level. The amount of calcium in the serum (2.5 mM), however, is even greater than what is found in storage buffer (1 mM). These experiments should, therefore, be repeated in the presence of physiological and various higher levels of calcium and other ions found in plasma (Ca++, 2.5 mM; Mg++, 0.75 mM; K+, 4.8 mM) and should be repeated using physiological saline, to exclude the effects of these ions. This would finally establish whether extracellular calcium or other ion levels are responsible for the TIMP-l secretion seen in this study. It would be unlikely that secretion of a granule would be regulated by the extracellular levels of an ion, but it is always possible and should be investigated.
Further studies should also be performed using different PMN agonists, such as £MLP, and the triggering of other complement receptors to try to identify the intracellular signal transduction pathway by which TIMP-l is released. This knowledge is, in any case, very important and is required for prediction of how PMN responses may be manipulated for different purposes. i.e. to block the release of some granules while triggering others as this could be of great importance in handling inflammatory disease and invasive cancers.
The mechanism by which increased secretion of MMP-9 was induced also needs to be checked. PMA was included in studies to check whether MMP or TIMP-l secretion was upregulated via a PKC-dependent mechanism. It is now realized that the inclusion of PMA in the latex bead experiments may be very useful as PMA, at different levels, may enhance or depress the expression of certain PMN receptors. Clq binding and stimulation of MMP-9 release may be confirmed by stimulation of PMNs with PMA as incubation of PMNs with PMA at concentrations of < 10 ng/ml may cause PMNs to shed their Clq receptors but increase the surface expression of CRI and CR3.
Incubation with PMA at a concentration of> 10 ng/ml, causes a decrease in Clq and CRI receptors but not CR3 receptors (Eggleton et al., 1994). If greater than 90 ng/ml of PMA was used in the current study CR3-mediated granule release would have been triggered. IfMMP-9 and TIMP-l secretion was not induced by the opsonin (which would be C3b) in the case of the CR3 receptor, this would imply that C3b was either not assembled, as assembly stopped at Clq, or that the CR3 receptor does not trigger the
release of MMP-9 or TIMP-1. If no MMP-9 release was seen when Cl q receptors were shed (due to high levels of PMA). This would support a C1q receptor-mediated release ofMMP-9. Another control which may have been included to check that MMP-9 release occurred via a C1q-mediated receptor pathway and not due to the assembly of other complement factors, would have been to include rosmarinic acid. Rosmarinic acid is an organic substance which contains four hydroxyl groups which compete for binding for the receptive surfaces on the microorganisms (with C4 and C3) (Sahu et al., 1999). If, after addition of rosmarinic acid, MMP.:;9 is still secreted this would further indicated the involvement of C1q. However, C4 may also bind through amine residues (Sahu et al., 1994) so other alternatives, such as the use of specific inhibitors for signalling pathways, should perhaps also be used to verify the binding of C1q and signal transduction pathway triggered.
Why active MMPs are required for killing of microorganisms, however, is still at this stage unclear. Active MMP-9, is most likely to be involved in limited cleavage of the extracellular matrix allowing movement towards a site of infection, through the barrier ECM (Price et al., 2000). This activity would have to be tightly regulated by the almost simultaneous release of TIMP-I as MMP-9 may cleave many other proteins which have collagen-like domains. Cleavage of Cl-inhibitor by PMN MMP-8 and MMP-9 has previously been shown (Knauperet al., 1991). The release oflow levels ofTIMP-1 may be required to inactivate the low levels of MMPs present under most conditions, allowing the Cl complex to be dissociated by Cl-inhibitor, C1q binding to PMN receptors, phagocytosis and respiratory burst, all important processes for the removal and killing of microorganisms. Cl q can also be degraded by elastase (Ruiz etal., 1995).
Since the fragments of Cl q can trigger an increased generation of superoxide production (Ruiz etal., 1995), stimulation of elastase and MMP release can also potentially increase the effectiveness of respiratory burst and hence killing of microorganisms.
This study seems to indicate the importance of serum factors (possibly complement) in the release of MMP-9. The role of MMP-9, and TIMP-1 in the killing of microorganisms seems still very unclear, however, and needs further investigation. A
study of the effects of MMP-9 and TIMP-1 release on opsonisation and subsequently respiratory burst should be pursued, both using polystyrene beads, and in the presence of bacteria. Where beads are replaced by microorganisms a different mechanism of release of MMP-9 and TIMP-1 could be brought about by bacterial interference with PMN signal transduction. In the absence of such experiments no conclusions on this subject can be drawn, however. It is known that certain microorganisms such as the pseudomonas species secrete elastases which may activate MMP-9 (Okamato et al., 1997) and may also inactivate Cl q-inhibitor, preventing phagocytosis which may occur via a Cl q-mediated mechanism. TIMP-l also does not inhibit bacterial collagenases (Hayakawa et al., 1992). Bacterial components may, therefore, induce the differential release of MMP-9 and TIMP-l and result in different outcomes.
The effect of MMP-9 and TIMP-l release on phagocytosis and respiratory burst using purified proteins and checking the activity of MM Ps after PMN stimulation and inhibitor secretion, is still untested and should be further investigated. The inactivation of Cl- inhibitor under such conditions may also be assessed. Knowledge of whether TIMP-l is totally inhibitory to MMP activity under various conditions would be useful in understanding the role of TIMP-l release in facilitating infection and preserving anti- inflammatory conditions. Even though the complement system has such an important task in the clearance of microorganisms, as has been discussed, its components may be inactivated by some PMN enzymes and the release of these enzymes may also have both adverse and beneficial effects. It is recommended that further tests be carried out in a reductionist fashion using proteins in pure forms, to carefully further explore these effects.