PLANIFICACIÓN DE LA GESTIÓN DE CALIDAD

The Yin Van hypothesis for the role of cyclic nucleotides as intracellular second messengers was introduced for the first time in the early seventies (Goldberg et al. 1973). This hypothesis proposed that the reactions of cyclic adenosine-3,4,- monophosphate (cAMP), an increase in the levels of which was known to inhibit platelet aggregation (Mills and Smith 1971), and cyclic guanosine-3,4,- monophosphate (cGMP) were antagonistic. Later in 1977, Katsuki and colleagues demonstrated activation of soluble guanylate cyclase prepared from various tissues by sodium nitroprusside, nitroglycerin and sodium azide. It was postulated that these agents may activate soluble guanylate cyclase via NO which they all released when in solution (Katsuki et al. 1977). Katsuki and co-workers demonstrated that relaxation of bovine tracheal smooth muscle accompanied by an increase in cGMP but not cAMP levels. In 1981, it was found that the nitrovasodilators and NO, which stimulate the soluble guanylate cyclase and increase the levels of cGMP inhibit platelet aggregation (Mellion et al. 1981). Hence the Yin Yan hypothesis was no longer acceptable since an increase in cGMP was also related to inhibition of platelet aggregation.

The last 30 years of research since the initial discovery o f cGMP have shown that cGMP is a physiologically important intracellular second messenger for a variety of first messengers such as hormones, autocoids, drugs, toxins and other regulatory agents. Cyclic GMP achieves its physiological effects by regulating a variety of enzymes and proteins including cGMP-gated ion channels, cGMP regulated phosphodiesterases, and cGMP-dependent kinases.

The early literature about the role of cGMP in the regulation of platelets has been controversial since increases in platelet cGMP levels had been observed in response to agents which either stimulate or inhibit the activation of platelets. However, it is now clear that elevation of cGMP levels in platelets is associated with the inhibition of platelet activation (Mellion et al. 1981; Radomski et al. 1990).

While there is little doubt that the elevation of cGMP causes platelet inhibition, the underlying molecular mechanism of cGMP action has not been clearly elucidated. Most platelet agonists (e.g. thrombin, thromboxane A2, platelet activating factor,

collagen and ADP), activate phospholipase C (PLC) and elevate cytosolic free calcium, resulting in activation of protein kinase C (PKC) and myosin light chain kinase (MLCK), and ultimately in platelet adhesion, aggregation and degranulation (Seiss 1989; Rink and Sage 1990).

Platelet aggregation is the end result of a series of reactions thought to start with receptor-mediated activation of a G-protein (Brass et al. 1986) and the cell membrane associated enzyme, phospholipase C (Boyer et al. 1994), which in turn cleaves phosphatidylinositol 4,5-biphosphate into 1,2-diacylglycerol and inositol 1,4,5-triphosphate. A rise in intracellular calcium concentration follows after inositol 1,4,5-triphosphate mobilises calcium from intracellular stores and protein kinase C becomes activated by 1,2-diacylglycerol (Walter et al. 1993; Siess 1989 reviews). Activation of protein kinase C, as judged by the phosphorylation of its 47 kDa substrate p47 (Nguyen et al. 1991), and an increase in intracellular calcium concentration, have been both closely linked with platelet aggregation and the release of granule content.

It has been shown that nitric oxide donors and analogues of cGMP such as 8-

bromo-cGMP (8-BrcGMP, a cell permeant cGMP analogue) respectively inhibit

the rise in intracellular calcium levels induced by agonist and phosphatidylinositol turn over induced by thrombin (MacIntyre et al. 1985, Nakashima et al. 1986). In addition, these compounds (NaNP and 8-BrcGMP) can inhibit phosphorylation of

both the p47 (PKC substrate) and 20kDa myosin light chain kinase induced by fluoroalaminate (a compound thought to bypass surface receptors to directly activate G proteins), intracellular calcium rise, platelet aggregation and granular content release. These observations suggest that NO inhibits platelet aggregation by a cGMP-mediated effect on intracellular signal transduction rather than by directly interfering with the interaction of agonists or adhesive ligands such as fibrinogen with the platelet surface (Nguyen et al. 1991).

Whether the anti-platelet effect of NO results from cGMP-mediated direct interference with PKC activation and intracellular calcium rise, indirect inhibition of G-protein-phospholipase C interaction, or a joint effect of both cGMP and cAMP-mediated inhibition o f platelet action is unknown. A recent study by Garg and Hassid showed that NO may inhibit functions of 3T3 fibroblasts by a mechanism independent o f a rise in cGMP (Garg and Hassid 1991) raising the possibility that NO may have effects on platelets not shared by cGMP analogues. An elegant study by Nguyen and co-workers showed that NO inhibited aggregation, intracellular calcium mobilisation, PKC activation, serotonin release and phosphatidic acid production induced by either thrombin or U-46619 (thromboxane A] analogue; Nguyen et al. 1991).

These data support the hypothesis that NO induces a rise in cGMP that blocks platelet activation by agonists that depend on the phosphoinositol signal transduction pathway to exert their effects.

3.1.3 Biological significance of NO.

The best examples of biological reactions controlled by NO include vasodilation and regulation of vascular tone, inhibition of platelet aggregation and adhesion, neural transmission and cytostasis.

Nitro- and nitroso-containing vasodilator drugs have been used for many years. Until the early 1980’s, the mechanism of action was completely unknown. The observation that NO and certain nitro-compounds activate guanylate cyclase and stimulate cGMP formation in tissues show that NO and nitroso compounds, that chemically liberate NO, are potent vasodilators (Gruetter et al. 1979). Although NO directly activates cytosolic guanylate cyclase, certain nitroso and nitro compounds activate guanylate cyclase better or only in the presence of added thiols such as cystine or glutathione. The effects of NO on vascular smooth muscle and platelets are attributed to cGMP. Cyclic GMP acts as an intracellular amplifier and second messenger to rapidly lower intracellular free calcium levels and inactivate myosin light chain kinase. These effects are attributed to the action of cGMP dependent protein kinase and protein phosphorylation.

3.1.4 Inhibition of platelet aggregation and adhesion and NO biosynthesis in platelets.

Over recent years the significance of the L-arginine : NO pathway as an important physiological regulatory system controlling platelet adhesion and aggregation has been suggested in a number o f studies demonstrating that platelets may be exposed to the action of NO and possibly produce NO themselves. It was first reported by Ignarro’s group that NO was capable o f inhibiting platelet aggregation (Mellion et al. 1981). Later in 1986 Azuma and colleagues reported that the effluent from a perfused aorta strip also inhibited platelet aggregation (Azuma et al. 1986). A subsequent study by Moncada and co-workers showed that inhibition of platelet aggregation was mediated through the stimulation of guanlylate cyclase and the resultant increased cyclic GMP levels (Radomski et al. 1987a). Release of NO and prostacyclin by the vascular endothelium has an important role in its thromboresistant properties, since both agents (NO and prostacyclin) act synergistically to inhibit aggregation and disaggregate platelets (Radomski et al. 1987b).

It has also been shown that NO inhibits platelet adhesion to collagen fibrils, endothelial cell matrix and endothelial cell monolayers (Radomski et al. 1987c, d), a process mediated via cyclic GMP. Both prostacyclin and NO which are produced by endothelial cells act together to exert a powerful anti-thrombotic action. However, prostacyclin has only a weak inhibitory effect on platelet adhesion, a process which is mediated by cAMP (Radomski et al. 1987c). The anti­ aggregatory effect o f prostacyclin is mediated via cAMP, suggesting that cGMP

regulates adhesion, whereas both cAMP and cGMP regulate aggregation (Moncada et al. 1991 review).

It has recently been shown that platelets themselves generate NO. The L-arginine : NO pathway acts as a negative feedback mechanism to regulate platelet aggregation (Radomski et al. 1990). An increase in cGMP with sodium nitroprusside, with L-arginine and to a lesser extent with L-homoarginine was observed in platelet cytosol. The increase of cGMP by L-arginine was inhibited by L-NMMA and was dependent on NADPH. This demonstrated an L-arginine and NADPH-dependent formation of NO which was inhibited by L-NMMA, providing conclusive evidence for the existence of L-arginine ; NO pathway in platelets (Radomski et al. 1990; Moncada et al. 1991, review). The formation o f NO from L-arginine in platelet cytosol was dependent on the free calcium concentration, since in the absence of calcium (+lmM EGTA) L-arginine (lOOpM) did not induce NO formation. However addition of calcium (0.03-3.00|iM) caused a concentration dependent increase in the rate of NO formation (Radomski et al

1990).

In 1993 Malinski and colleagues showed direct electrochemical measurement of nitric oxide release from human platelets (Malinski et al. 1993). These authors used a porphyrinic microsensor to investigate the release of nitric oxide from human platelets in whole blood and washed platelet suspensions. In the absence of exogenous L-arginine, the amounts of NO generated during aggregation of platelets with collagen (5 and ISpg/ml) were 20 ± 2 and 140 ± 15nM respectively. This corresponded to the release of 4 x 10'^* mol per platelet. However, in the presence of ImM L-arginine, a significant addition o f collagen (1-lOpg/ml)

resulted in a concentration dependent release of NO to a maximum of 10.5 ± 2.4 x lO'^* mol per washed platelet. NO release from washed platelets was 2.5 times higher than those measured in whole blood indicating that NO was probably inactivated more rapidly by chemical reactions in the blood than in platelets (Malinski et al 1993).

Experiments investigating the action of NO, or NO donors, on platelets have so far only been concerned with platelets in suspension. It has been shown that nitric oxide inhibits platelet adhesion and aggregation (Radomski et al 1990; Kowaluk and Fung 1990), but so far a likely mechanism o f this anti-aggregatory effect has not been found. Unlike previous studies, the experiments discussed in this thesis investigate the effects of the addition of NO donors to platelets contacting glass surfaces coated with either fibrinogen, fibronectin or collagen. This allows the influence of NO upon the interaction between platelet receptors and these extracellular matrix molecules to be investigated. A quantitative technique, reliant on computer image processing, has been developed in this work which allows platelets to be directly visualised as they adhere and spread on protein coated surfaces. Assessing the degree of platelet-substratum adhesion in this manner offers an added dimension over measuring the degree of aggregation in platelet suspension. The process o f attachment and spreading can be monitored continuously giving a qualitative impression of shape and behaviour under control and agonist conditions as well as allowing measurement of platelet number and area. Also, observation of platelets attaching and spreading on protein surfaces corresponds more closely with platelet behaviour

in vivo

Method.

See General Methods chapter for preparation of washed human platelets, preparation of chambers and microscopy, method of image acquisition and analysis, treatment of glass coverslips with protein solutions and statistical test used to calculate the significance values.

To study the anti-adhesion/anti-spreading effect o f nitric oxide appropriate concentrations of the NO donors, s-nitroso-glutathione (GSNO), sodium nitroprusside (NaNP) and s-nitroso-acetylpenicillamine (SNAP), were prepared in fresh Tyrode’s buffer (pH 7.40) and mixed with 1ml of platelet suspension (2 x 10* cells/ml) just before incubation on protein coated substrates at 37°C.

When the effect of nitric oxide donors on platelet attachment and spreading was to be analysed, five fields were chosen 5 minutes after the addition of platelets to the chamber and the same five fields were monitored again at 10, 15 and 30 minutes. When NO donors were added to already attached cells or where cells were pre­ treated with anti-GPIIb/IIIa again 5 fields were first chosen at random and monitored before and after the addition of NO donors.

The NO donors were added 14 minutes after platelet incubation at 37°C and any effect was observed 1, 6 and 11 minutes after NO exposure (15, 2 0 or 25 minutes

total incubation time respectively).

In another set o f experiments, the NO donors were introduced 29 minutes after the incubation at 37°C and the effect was investigated 1 and 11 minutes after NO donor exposure (30 and 40 minutes total incubation time respectively). In the experiments where the antibody was used, 1ml platelet suspension was incubated

with Zpg anti-GPIIb/IIIa for 30 minutes at 37°C before the addition of cells to the protein coated substrates.

To study the effect of divalent cation chelation on platelet attachment and spreading, platelets were incubated with 1 pM EDTA for 3 minutes at 37°C and the observations were made 30 minutes after incubation of cells on the protein substrates or uncoated coverslips. Platelets were prepared in Ca^^/Mg^^ free Tyrode’s buffer in these experiments.

To investigate the effect of oxy-haemoglobin on NO donors, lOpM oxy- haemoglobin was mixed with 1ml platelet suspension. This was then followed by the addition of NO donors just before the introduction into the protein coated chamber. The effect of oxy-haemoglobin on NO donors was studied at 5, 10, 15 and 30 minutes after the addition of mixture (cells + Oxy-Hb + NO donor) to the chamber.

Results.

3.2.1 Characteristics of platelet attachment and spreading.

1ml of washed human platelets suspended in Tyrode’s solution was added to a protein coated well kept at 37°C. Platelets attached and spread rapidly on the protein coated substrates within the first few minutes after addition to the well. Interference reflection microscopy (IRM) showed that within 5 mins o f contacting the surface, most cells changed from a rounded form in suspension into thin lamella forms -10pm in diameter. Spreading was accompanied by the development of dark peripheral zones with bright centres. Plate 3 .1 shows an image of platelets spread on a 1 mg/ml fibrinogen coated glass surface 30 mins after addition to the chamber. Spread platelets show black circular rings around the centre (indicated by arrows). The area inside the fringes corresponds to the central region of platelets containing granules (granulomere). It is likely that these fringes are due to a large change in the thickness from the thin cytoplasmic region (hyalomere) towards the thicker central granulomere region (Gingell 1981). The black circular fringes therefore indicate the edge o f the central granulomere region. As platelets spread fully, the central granulomere region becomes bright. This suggests that the central part of the platelet membrane and its subjacent microfilamentous network are separated from the surface by more than lOOnm (Park and Park 1989).

Single, typical fields were followed for controls on all proteins in order to visualise the process of platelet attachment and spreading. The effect of the nitric oxide donors on this process was also observed by following single typical fields for each NO donor. In the ensuing sections, the appearance of platelets as they adhere to the

Plate 3.1. Interference reflection microscopy (IRM) image of human blood platelets attaching and spreading on fibrinogen coated glass incubated at 37°C for 30 mins. This image shows the typical shape of platelets attached and spread on fibrinogen. IRM helps reveal the pattern o f contacts made between cells and the glass. On fibrinogen, after 30 minutes, well spread cells appear discoid with bright central zones surrounded by dark fringes. These bright zones correspond to areas where the cell membrane is separated from the surface by more than lOOnm. The dark zones generally correspond to areas of close apposition. Bar = 1 pm.

'9

surface in single experiments is first described, followed by a summary of data collected in several (usually four) equivalent experiments.