EDTA/gelatin zymography method to identify C1s versus activated MMP‐9 in plasma and immune complexes of patients with systemic lupus erythematosus

Abstract Gelatin zymography analysis is a sensitive method and commonly used to characterize and quantify the presence of the gelatinases (MMP‐2 and MMP‐9) in biological samples. In human plasma samples from healthy controls and systemic lupus erythematosus (SLE) patients, we observed a gelatinolytic molecule at 80 kDa, suggestive for activated human MMP‐9. However, by developing and using the EDTA/gelatin zymography method and after purification of the 80 kDa entity, we proved that this molecule was the C1s subunit of the complement system. The zymolytic capacity of C1s was validated and found to be enhanced, in the absence of calcium and in the presence of EDTA. Our findings indicate that for correct identification of gelatinolytic proteins in complex biological samples the use of EDTA/gelatin zymography for enzyme development is advised. In addition, by quantification of EDTA/gelatin zymography analysis and ELISA, we observed that the levels of C1s were higher in plasma and immune complexes of SLE patients than of healthy individuals. Therefore, our data imply that C1s may become a marker for the diagnosis of SLE.

mediators such as cytokines, chemokines and matrix metalloproteinases (MMPs). [4][5][6][7] The complement system consists of more than 30 proteins that are soluble or covalently attached to target (cell) surfaces. 8 It regulates many biological functions such as activation of B lymphocytes, degranulation of mast cells and granulocytes, solubilization and clearance of IC and membrane attack-complex-associated cell lysis. 8 Although the activation of the complement system is important to clear infections, it also contributes to the inflammatory response triggered by IC deposition in tissues in autoimmune diseases, thereby playing an important role in SLE. [9][10][11] The classical complement activation pathway forms a bridge between adaptive and innate immunity. The first component C1q binds to antibodies within IC. C1q binds with two zymogens of C1r and two of C1s. Once the C1q is bound to IC or pathogens, the complex undergoes a conformational change, which leads to C1r activation, that immediately activates C1s, generating an active serine protease. 12 Thereafter, C1s activates the next two components of the classical pathway C4 and C2, which at the same time will cleave C3 into C3a and C3b. Most of these complement components are dysregulated in SLE. 9 Genetic deficiencies of C1q, C1r, C1s, C2, C4, and C3 are associated with SLE development. In these cases, the disease usually starts early in life, has an equal incidence among males and females and is characterized by multi-organ involvement with glomerulonephritis in approximately 30% of the cases. 9 The levels of complement molecules and their functional activities in serum are used as markers of disease activity in SLE.
During periods of active disease, serum complement activity is usually reduced. Autoantibodies against complement system components are detected in SLE patients. 13,14 Interestingly, autoantibodies against C1s are present in SLE patients and enhance the proteolysis/enzymatic capacity of C1s, thereby inducing C4 activation and the complement system. 14 Matrix metallopeptidase 9 (MMP-9), also called gelatinase B, mediates vital physiological processes. The dysregulation of this molecule is associated with several pathological conditions, including SLE, 15 although its role in SLE is still unclear. The data in the literature are not consistent, because researchers utilized different sources and techniques to study MMP-9 in relation with SLE. Some groups measured MMP-9 levels in plasma or serum samples by ELISA, 16,17 whereas others studied the levels of MMP-9 secreted by peripheral blood mononuclear cells. 18 In addition, others measured MMP-9 levels by gelatin zymography. 19,20 Because of discrepancies about the pathogenic role of MMP-9 in SLE and the finding of opposing results, we hypothesized that some results of the used analytical methods might have been misinterpreted. Gelatin zymography is the most sensitive method to detect gelatinases. However, other proteases, including the serine protease trypsin, cleave denatured collagen or gelatin. 21 In zymography analysis, information about molecular weights is used to identify MMP-9 forms, including multimeric MMP-9, NGAL-MMP-9 complexes, proMMP-9, activated MMP-9 (actMMP-9) and differences in glycosylation. Other proteases, co-migrating with such MMP-9 forms, maybe misinterpreted as MMP-9 forms. We analyzed the gelatinolytic banding patterns of plasma samples from healthy controls and patients suffering from SLE. Surprisingly, by introduction of EDTA/gelatin zymography, we discovered the presence of complement C1s zymolysis at around 80 kDa, which might be misinterpreted as actMMP-9. We also provided the methodology to detect C1s and to discriminate it from actMMP-9. We observed that C1s levels were significantly higher in plasma and IC from SLE patients than in control samples and this finding might be relevant for SLE diagnosis and staging.

| Plasma samples
Blood samples from healthy controls and SLE patients were centrifuged at RT and 1500 g for 5 minutes. The resulting supernatants were used as plasma samples. These were collected and stored at −80°C in several small aliquots for optimal conservation. All donors gave written consent and all procedures were according to the terms of the declaration of Helsinki and following Belgian and European legislation. In Table S1 information about the SLE patients is provided. All patients suffered from clinical SLE symptoms and were under various anti-inflammatory treatments at sampling. The control samples came from healthy donors (3 males, 17 females).

| Gelatin zymography
The proteins were separated in 7.5% polyacrylamide gels containing sodium dodecyl sulfate (SDS) and 1 mg/ml gelatin. 21 After electrophoresis, the gels were washed with 2.5% Triton X100 to remove the SDS and incubated overnight in 50 m mol L −1 Tris-HCl pH 7.5, In the EDTA/gelatin zymography analysis, the two visualized bands contained the target protein(s). In the center of the outlined zone, faintly stained bands were seen. The bands were carefully sliced out and analyzed by nanoLC/TOF/MS analysis after in-gel trypsin digests.

| nanoLC/TOF/MS and protein identification
The nanoLC/TOF/MS and protein identification were done by Alphalyse (Alphalyse A/S, 5220 Odense, Denmark). Briefly, the protein samples were reduced and alkylated with iodoacetamide, ie, carbamidomethylated, and subsequently digested with trypsin. The

| Enzyme-linked immunosorbent assays for C1s
For ELISAs, 96 well plates were coated with 2 μg of sheep antihuman C1s antibody. After washing and blocking the plates, the studied samples were added and incubated overnight at 4°C.
The plates were then washed and the samples were incubated with the C1s detection antibody (of mouse origin). After washing, a specific secondary anti-mouse IgG-HRP-conjugated antibody was added and incubated in the plate for 1 hour at RT. Finally, the plate was washed and developed.

| Statistical analyses
Statistical analysis were performed using GraphPad Prism 6 software. Significant differences between experiments were evaluated using a non-parametric ANOVA Kruskal-Wallis test. All P values of 0.05 or less were considered significant.

| RESULTS
3.1 | Identification of a gelatinolytic 80 kDa molecule in human plasma samples, which is not MMP-9 With the intention to study the levels of MMP-9 in plasma samples from SLE patients, we performed gelatin zymography analysis on 2 μl of plasma samples. Gelatin zymography revealed several gelatinolytic proteins of different molecular weights ( Figure 1A). High molecular weights bands (>250 kDa) represented MMP-9 trimers and complexes of MMP-9/-2 with α2-macroglobulin. The NGAL-MMP-9 complex was detected at 150 kDa and proMMP-2 at 72 kDa. In the electrophoretic mobility range of proMMP-9 and its activated (actMMP-9) forms, approximately between 95 and 80 kDa, respectively, a collection of zymolytic bands was visible. To distinguish the presence of MMP-2 or MMP-9 from non-metalloproteases, we incubated gelatin zymography gels in the presence of EDTA to inhibit metalloproteinase activity ( Figure 1B). A major 80 kDa band was not inactivated by EDTA. This molecule was barely visible in normal plasma, but was abundant in SLE patient plasma samples ( Figure 1B). The dark protein bands observed in the gels mainly corresponded to immunoglobulins and IC staining with Coomassie blue.
To confirm the different nature of this EDTA-resistant enzyme, we examined its gelatin binding property to gelatin-Sepharose (Figure 1C). As expected, proMMP-9 (92 kDa), actMMP-9 (80 kDa) and MMP-2 (72 kDa) were detected in the bound fractions. However, in the unbound fractions, a protease with a molecular weight, which might be misinterpreted as the activated form of MMP-9 (80 kDa), was detected ( Figure 1C) in SLE patients P1 and P2. Quantification of this specific zymolysis, in the presence of EDTA, showed significantly increased levels of this protease in plasma samples of the SLE patients analyzed compared with the healthy controls ( Figure 1D).

| Purification and identification of the 80 kDa protease
Since the levels of the unknown 80 kDa protease detected upon zymography were significantly higher in plasma from SLE patients than in healthy control samples, we hypothesized that this molecule might play a role in SLE. In an attempt to purify and identify the unknown protease, we performed anion exchange chromatography (Material and Methods). All the obtained fractions were analyzed by gelatin zymography ( Figure S1A). However, the gelatinolytic bands did not disclose any information whether the 80 kDa protease was  were quantified as is shown in Figure 2B. The results from the Western blot analysis of C1s showed that this protease was present in the same fractions (mainly fractions 63-71) in which we observed gelatinolytic activity in the presence of EDTA. In contrast, the MMP-9 signal in the Western blot (mainly fractions 53-71) did not perfectly correlate with the zymolytic activity detected in the standard gelatin zymography. In the case of MMP-2, Western blot immunoreactivity and gelatinolytic bands detected at 70 kDa were present in similar fractions. C1r levels were also studied, for the reason that it is another protease that forms a complex with C1q and C1s and because C1r has considerable homology with C1s. However, C1r was distributed differently than the EDTA-resistant protease in the Western blot analysis ( Figure 2B). We herewith provided a quality control of the used antibodies and complementary information excluding C1r and reinforcing C1s as the unknown EDTA-resistant protease. In conclusion, these analyses validated that C1s was the searched molecule and that it possesses newly discovered gelatinolytic activity. The co-migration of human C1s with human MMP-9 upon classical gelatin zymography stimulated a critical re-appraisal of previously published data.

| Characterization and optimization of the detection of C1s in zymography analysis
After corroboration that C1s was the EDTA-resistant protease detected at the molecular weight of 80 kDa in gelatin zymography, besides actMMP-9, we optimized the detection of C1s molecule with the EDTA/gelatin zymography technique. We studied the gelatinolytic capacity of C1s in the presence or absence of EDTA and/or Ca 2+ . As expected, in the presence of EDTA all the signal from metalloproteases was inhibited and solely the proteolytic C1s band remained visible. Interestingly, the signal from C1s in gelatin zymography was also enhanced in the absence of Ca 2+ (Figure 3A,B).  Figure 4A,B). In parallel, we performed ELISA to detect C1s in the same plasma samples (Figure 4C). Both complementary analytical techniques gave similar results (P = 0.0004) ( Figure 4D). Therefore, our results corroborated the finding (Figure 1) of significantly higher levels of C1s in plasma of SLE patients (n = 22) versus healthy controls (n = 20) (Figure 4).
It is known that C1 binds to IC. Therefore, we studied if the presence of C1s in IC might be analyzed by gelatin zymography.
Since the levels of IC in SLE plasma are high, we compared plasma samples of patients suffering from SLE and healthy individuals. First, we purified the IgG and IgG-IC by protein G-Sepharose precipitation.
The results showed that plasma of the studied patients suffering from SLE had higher levels of C1s in the protein G-bound fractions, thus in the IC (Figure 5A,B). Additionally, we showed that the levels of C1s had a trend towards higher levels in the unbound fractions, confirming that the levels C1s were higher in the plasma from the SLE patients analyzed compared with the healthy control samples ( Figure S3).
To confirm the presence of C1s in IC, we analyzed the presence of C1s by Western blot (Figure 5C). The quantification of the EDTA-zymography analysis ( Figure 5D) and the Western blot for C1s ( Figure 5E

| DISCUSSION
In this manuscript, we revealed that (a) previously annotated human activated MMP-9 in plasma or serum samples may be C1s; (b) with the use of EDTA/gelatin zymography these entities were identified and discriminated; (c) the EDTA/gelatin zymography technique was optimal in the absence of Ca 2+ and (d) we discovered that the levels of C1s were higher in plasma and IC of the SLE patients analyzed versus control individuals.
Gelatin zymography assays are commonly used to study gelatinases. 21 However, analysis and interpretation of gelatin zymography data from complex sources, like plasma or tissue extracts, may be a difficult subject, since not only MMP-2 and MMP-9 degrade gelatin, but also other proteases such as neutrophil elastase or trypsin.
Consequently, a critical analysis and complementary assays need to be performed to ratify the data obtained from gelatin zymography.
MMP-9, as other MMPs, is secreted as a zymogen and has to be activated to execute catalytic activity. The cysteine switch from the propeptide interacts with the catalytic Zn 2+ in the active site, thereby keeping the proenzyme inactive. 22 Hence, proMMP-9 is often activated by cleavage of the propeptide, which may be accomplished by proteolytic activators such as kalikrein, plasmin, neutrophil elastase or active MMP-3. This proteolytic removal of the propeptide shifts the molecular weight of human MMP-9 from about 90 to 80 kDa. Both entities, proMMP-9 and actMMP-9 are detectable in gelatin zymography. 21 The gelatin zymography from plasma samples in Figure 1A showed several bands besides the expected NGAL-MMP-9 complex, . This result was corroborated by a specific ELISA for C1s. Therefore, our data suggest that the levels of C1s are higher in the plasma and IC of SLE patients. After future validations with additional patient and control cohorts, the measurement of C1s in plasma and IC may become a marker to assist with the diagnosis of SLE.
Patients with deficiencies in C1 components usually develop SLE: specifically, 91% of individuals with C1q deficiency and 60%-66% of individuals with C1r-C1s deficiencies. 9,11 The levels of C1s in SLE patients have not been studied in detail. De Bracco et al measured the levels of C1q, C1r and C1s and they made correlations between these 3 components of C1. They concluded that in the majority of SLE cases C1q, C1r and C1s presented a good correlation but also low C1q levels coexisted with high levels of C1s and C1r. 24 Furthermore, in patients with lymphopenic hypogammaglobulinaemia the levels of C1s were normal, whereas C1q levels remained unusually low. 25 All these data suggest that C1q and C1s are differently regulated in (auto)immune pathologies. The basis for this might be differences in the cell sources for C1s and C1q. Whereas almost all serum complement system proteins, including C1s and C1r, are synthetized in the liver, 26 C1q is mostly secreted by macrophages and immature dendritic cells. [27][28][29] Here we showed that C1s levels are considerable in the IC of the SLE plasma samples studied. Since the presence of C1s autoantibodies has been demonstrated in SLE plasma samples, 14 we cannot exclude that some of the autoantibodies detected in the IC might be against C1s. Therefore, in the IC, we may be detecting the C1s attached to autoantibodies against C1s but also to C1q in IC. SLE diagnosis is very challenging because no generally accepted diagnostic criteria exist. Usually, the diagnosis is based on clinical manifestations and laboratory tests, including detection of autoantibodies, functional tests and imaging analysis. Sometimes the diagnosis is made after exclusion of alternative diseases. 30 Therefore, the discovery of new markers for better diagnosis of this disease is important. In this manuscript, we provide preliminary evidence that the levels of C1s in plasma and IC of SLE patients are significantly higher than in control individuals (n = 20) suggesting that C1s might become useful for diagnosis.
We conclude that EDTA/gelatin zymography is a relevant new exploratory method for detection of C1s in SLE and probably also in other pathologies.