MATRIX METALLOPROTEINASES (MMPS)
 

Biotrak Tutorials   4. Analytes 4.1: Prostaglandins 4.2: Leukotrienes 4.3: Rat Hormone Assays 4.4: Reference and further reading 4.5: Steroids 4.6: Cytochrome P450 4.7: Reference and further reading 4.8: Cytokines 4.9: Adhesion Molecules 4.10: Signal-transduction Assays 4.11 Cell Proliferation Immunoassay 4.12: Matrix Metalloproteinases (MMPs) 4.13: Cardiovascular Peptides

MATRIX METALLOPROTEINASES (MMPS)

1. THE FAMILY OF MMPS AND THEIR ROLE

Many normal physiological and pathological processes such as embryogenesis, morphological growth changes, ovulation and pregnancy, wound healing, atherosclerosis, inflammation, tumor invasion and metastasis involve breakdown and remodeling of the extracellular matrix. This degradation is due to the family of important enzymes known as the matrix metalloproteinases (MMPs).

There are now at least 16 MMPs identified (Table 4). They are quite a homologous family due to the high degree of similarity in their domain structures. They can be further classified into substrate specific groups although the distinctions are becoming less clear as more becomes known about them. The main types of MMPs can be categorized as collagenases, stromelysins or gelatinases.

Table 4 Members of the MMP family

MMP No.	Human Enzyme	Latent MW	Active MW	Substrates
MMP-1	Interstitial collagenase-1	55,000	45,000	collagens, gelatin, proteoglycan, etc.
MMP-8	Neutrophil collagenase	75,000	58,000	"
MMP-13	Collagenase-3	60,000	48,000	collagens, gelatin, PAI2, aggrecan
MMP-18	Xenopus collagenase	55,000	?
MMP-3	Stromelysin-1	57,000	45,000	collagens, gelatin, aggrecan, fibronectin, laminin, casein, MMPs
MMP-10	Stromelysin-2	57,000	44,000	"
MMP-11	Stromelysin-3	51,000	44,000	a 1-proteinase inhibitor
MMP-19		?	?
MMP-20	Enamelysin	?	?	Amelogenin
MMP-2	72 kDa gelatinase	72,000	66,000	gelatin, collagens, fibronectin, MMPs
MMP-9	92 kDa gelatinase	92,000	86,000	gelatin, collagens, fibronectin
MMP-7	Matrilysin	28,000	19,000	as MMP-2
MMP-12	Macrophage metalloelastase	54,000	45,000/22,000	collagen, gelatin, elastin
MMP-14	MT-MMP-1	66,000	56,000	collagens, casein, fibronectin, MMP-2,13
MMP-15	MT-MMP-2	72,000	?	MMP-2, gelatin, fibronectin
MMP-16	MT-MMP-3	64,000	52,000	MMP-2
MMP-17	MT-MMP-4	?	?

 

There is now evidence that a fourth group, the so-called membrane-type or MT-MMPs, can be added to this list.

Several collagenases have now been identified—the more ubiquitous fibroblast collagenase (MMP-1), neutrophil collagenase (MMP-8) and more recently MMP-13 which may be involved in breast tumors and may also represent the principal murine collagenase. They are unique in their ability to degrade collagens.

Stromelysins have a broader substrate specificity, including proteoglycans collagen type IV and IX, fibronectin and laminin, and they can activate MMP-1. The archetypal enzyme is stromelysin-1 (MMP-3) which has strong homology to stromelysin-2 (MMP-10). Also included in this group is matrilysin (MMP-7) which has a quite distinctive truncated domain to the other MMPs and may represent the most evolutionary primitive MMP.

The gelatinases are 72 kDa gelatinase (gelatinase A, MMP-2) and 92 kDa gelatinase (gelatinase B, MMP-9). They degrade denatured collagens (gelatins), type IV basement membrane collagen, elastin and fibronectin.

MT-MMPs have an additional transmembrane sequence to anchor them to the cell and are thought to be involved in the activation of other MMPs.

These very potent enzymes have their activity tightly controlled at the transcription and activation stages to prevent inappropriate matrix degradation. Furthermore, specific inhibitors, the tissue inhibitors of metalloproteinases (TIMPs) bind to the active MMPs in a 1:1 ratio to form inactive complexes. So far, four TIMPs have been identified, with again, a high degree of homology.

The MMP activation cascade is shown schematically in Figure 5. In principle, this applies to all the known MMPs. The enzymes are secreted as inactive (latent) proMMPs, which require another agent to cleave a propeptide sequence and/or perturb the conformation leading to an unstable intermediate. This autocatalytically processes itself to the fully active enzyme by further propeptide sequence cleavage. Residues such as cysteine that are important in bonding the active site Zn²⁺ , are removed during this process to reveal the active catalytic site. The fully active enzyme is then capable of degrading the appropriate matrix substrates. However, it is then also open to inhibition by high affinity (10 ^-10 M) binding of TIMPs.

MMPs can be activated in vitro by proteinases (e.g. trypsin and plasmin), mercurials (e.g. 4-aminophenylmercuric acetate, APMA), and in vivo by proteases such as plasmin and other MMPs.

The figure shows TIMP-1 and 2 forming complexes with the active MMP. However, as far as is known, all active MMPs are capable of forming complexes with all the TIMPs. There are two particular additions to this situation in that proMMP-9 and proMMP-2 (latent zymogens) can form complexes with TIMP-1 and TIMP-2 respectively. It is thought the TIMPs may play a role in stabilizing the proenzyme to autoactivation but also participate in the activation process. In addition, the non-specific protease inhibitor a₂ -macroglobulin (a ₂ M) can also entrap active enzyme providing a further inhibitory mechanism. These all serve to tightly regulate MMP activity.

Due to their implication in many disease states the MMPs are a potential target for therapeutic intervention and this is currently an active area of drug development for several pharmaceutical companies. Most approaches are based on the use of low-molecular weight inhibitors of the activity of the MMP. For example, British Biotech currently have Marimastat™ in clinical trials for cancer treatment which is an MMP inhibitor. The balance between MMPs and their inhibitors is thought to be altered in disease states and the ability to measure their relative levels is crucial to the investigation of their role in these conditions.

2. MEASUREMENT METHODOLOGY

It is only recently that immunoassays have been applied to the measurement of MMPs. Typically they have been measured by substrate cleavage assays and/or zymography, which rely on the enzyme's biological activity in degradation. However, these techniques do have several disadvantages. Substrate cleavage assays e.g. using radiolabeled collagen as substrate, can lack convenience, specificity and sensitivity, and are prone to interference from TIMPs which, by definition, will inhibit enzyme cleavage. Zymography can also be variable and labor-intensive, and is only semi-quantitative.

At Amersham Pharmacia Biotech we have developed the first commercially available immunoassays to a range of MMPs and TIMPs. The assay methodology is conventional "sandwich" ELISA using HRP as detection enzyme and a colorimetric signal. Some of these assays have been developed in collaboration with Fuji Chemical Industries of Japan who also had ELISAs available for sale, mostly in Japan. However the Fuji assays use a large plastic bead as the solid-phase rather than a microtiter plate.

An ELISA is clearly measuring something different to that which is detected in "bioassays". An ELISA is a quantitative measure of the amount of MMP or TIMP present and may not provide information on how much "activity" ; there is. One of the main challenges for manufacturers/ developers of kits and for customers starting to use these assays is to define what is being measured.

The expression and activation of MMPs is complex with several different MMPs being produced at the same site or by the same cell. The MMPs may interact to activate each other as well as displaying their matrix substrate specificities. For example, MMP-3 can activate MMP-1 to a highly activated form. In addition, the inibitory TIMPs are found in close proximity and again can be produced by the same cells. Therefore it is necessary to specifically identify the activity of each MMP and their inhibitors that may be present in a mixture of MMPs in order to fully understand their action.

Furthermore, the complicated activation and subsequent inhibition mechanisms of the MMPs can cause several forms of these enzymes to be present in a single sample. As a result, researchers studying MMPs are often perplexed as to what to measure and how best to do it.

So, before considering an ELISA system to measure MMP, one should ask:
1. In what form are the MMPs present in the sample?
2. Is the assay's specificity appropriate for this form?

There are potentially several forms in which a given MMP can exist given its activation and inhibition cascade. However, the cross-reactivity profile of an assay can be defined quite accurately for a given ELISA. It is very important to define the cross-reactivity for these forms and not just related MMPs. An example of the cross-reactivity profile for the Amersham Pharmacia Biotech MMP-3 assay is shown as an example (Table 5). In this particular case the assay cross-reacts virtually equally with pro, active and complexed active forms of MMP-3 i.e. it can be regarded as measuring "total" MMP-3 levels. Another particularly common profile found is for an assay to measure the pro form only, with no cross-reaction with active MMP. TIMPs are interesting in that potentially their ELISA reactivity when complexed to some MMPs may not be equal to that when they are free posing the question as to which form is actually the most significant in a given sample.

 

---

Table 5 Cross-reactivity of MMP-3 ELISA (APB)

Compound	% Cross-reactivity
MMP-3 Standard	100.0
Activated MMP-3*	96.45
MMP-1	<0.118
MMP-2	<0.118
MMP-9	<0.118
TIMP-1	<0.118
TIMP-2	<0.118
MMP-3/TIMP-1 Complex	112.68
MMP-3/TIMP-2 Complex	90.10

Note: For certain compounds the measured concentrations were below the sensitivity (detection limit) of the assay. Hence the assay sensitivity value was used in the calculation of cross-reactivity.

* APMA activated

 

Therefore the other element required to judge assay suitability is a knowledge of what form of MMP may be present in a sample. The ELISAs themselves can help answer this. For example, it has been shown that MMP-9 is elevated in the plasma of hepatocellular carcinoma (HCC) patients. By using gel filtration and a variety of assays and antibodies it was shown that the predominant form of plasma MMP-9 immunoreactivity is proMMP-9 complexed to TIMP-1. In this case the Amersham Pharmacia Biotech ELISA measures proMMP-9 and so would be appropriate for such samples.

3. SAMPLE LEVELS

Research assays are applied to many more sample types than clinical assays. These assays have been validated for use in serum and plasma as they are commonly found sample types. Typically for MMPs, this has involved testing for linearity of dilution and recovery values and a small-scale estimation of the levels likely to be encountered in normals. One note of caution about recovery testing should be that researchers who wish to do this should ensure that, for spiking, the pro-MMP is as pure as possible. Active MMPs are likely to be trapped by a ₂ -macroglobulin which is present in great excess in serum. Even using pure proMMP is no guarantee of successful recovery because serum proteases may well occur.

Normal levels of MMPs and TIMPS in serum and plasma have been determined as mentioned above. The results obtained are for guidance only and are shown in Table 6.

 

Table 6 Ranges of normal levels of MMPs and TIMPs in serum and plasma (values as ng/mL)

 

	Sample type
MMP	Serum	EDTA plasma	Citrate plasma	Heparin plasma
MMP-1	ND	ND	ND	ND
MMP-2	470–800	365–649	NT	NT
MMP-3	28–99	17–41	23–79	25–136
MMP-9	ND	ND	ND	ND
MMP-8	40–57	1.4–10.3	1.0–1.8	1.4–4.7
MMP-1/ TIMP-1 complex	ND	ND	ND	ND
TIMP-1	280–520	142–198	64–100	104–162
TIMP-2	29–108	21–108	NT	NT

ND – Not detectable
NT – Not tested
 

There is little indication at this stage that any sample type should be avoided for the MMPs. Serum MMP-8 is higher than plasma due to neutrophil release during clotting. For TIMP-1 the serum level may be higher and more variable due to platelet lysis than that found in plasma. It has also been reported that TIMP-2 levels are elevated in heparin plasma although the reason for this is unclear at present.

Birkedal-Hansen, H. Proteolytic remodelling of extracellular matrix. Current Opinion in Cell Biology7, 728–735 (1995).

Cawston, T. Mol. Matrix metalloproteinases and TIMPs: properties and implications for the rheumatic diseases. Molecular Medicine Today , 130–137 (March 1998).

Fujimoto, N., Hosokawa, N., Iwata, K., Shinya, T., Okada, Y. & Hayakawa, T. A one-step sandwich enzyme immunoassay for inactive precursor and complexed forms of human matrix metalloproteinase 9 (92 kDa gelatinase/type IV collagenase, gelatinase B) using monoclonal antibodies. Clin. Chim. Acta 231, 79–88 (1994).

Jung , K., Nowak, L., Lein, M., Henke, W., Schnorr, D. & Loening, S. A. Role of specimen collection in preanalytical variation of metalloproteinases and their inhibitors in blood. Clin. Chem. 42 , 2043–2045 (1996).

Jung , K., Laube, C., Lein, M., Lichtinghagen, R., Tschesche, H., Schnorr, D. & Loening, S. A. Kind of sample as preanalytical determinant of matrix metalloproteinases 2 and 9 (MMP-2; MMP-9) and tissue inhibitor of metalloproteinases 2 (TIMP-2) in blood. Clin. Chem. (in press).