PG synthesis during an inflammatory reaction.¹²

PGHS, the well-known target of NSAIDs, exists in two isoforms. These isoforms are similar in terms of amino acid identities (about 60%), catalytic properties, and substrate specificity, but they differ in their genomic regulation.¹³

Prostaglandins

The basic structure of all PGs is a “prostanoic acid” skeleton, a 20-carbon fatty acid with a five-membered ring at C8 through C12 (see Figure 24-4, inset). The term prostaglandins is employed widely but should be used only to describe the oxygenation products that contain the five-membered carbon ring. A family of acidic lipids found first in human seminal fluid, PGs were misnamed because it was thought they were produced in the prostate gland rather than in the seminal vesicles.^28–30 The alphabetic PG nomenclature (e.g., PGE, PGF, PGD) is related to the chemical architecture of the cyclopentane ring. PGE and PGF differ only in the presence of a ketone or hydroxyl function at C9 (see Figure 24-4). These compounds are made by a variety of cells (e.g., PGE₂ and PGD₂ by isomerases, PGF_2α by a reductase). In the nomenclature a subscript numeral after the letters indicates the degree of unsaturation in the alkyl and carboxylic acid side chains. The numeral 1 indicates the presence of a double bond at C13-C14 (PGE₁), 2 marks the presence of an additional double bond at C5-C6 (PGE₂), and 3 denotes a third double bond at C17-C18 (PGE₃).

PGs are produced on demand and seem to exert their effects on the cell of origin or nearby structures. They are not stored in cells and are degraded rapidly in vivo by 15-hydroxyprostaglandin dehydrogenase (PGDH) during one passage through the lungs. Abundant experimental evidence supports the view that PGs participate in development of the inflammatory response. Administration to rats of a monoclonal antibody to PGE₂ prevents carrageenan-induced pain and inflammation.³¹ PGs are probably better at potentiating the effects of other mediators of inflammation than they are at inducing inflammation directly. PGE compounds and intermediate hydroperoxides of AA increase pain sensitivity to bradykinin and histamine. The effects of PGE are cumulative, depending on concentration and time. Even small amounts of PGs, if allowed to persist at the site of injury, may in time cause pain.

PGE₂ stimulates bone resorption,³² and its 13,14-dihydro derivative is nearly as potent, which is of interest because derivatives of the biologically active PGs are usually assumed to be of no functional significance. Addition of serum to the culture medium stimulates bone resorption, a process that is complement dependent and PG mediated. The mechanism may help explain bone erosion in joints of patients with RA, in which complement is activated, and PGE₂ concentrations are high. The observation that PGE₁ can stimulate bone formation³³ suggests that PGs physiologically participate in coordination of bone formation and resorption. For example, primary idiopathic hypertrophic osteoarthropathy, a familial disorder, is associated with mutations in PDGH and subsequent impairment of PG degradation. These patients have chronically elevated levels of PGE and exhibit digital clubbing and evidence of both increased bone formation and resorption in their phalanges.³⁴ Many effects of IL-1 and TNF on cells are associated with stimulation of PG production and inflammation. Cartilage explants from osteoarthritis patients express COX-2 (but not COX-1) and release 50 times more PGE₂ in culture than cartilage from healthy subjects and 18 times more PGE₂ than normal cartilage stimulated with cytokines plus lipopolysaccharide. Explants from osteoarthritis patients release IL-1β, whereas normal cartilage does not express mRNA for pro-IL-1β or release IL-1β spontaneously. It seems that in osteoarthritis—and probably in RA—upregulation of cartilage IL-1β and subsequent production of PGE₂ leads to cartilage degradation.^35,36

Mast cells, often overlooked as important in inflammatory responses, are seen in large numbers in synovium from patients early in the course of RA.³⁷ PGD₂, the major PG formed by mast cells, also can mediate histamine release from mast cells exposed to anti-IgE antibody. PGJ₂, formed from the dehydration of PGD₂, seems to function as a brake on the inflammatory response. It reduces macrophage activation, reduces nitric oxide production from stimulated cells, and induces apoptosis in tumor cell lines. PGJ₂ is metabolized to 15-deoxy-δ12,14 PGJ₂ and δ12 PGJ₂, which are also biologically active.³⁸ It is not clear whether PGA, formed from loss of water from the cyclopentane ring, has biologic activity. However, in vitro, PGA₂ induces apoptosis in HL-60 cells.³⁹

Prostacyclin

Prostacyclin, discovered in 1976,⁴⁰ has been purified, and the cDNA for prostacyclin synthase has been cloned. In addition to a cyclopentane ring, a second ring is formed by an oxygen bridge between carbons 6 and 9. It is generated from PGH₂ by a distinct prostacyclin synthase, a 56-kD member of the cytochrome P-450 superfamily of enzymes found predominantly in endothelial and vascular smooth muscle cells.¹⁵ Production of prostacyclin can be stimulated by thrombin or generated by transfer of PGH₂ from platelets (the endoperoxide steal), contact with activated leukocytes, or stretching of the arterial wall. It is a powerful vasodilator and inhibits platelet aggregation through activation of adenylate cyclase, which leads to an increase in intracellular cAMP. It is metabolized rapidly (half-life in plasma is less than one circulation time) to the more stable, less biologically active 6-keto-PGF_1α. The enzymatic products of its conversion—2,3-dinor-6 keto-PGF_1α and 6,15-di-keto-2,3-dinor PGF_1α—are also chemically stable and have little biologic activity. They are the major metabolites of prostacyclin excreted in urine, in which they can be assayed as indicators of prostacyclin generation.

Prostacyclin generated in the vessel wall has antiplatelet and vasodilator actions, whereas TXA₂ generated by platelets from the same precursors induces platelet aggregation and vasoconstriction. These two eicosanoids represent biologically opposite poles of a mechanism for regulating the interaction between platelets and the vessel wall and of formation of hemostatic plugs and intra-arterial thrombi. Given the central role of platelets in inflammatory reactions, an appropriate prostacyclin-TX balance is important to regulation of inflammation. The balance may be altered in patients with antiphospholipid antibody syndrome; in patients treated with cyclosporine; and in patients treated with NSAIDs, especially patients treated with selective COX-2 inhibitors. Although COX-2 inhibitors reduce recurrence of colorectal adenomas, they increase the risk for cardiovascular events such as myocardial infarction and stroke.⁴¹

Intravascular infusion of prostacyclin also reduces some of the clinical changes associated with pulmonary embolism. The instability of prostacyclin makes it cumbersome to administer therapeutically. Nonetheless, it has been used with limited success to treat peripheral vascular disease including Raynaud’s phenomenon. New therapeutic approaches for treatment of pulmonary hypertension including patients with sarcoidosis and systemic sclerosis are prostacyclin analogues such as epoprostenol, beraprost, and iloprost, but these are not free of adverse events due to intravenous administration.^42,43 Several drugs including angiotensin-converting enzyme inhibitors and statins release endogenous prostacyclin from vascular endothelium in vivo, observations that pave the way for development of new anti-inflammatory and cardioprotective agents.⁴⁴ In addition to its vasodilator effects, prostacyclin suppresses endothelial cell proliferation. Prostacyclin analogues might prove useful as adjunct treatment for some cancers.

Thromboxanes

The endoperoxide PGH₂ can be converted into TXs after the action of the enzyme TX synthase, a microsomal 60-kD member of the cytochrome P-450 family, which is quite active in the platelet. The gene that encodes the enzyme has been cloned. TXs contain a six-member oxane ring instead of the cyclopentane ring of the PGs. TX synthase converts PGH₂ into equal amounts of TXA₂ and 12L-hydroxy-5,8,10-heptadecatrienoic acid. TXA₂ stimulates platelet activation, contributes to intravascular aggregation of platelets, and contracts arteriolar and bronchiolar smooth muscles. It is hydrolyzed rapidly (half-life is 30 seconds) to the inactive, stable, measurable product, TXB₂; its actions are limited to the microenvironment of its release.

The extraordinary rapidity with which platelets adhere to damaged tissue, aggregate, and release potent biologically active materials suggests that the platelet is well suited to be a cellular trigger for the inflammatory process. Efforts directed at suppression of TX synthesis and platelet aggregation may result in limitation of inflammatory responses, especially in coronary arteries. Inhibition of platelet aggregation may be important to the anti-inflammatory effects of aspirin and other NSAIDs. Long-term administration of low doses of aspirin (40 mg/day—the lowest dose predicted to cause total inhibition of TX formation in serum, according to mathematic modeling)—has inhibitory effects on platelet function ex vivo that are indistinguishable from the effects caused by giving 325 mg/day of aspirin.⁴⁵ Aspirin acetylation of COX in platelets occurs in the portal vein where the aspirin concentration is high before it is metabolized in the liver, which explains why the 81-mg dose of aspirin is so effective in prevention of heart attacks and strokes. Platelets lose their ability to aggregate until new platelets are formed in about a week.¹¹ Nonetheless, the response to aspirin can vary among individuals. It has been suggested⁴⁶ that serum TXB₂ levels be monitored to ensure the efficacy of aspirin therapy. Even low-dose aspirin has some depressive effect on prostacyclin synthesis, however. This means recovery of vascular prostacyclin production may be more difficult to attain in elderly patients.⁴⁷

Selective inhibition of TX synthase represents an approach that may be used to suppress TXA₂ synthesis without depressing prostacyclin formation. The endoperoxide steal seems to function in vivo after administration of a TX synthase inhibitor. Antagonists of the receptors shared by endoperoxide and TXA₂ have been developed, and these agents inhibit platelet aggregation in patients who are recalcitrant to TX synthase inhibition. Novel agents targeting TX receptors and TX synthase have improved treatment of vasculitis and cardiovascular and renal diseases.⁴⁸ More specific inhibition of TX action may become possible now that TX receptors have been cloned and characterized.⁴⁹

Lipoxygenase Pathways

In contrast to the COX pathway, in which stable products have three atoms of oxygen covalently attached to AA from 2 moles of molecular oxygen, lipoxygenases insert a single oxygen atom into the molecular structure of AA. Separate lipoxygenases exist in certain cells and have strict structural requirements for their substrates. Three Six major mammalian lipoxygenases insert their oxygen atoms into the 5, 12, or 15 position of AA, with formation of a new double bond and hydroperoxy group. The hydroperoxy fatty acids (hydroxyperoxy-eicosatetraenoic acid [HPETE]) can be reduced by peroxidases in the cell to yield the corresponding hydroxy fatty acids (hydroxy-eicosatetraenoic acid [HETE]). The exclusive lipoxygenase product of the human platelet is 12-HPETE, which on reduction of the hydroperoxy group yields 12-HETE. In contrast, the human neutrophil makes predominantly 5-HPETE, but when high concentrations of AA are added, 15-lipoxygenase can be shown. Lipoxygenases that act on AA are found in the cytosol fraction of cells.

The human 5-lipoxygenase (5-LO) gene has been isolated and characterized^50,51 and produces a 78-kD enzyme. In myeloid cells the 5-lipoxygenase pathway leads to formation of the biologically active LTs (Figure 24-5) that were originally found in leukocytes and that contain three conjugated double bonds (trienes). Cell activation leads to translocation of 5-lipoxygenase from cytosol to the nuclear membrane, where it encounters the 18-kD, 5-lipoxygenase-activating protein (FLAP). AA is also translocated to 5-lipoxygenase-activating protein FLAP for presentation to 5-lipoxygenase (5-LO). In addition, upon cell stimulation, cPLA₂ is activated and also associates with the nuclear membrane, close to FLAP. The ability of macrophages and dendritic cells to respond appropriately during innate immune responses is likely regulated by 5-LO and 12-LO.^52,53 The unstable HPETE is the initial metabolite of each lipoxygenase pathway. HPETE is reduced to the more stable HETE or is converted by 5-LO to LTA₄. LTA₄ can be converted to LTB₄ (in neutrophils and macrophages) or conjugated with reduced glutathione to form LTC₄ (in eosinophils, mast cells, endothelial cells, and macrophages). In contrast to lipoxygenase, which is mainly distributed in myeloid cells, LTA₄ hydrolase (5,12-dihydroxy-eicosatetraenoic acid), a zinc-requiring enzyme that converts LTA₄ to LTB₄, is widely distributed. From the cDNA sequence, it was suggested that mRNA for LTA₄ may have a short half-life, which could account for the properties of extremely rapid production and shutdown of LTB₄ and other eicosanoid biosynthesis.

Figure 24-5 5-Lipoxygenase pathway of arachidonic acid metabolism. HETE, hydroxy-eicosatetraenoic acid; HPETE, hydroperoxy-eicosatetraenoic acid; LT, leukotriene.

LTA₄ can be exported from the cell of origin and converted in other cells by LTA₄ hydrolase to LTB₄. This variation on the endoperoxide steal—perhaps better called transcellular metabolism—also applies to conversion of LTA₄ to LTC₄ by LTC₄ synthase, a glutathione-S-transferase.⁵⁴ Although human endothelial cells do not produce terminal products of the 5-lipoxygenase system, they do generate LTC₄ from LTA₄ provided by neutrophils. LTC₄ and its products, LTD₄ and LTE₄, constitute the biologic mixture previously known as slow-reacting substance of anaphylaxis. LTD₄ and LTE₄ arise from LTC₄ after sequential removal of γ-glutamic acid and glycine from LTC₄. The enzyme γ-glutamyl transpeptidase is present in many cells as part of a complex enzymatic system involved in glutathione biosynthesis and amino acid transport. In many systems the major sulfidopeptide LT has been reported to be LTD₄, rather than the precursor LTC₄. Removal of glycine from LTD₄ results in LTE₄ with concomitant loss of a significant amount of biologic activity. The principal route of inactivation of LTB₄ is by omega oxidation.