Gain of function of a metalloproteinase associated with multiple myeloma, bicuspid aortic valve, and Von Hippel–Lindau syndrome

Abstract

A patient diagnosed with multiple myeloma, bicuspid aortic valve, and Von Hippel–Lindau syndrome underwent whole-exome sequencing seeking a unified genetic cause for these three pathologies. The patient possessed a single-point mutation of arginine to cysteine (R24C) in the N-terminal region(pro-domain) of matrix metalloproteinase 9 (MMP-9). The pro-domain interacts with the catalytic site of this enzyme rendering it inactive. MMP-9 has previously been associated with all three pathologies suffered by the patient. We hypothesized that the observed mutation in the pro-domain would influence the activity of this enzyme. We expressed recombinant versions of MMP-9 and an investigation of their biochemical properties revealed that MMP-9 R24C is a constitutively active zymogen. To our knowledge, this is the first example of a mutation that discloses catalytic activity in the pro-form in any of the 24 human MMPs.

Introduction

Multiple myeloma (MM) is characterized by aberrant proliferation of plasma cells located in the bone marrow and by the secretion of a monoclonal immunoglobulin [1,2]. Symptoms include hypercalcemia, renal insufficiency, anemia, and bone lesions. We describe the case of a MM patient who was concurrently affected by two other diseases: (i) Von Hippel–Lindau (VHL) syndrome, a hereditary condition characterized by hemangioblastomas of the retina and the central nervous system and (ii) bicuspid aortic valve (BAV), a congenital heart defect. A recent study showed that over 50% of MM patients have a comorbidity at diagnosis [3]. Although a genetic mechanism linking the VHL gene expression and MM has been hypothesized [4], there is no clear evidence of association to the above diseases.

In search for a potential genetic cause that could explain the three disorders, the patient underwent whole-exome sequencing. A candidate missense mutation was found in the matrix metalloproteinase 9 (MMP-9) gene. To ascertain whether this mutation could link mechanistically to the patient’s disorders, we investigated the mutant protein in recombinant form and biochemically characterized in comparison with the wild-type enzyme.

Matrix metalloproteinases (MMPs) represent a class of at least 24 different Zn²⁺ dependent proteases with a role in modulating cellular environments and responses [5,6]. The MMP pro-domain shields the catalytic domain in the inactive form of MMPs secreted from cells. Like most MMPs, MMP-9 is secured from activation by the presence of the pro-domain because a Cys residue (C99) in this domain co-ordinates with the catalytic Zn²⁺ ion preventing it from participating in catalysis. Dissociation of this interaction, resulting in activation of pro-MMP-9 can occur by proteolytic removal of the pro-domain or inactivation of the latency-inducing Cys-Zn via the ‘cysteine switch’ mechanism [7–9]. MMPs are widely expressed in epithelial and endothelial cells as well as cells of the hematopoietic lineage [10,11]. MMP-9 has a wide range of physiologic substrates, including connective tissue proteins, chemokines, cytokines, and other inflammatory mediators [5].

Dysregulated MMP activity has been implicated in proliferative and degenerative diseases as well as normal physiological functions [12]. Thus, MMP-9 and its endogenous inhibitor, tissue inhibitor of metalloproteinase-1 (TIMP-1), are key mediators of acute inflammation and regulators of the wound healing process [13]. Consequently, this protease is considered to represent a relevant therapeutic target [14].

In addition to regulation by TIMP-1, MMP-9 activity is also controlled at the level of zymogen activation. The secreted zymogen can be activated by proteolytic removal of its pro-domain by many proteases including MMP-3 and plasmin [15]. It can also be activated by reagents that block the critical Cys-99 in the pro-domain. The mutation in MMP-9 (R24C) resulted in a novel cysteine in the pro-domain. Based on simplistic structural modeling we hypothesized that Cys-24 may form a disulfide with Cys-99 thereby blocking the shielding activity and resulting in a constitutively active (gain of function (GOF)) protease. Most of the biological functions of MMP-9 have come from studying mice that lack this gene [11,16]. We propose that studying the human gene from a GOF mutation will provide additional insight into the biological roles of MMP-9.

Results

Identification of the mutation

An adult patient diagnosed with MM, VHL syndrome, and BAV underwent exome sequencing and annotation according to the Scripps Genome ADVISER [17] uncovering 98 053 variants (nucleotides that differ from the corresponding sequences in the reference human genome); 333 of these changes are linked to pathogenic or damaging conditions (Supplementary Table S1). Remarkably, the patient did not present a typical germline mutation in the VHL gene. We searched PubMed for abstracts mentioning one of the three patient diseases in combination with any of the 333 genes presenting pathogenic and damaging variants. For 10 of the genes, we found at least one article mentioning MM, at least one article mentioning VHL syndrome, and at least one article mentioning BAV (Table 1).

Table 1.

Literature search.

Gene	Myeloma	Bicuspid aortic valve	Von Hippel–Lindau
VEGFA	89	1	88
EGFR	157	2	44
MMP9	18	9	7
NOTCH1	48	51	6
ENG	12	1	4
NOS3	5	11	3
TGFBR1	2	6	3
ACTA2	2	10	2
TGFBR2	7	9	1
SMAD3	5	4	1

Search of existing literature showing the 10 top genes associated with the three diseases affecting the patient. For each gene, the numbers indicate how many individual article abstracts were found citing MM, or VHL syndrome, or BAV. The search was performed on PubMed in March of 2019.

MMP-9 Arg24Cys

One of the top genes identified from this search was MMP-9. We chose to investigate the MMP-9 variant as there is evidence that the protease level increases in MM patients [18] and because the MMP-9 protein activity can be determined in vitro. Located on exon 1 of the MMP-9 gene, this variant generates an arginine to cysteine substitution at position 24. The patient was found to be heterozygote for the mutation. This single nucleotide variant (C to T nucleotide substitution) is present in the NCBI dbSNP (build 155) database of Short Genetic Variations as rs144023823 (https://www.ncbi.nlm.nih.gov/snp/rs144023823), which, according to the ClinVar catalog, has been reported to be of ‘Benign’ or ‘Likely-Benign’ clinical significance. The variant is also identified as 20–44637635-C-T (GRCh37) in gnomAD Browser v2.1.1 (https://gnomad.broadinstitute.org/variant/20–44637635-C-T?dataset=gnomad_r2_1) without any associated phenotype. The global allelic frequency is 0.0038. The allelic frequency is higher than 1% in some European populations, such as 0.011 in the Finnish and 0.019 in the Estonian. These data indicate that the MMP-9 R24C variant is relatively frequent in the context of specific populations, where it may have only moderate consequences. Access to U.K. Biobank (https://www.ukbiobank.ac.uk/) containing genetic and health information from half a million U.K. participants, allowed a much more an in-depth analysis, which indicated that the MMP-9 R24C variant is significantly (P = 8.9 × 10⁻⁷) associated with a ‘Nucleated red blood cells’ phenotype. Nucleated red blood cells are immature and their presence in the circulation in adults may indicate a bone marrow abnormality. We concluded that in combination with other genetic changes, the MMP-9 R24C mutation might play a significant role in the etiology of the diseases affecting this specific patient and warranted a biochemical characterization.

Expression and purification of recombinant MMP-9 variants

Wild-type pro-MMP-9 and the pro-MMP-9 R24C mutant were expressed from a modified pCEP vector containing a C-terminal HIS purification tag in HEK293 cells. We also expressed pro-MMP-9 R24S, with an arginine to serine to control for the influence of the cysteine side chain at position 24. Ni-NTA resin purification resulted in the recombinant MMP-9 variants shown in Figure 1.

Figure 1. Coomassie stained SDS–PAGE gel of recombinant proteins demonstrating purity.

The three indicated HIS tagged MMP-9 constructs were expressed in 293T Freestyle cells and purified by Ni-NTA chromatography. Expressed proteins are secreted into culture media [1] and purified over a Ni-NTA column. Flowthrough material [2]. Captured material is eluted with 200 mM imidazole. Fractions of 1 ml were collected [3,4].

Assessing the activity status of MMP-9 variants

The variants were incubated at 37°C in MMP activity buffer (see Materials and methods) in the presence of the MMP activator p-Aminophenylmercuric Acetate (APMA) (Figure 2A) and analyzed with a quenched fluorogenic peptide substrate. APMA is an organomercurial agent that presumably interacts with the Cys-99 in the pro-domain of MMPs resulting in their activation [19]. We noticed that without the addition of APMA (time = 0, Figure 2A) MMP-9 R24C has enzymatic activity, implying that it is constitutively active. In contrast, the WT and R24S MMP-9 variants are essentially inactive. During the 22-h incubation with APMA MMP-9 R24S and WT activities increased 75 and 100-fold while MMP-9 R24C increased only 3-fold. We conclude that APMA addition significantly increases enzyme activity while this effect is minimal on the constitutively active MMP-9 R24C. The variants were subjected to SDS–PAGE to monitor pro-domain removal, one of the consequences of becoming active, as well as other auto-processing fragments (Figure 2B). Based on the zymogen status of all three enzymes before APMA addition and the ability for MMP-9 R24C to cleave a peptide substrate, pro-domain removal is not required for the activity status of MMP-29 R24C. Thirty minutes post AMPA addition all variants had removed their pro-domain (Figure 2B). Pro-domain removal by cleavage at the Ala⁹³–Met⁹⁴ bond results in the first active enzyme and activity increases with time and further auto processing (Figure 2A,B) [20].

Figure 2. MMP-9 R24C is constitutively active.

(A) Recombinant MMP-9 variants were incubated with the chemical activating agent APMA over a time course of 22 h and assayed for enzymatic activity towards a fluorescent peptide substrate. Data represent the means and SD (error bars) of three technical replicates. (B) Coomassie stained SDS–PAGE gel of MMP-9 variants showing auto processing cleavages over time. Twenty microliter portions from each timepoint were analyzed without adjusting for protein concentration.

To test the prediction that pro-MMP-9 R24C is constitutively active we incubated in the absence of APMA (Figure 3). The activity of pro-MMP-9 R24C is sustained during the 22-h time course in contrast with the lack of activity from the WT and R24S mutant (Figure 3A). The variants were subjected to SDS–PAGE to monitor pro-domain removal and other auto-processing fragments (Figure 3B). Pro-domain removal from the MMP-9 WT and R24S mutant were delayed until the 22-h time point. Whereas pro-domain removal from the MMP-9 R24C mutant is visible at 30 min and further processing occurs over the 22-h time course. These data confirm that pro-MMP-9 R24C is a constitutively active protease.

Figure 3. Assessing the constitutive activity of MMP-9 variants.

(A) Recombinant MMP-9 variants were incubated over a time course of 22 h and assayed for enzymatic activity towards a quenched fluorescent peptide substrate. Data represent the means and SD (error bars) of three technical replicates. (B) Coomassie stained SDS–PAGE gel of MMP-9 variants showing the auto processing cleavages over time. Equivalent portions from each timepoint were analyzed after adjusting for protein concentration.

We investigated the molecular forms of MMP-9 by gelatin zymography, to confirm their ability to digest larger peptides and proteins (Figure 4). Following incubation at 4°C for 2 h MMP-9 R24C shows processing in the absence of APMA indicating proteinase activity without activation. Following incubation at 37°C MMP-9 R24C has undergone extensive processing, and the other variants also demonstrate a small amount of processing. Zymogram gels always show the pro-form as an active gelatinase because during denaturation and renaturation the enzyme becomes active [21,22]. More importantly, the extensive processing revealed in MMP-9 R24C confirms that this mutant has undergone extensive autolysis due to its constitutive activity.

Figure 4. Gelatin zymography showing the ability of MMP-9 variants to cleave gelatin.

Three variants were incubated in MMP buffer with or without 1 mM APMA for 2 h at 4°C or 37°C. Samples were loaded at a concentration of 5 μg/ml and run in a zymogram gel, followed by soaking in renaturation buffer and stained with Coomassie. Gelatinase activity appears as white bands against a dark background.

We show that MMP-9 R24C is active without the addition of APMA on both peptide substrates and on gelatin. To assess the proteolytic activity of MMP-9 R24C on a physiologically relevant protein substrate, we analyzed the cleavage of α1-Proteinase inhibitor (α1-PI) [23,24]. We found that MMP-9 R24C cleaves α1-PI (Figure 5). MMP-9 R24C had a catalytic efficiency (k_cat/K_m) of 2.0 × 10⁴ M⁻¹ s⁻¹, which compares favorably with the value of 6.3 × 10³ M⁻¹ s⁻¹ for APMA-activated MMP-9 WT.

Figure 5. α1-PI cleavage by MMP-9 variants.

The serpin α1-PI was used as a protein substrate for MMP-9. (A) WT MMP-9 was pre-activated with APMA and allowed to process α1-PI. (B) Constitutively active MMP-9 R24C processing of α1-PI. The inactivating cleavage product of α1-PI (black arrowhead) appears as MMP-9 concentration increases. The concentration to cleave half the substrate (E_1/2) was measured based on the disappearance of the full-length substrate (%) using densitometry. These values were used to calculate k_cat/K_M as described in the methods.

Discussion

The activities of the MMP family have been correlated to the initiation and progression of cancer, osteoarthritis, neurodegenerative, infectious, and cardiovascular diseases [25–27]. However, MMPs also have beneficial roles in homeostasis such as wound healing, inflammatory response, and angiogenesis [28–32], possibly through their involvement in the processing of growth factors and their receptors, cytokines, chemokines, and adhesion molecules [10]. Originally identified as a gelatinase [33], MMP-9 itself is implicated in the control of inflammation through the inactivation of serpins that control inflammatory serine proteases [23,24].

In this study, we describe a missense mutation (R24C) in the MMP-9 protease affecting an individual suffering from MM, BAV, and VHL syndrome. To determine if this mutation might in principle be involved in biological mechanisms that contribute to these conditions, we produced the mutant protein and identified MMP-9 R24C as a constitutively active zymogen revealing that this mutation generated a GOF effect in the pro-form. Seeking to understand the molecular basis for the GOF, we show that the biochemical characteristics of a control MMP-9 R24S mutant to be very similar to the wild-type. Accordingly, the reason for the activated status of the R24C mutant is not due to the lack of the Arg side chain, but most likely, to the presence of the SH group found in the patient’s mutation.

Cys-99 is the only cysteine residue present in the wild-type pro-domain. We hypothesize that Cys-24 may form a novel disulfide bond with Cys-99 thereby preventing its interaction with the Zn²⁺ ion in the catalytic site. This would result in a constitutively active protease much in the same way as artificially engineered mutations in the Cys switch region [34]. Accordingly, it may be possible to obtain constitutively active forms of other MMPs by introducing a cysteine residue in the amino-terminal region of the pro-domain.

A SNP in the MMP9 promoter leading to enhanced transcription is reported to associate with the severity of atherosclerosis [35] and tracks with the risk of developing smoke-induced emphysema [36]. However, there is no evidence that enhanced expression would lead to enzyme activity. To our knowledge, R24C is the first characterized natural mutation of any MMP displaying constitutive catalytic activity in the proenzyme state. Because the mutation results in a GOF any pathology associated with it is likely to be observed in heterozygotes and homozygotes. In considering the role of constitutively active MMP-9 it is also important to consider the balance between the enzyme and its relation to endogenous inhibitors such as TIMPs. It is not clear to what extent this GOF mutant would be associated with the development of MM, VHL syndrome, and BAV exhibited by the patient. However, the role of MMP-9 in activating inflammatory mediators and/or modifying the connective tissue environment would be consistent with the development of these diseases.

Materials and methods

Exome sequencing and analysis

Detailed procedures for whole-exome sequencing and downstream analysis have been described previously [37]. In brief, whole blood samples were preserved using Paxgene DNA tubes (PreAnalytiX, Hombrechtikon, CH), and genomic DNA was extracted using the QIAamp system (Qiagen, Valencia, CA). Enriched exome libraries were captured using the Agilent SureSelect v7 systems according to the manufacturer’s instructions (Agilent, Santa Clara, CA). Final libraries were generated using Illumina TruSeq sample preparation kits and underwent 100 bp paired-end sequencing on a HiSeq 2500 (Illumina, San Diego, CA). Variants were determined following GATK best practices [38], annotated with SG-ADVISER [17], and filtered for candidate pathogenic variation as previously described [37].

Pubmed search

A literature search was performed as described in the legend of Table 1.

Expression vector and MMP insertion

The nucleotide sequence corresponding to Uniprot limits 1–707, encoding the ORF of human MMP-9 and mutants, were ordered as gBlocks from IDT and cloned into a modified pCEP vector in frame with a C-terminal 8× histidine affinity tag using the restriction enzymes NheI and AgeI. The entire MMP-9 coding sequence relative to all MMP-9 expression vectors were verified by DNA sequencing.

Production of MMP-9 proteins

Expression DNA vectors purified by Nucleobond Xtra Midi (Machery-Nagel) were transfected into FreeStyle^™ 293 suspension cells. Cells were grown in FreeStyle^™ 293 serum-free medium (Invitrogen) in a 37°C incubator containing a humidified atmosphere of 8% CO₂ in air, on an orbital shaker platform rotating at 125 rpm. For each protein preparation, cells were suspended in 125 ml at a concentration of 10⁶ live cells/ml in a 1 l flask. After 2 h of further incubation, cells were transiently transfected with 250 μg DNA using the PEI derivative Transporter^™ 5 Transfection Reagent (Polysciences, Inc.) according to the manufacturer instructions. After 2 h of additional incubation, the culture volume was increased with 125 ml prewarmed FreeStyle^™ 293 medium and returned to the incubator. Cultures were collected 4 days post-transfection and centrifugated 10 min at 7000 g at 4°C. The supernatant, containing the secreted recombinant MMP-9, was recovered for protein purification.

Purification of MMP-9 8xHis fusion proteins

Approximately 250 ml of culture supernatant was stirred for 2 h at 4°C with 2 ml of MMP activity buffer washed Ni-NTA agarose slurry (Qiagen #30230). The mix was packed in a column (Bio-Rad Econo-Column, 1.0 × 5 cm) and washed three times with 10 ml of MMP activity buffer (50 mM Tris pH = 8.0,100 mM NaCl, 10 mM CaCl₂, 0.05% Triton-X 100).

The MMP-9 8xHis fusion proteins were eluted using 4 × 1 ml volumes of 200 mM Imidazole-MMP activity buffer. Fractions were visualized by 4–12% SDS PAGE (Invitrogen Bolt Bis-Tris Plus Gel). Protein final concentration was measured via absorbance at 280 nm.

MMP-9 enzyme APMA activation

A 2 ml reaction mixture was made by adding 20 μl of 100 mM 4-aminophenylmercuric acetate (Sigma #A9563) to 1980μl of purified protein diluted in MMP activity buffer. The reaction was incubated for the time intervals indicated in Figures 2 and 3 at 37°C. At the indicated time points 50 μl of the reaction mixture was added to 50 ml of substrate mixture containing 200 μM ACC-GRGPYGYHRK(Dnp)G-NH_2. Enzymatic activity was measured in white 96-well opaque plates (Costar, Corning) on a fluorescent plate reader (BMG LabTech). 7-amino-4-carbamoylmethylcoumarin (ACC), fluorophore was detected at excitation/emission wavelengths of 355/460 nm. Reaction velocity was calculated as the linear portion of kinetic curves using MARS data analysis software (BMG LabTech). Twenty microliter of each time point was saved to run on a 4–12% SDS PAGE (Invitrogen Bolt Bis-Tris Plus Gel).

Active concentration determination

MMPs were activated in MMP activity buffer supplemented with 1 mM APMA for 4 h at 37°C. The active enzyme concentration was determined by titration with the MMP inhibitor batimasatat. Serial dilutions of batimastat (Selleckchem) were incubated with the purified enzymes for 30 min at 37°C. Batimastat is a potent, reversable, broad-spectrum MMP inhibitor. The activity assay was initiated by the addition of 100 μM of the fluorogenic substrate in the MMP activity buffer. Enzymatic activity was measured in white 96-well opaque plates (Costar, Corning) on a fluorescent plate reader (BMG LabTech). 7-amino-4-carbamoylmethylcoumarin (ACC), fluorophore was detected at excitation/emission wavelengths of 355/460 nm. Reaction velocity was calculated as the linear portion of kinetic curves using MARS data analysis software (BMG LabTech). Velocities were plotted against the inhibitor concentrations allowing for the calculation of the active enzyme concentrations.

Gelatin zymography

MMP samples were diluted to 5 μg/ml in MMP activity buffer with or without 1 mM APMA. The samples were then incubated at 37°C or 4°C for 2 h. 3× SDS loading buffer without DTT was added to the samples. Preparation of samples for zymography requires they are not boiled. Novex^™ 10% Zymogram Plus (Gelatin) gels were purchased from Invitrogen catalog # ZY00100BOX. Gels were kept cold by adding cold Laemmli running buffer and surrounded with ice. Gels were run at 125 V for 115 min. The gel was then washed two times in 2.5% Triton X-100, 50 mM Tris pH = 8.0,100 mM NaCl, 10 mM CaCl₂ at room temperature for 30 min. The high concentration of Triton-X replaces the SDS in the gel. The gel was rinsed by rocking 10 min in a low concentration Triton buffer: 0.05% Triton X-100, 50 mM Tris pH = 8.0,100 mM NaCl, 10 mM CaCl₂. The buffer was discarded and replaced with low Triton buffer and the gel was incubated at 37°C for 2 h. The gel was then stained with 40% methanol, 10% acetic acid, and 0.5% w/v Coomassie blue overnight. Gels were destained in a solution of 40% methanol and 10% acetic acid until bands are visible.

Kinetics of protein cleavage (kcat/KM)

Rates of cleavage (k_cat/K_M) were measured by treating 2 μM substrate with a dilution series of protease for 1 h at 37°C, followed by SDS–PAGE detection of cleavage products. The enzyme concentrations corresponding to 50% cleavage of the full-length substrates were determined by densitometry. This allowed us to determine their respective k_cat/K_M values from the following half-life equation: k_cat/K_M = ln2/(E_1/2 × t) where E_1/2 is the concentration of protease that produces 50% substrate cleavage in time = t [39]. α1-PI was a gift from Miles Inc. Clayton, NC.

Statistical analysis

Statistical analyses were performed using GraphPad Prism software (version 8.9; GraphPad Software Inc.).

Abbreviations

APMA

p-Aminophenylmercuric Acetate

BAV

bicuspid aortic valve

GOF

gain of function

MM

multiple myeloma

MMP-9

matrix metalloproteinase 9

MMPs

matrix metalloproteinases

TIMP-1

tissue inhibitor of metalloproteinase-1

VHL

Von Hippel–Lindau

Data Availability

All data needed to evaluate the conclusion in the paper are present in the paper and its Supplementary Table. Source data of this study are available from the corresponding author upon request.