J Mol Med (2017) 95:487–497 DOI 10.1007/s00109-017-1510-z

ORIGINAL ARTICLE

Matrix metalloproteinase-14 triggers an anti-inflammatory proteolytic cascade in endotoxemia Alina Aguirre 1 & Jorge Blázquez-Prieto 2 & Laura Amado-Rodriguez 2,3 & Inés López-Alonso 2,4 & Estefanía Batalla-Solís 2 & Adrián González-López 5 & Moisés Sánchez-Pérez 3 & Carlos Mayoral-Garcia 2 & Ana Gutiérrez-Fernández 6 & Guillermo M Albaiceta 2,4,7

Received: 6 July 2016 / Revised: 4 December 2016 / Accepted: 17 January 2017 / Published online: 24 January 2017 # Springer-Verlag Berlin Heidelberg 2017

Abstract Matrix metalloproteinases can modulate the inflammatory response through processing of cyto- and chemokines. Among them, MMP-14 is a non-dispensable collagenase responsible for the activation of other enzymes, triggering a proteolytic cascade. To identify the role of MMP-14 during the pro-inflammatory response, wildtype and Mmp14−/− mice were challenged with lipopolysaccharide. MMP-14 levels decreased after endotoxemia. Mutant animals showed 100% mortality, compared to 50% in wildtype mice. The increased mortality was related to a more severe lung injury, an impaired lung MMP-2 activation, and increased levels of the alarmin S100A9. There were no differences in the expression of other mediators including Il6, Cxcl2, Tgfb, Il10, or S100a8. A similar result was observed in lung explants of both genotypes cultured in presence of lipopolysaccharide. In this ex vivo model, exogenous activated MMP-2 ameliorated the observed increase in alarmins. Samples from septic patients showed a decrease in serum MMP-14 and activated MMP-2 compared to non-septic critically ill patients. These results demonstrate that the MMP-14MMP-2 axis is downregulated during sepsis, leading to a proinflammatory response involving S100A9 and a * Guillermo M Albaiceta [email protected] 1

Facultad de Ingeniería y Ciencias Agropecuarias, Universidad de las Américas, Quito, Ecuador

2

Departamento de Biología Funcional, Área de Fisiología, Instituto de Oncología del Principado de Asturias (IUOPA), Universidad de Oviedo, Oviedo, Spain

3

Unidad de Gestión Clínica de Medicina Intensiva, Hospital Valle del Nalón, Langreo, Spain

more severe lung injury. This anti-inflammatory role of MMP-14 could have a therapeutic value in sepsis. Key messages • MMP-14 levels decrease in lungs from endotoxemic mice and serum from septic patients. • Mmp14−/− mice show increased lung injury and mortality following endotoxemia. • Absence of Mmp14 decreases activated MMP-2 and increases S100A9 levels in lung tissue. • MMP-14 ameliorates inflammation by promoting S100A9 cleavage by activated MMP-2. Keywords Matrix metalloproteinases . MMP-14 . Endotoxemia . Sepsis . Alarmins Sepsis, defined as the systemic response to infection, is a severe condition linked to high morbidity and mortality rates [1]. Septic shock related to gram-negative infections is a major concern both in adults and neonates. One of the main pathogenetic mechanisms behind the septic syndrome is the inflammatory response. The paradigm of inflammation during sepsis has evolved from an uncontrolled pro-inflammatory state to a 4

Unidad de Cuidados Críticos, Área de Gestión Clínica del Corazón, Hospital Universitario Central de Asturias, Oviedo, Spain

5

Department of Anesthesiology and Operative Intensive Care Medicine, Charité Universitätsmedizin, Berlin, Germany

6

Departamento de Bioquímica y Biología Molecular, Instituto de Oncología del Principado de Asturias (IUOPA), Universidad de Oviedo, Oviedo, Spain

7

Centro de Investigación Biomédica en Red-Enfermedades Respiratorias, Instituto de Salud Carlos III, Madrid, Spain

488

complex, time-changing, and mix of pro- and antiinflammatory responses [2]. Within this changing framework, regulatory mechanisms of the inflammatory response are a major area of interest in order to identify suitable therapeutic targets. Experimental models of endotoxemia, after administration of a proinflammatory agent such as bacterial lipopolysaccharide (LPS) are widely used to characterize these responses and the associated lung injury [3]. Matrix metalloproteinases (MMPs) are a family of enzymes characterized by their ability to cleave the components of the extracellular matrix. Moreover, they have a large variety of substrates, including some inflammatory mediators. Currently, MMPs are identified as key regulators of the inflammatory response, as proteolytic processing may increase or decrease the biological activity of the substrate [4]. For instance, MMP-8 has a dual effect on inflammation, as processing by this enzyme can increase the chemotactic effect of LIX [5] and inactivate the alarmins S100A8 and S100A9 [6]. In opposite, MMP-2 degrades the alarmins S100A8 and S100A9, thus contributing to the resolution of inflammation [7]. Matrix metalloproteinase-14 (MMP-14, also known as MT1-MMP) is a membrane-bounded protease responsible for pericellular collagenolysis and activation of other proteases such as MMP-2 [8]. In contrast to other MMPs, absence of MMP-14 results in death in 4–5 weeks after birth, with an abnormal lung [9] and skeletal development [10]. This finding suggests that absence of MMP-14 cannot be compensated by other enzymes. However, inflammation has not been implicated in the pathogenesis of these defects and, the inflammatory response in absence of MMP-14 has not been fully studied. To test the hypothesis that MMP-14 could also regulate the inflammatory response, we have studied an experimental model of neonatal lung inflammation in response to LPS in mice lacking MMP-14 and their wildtype counterparts. The differences between genotypes would help to identify the targets of MMP-14 during endotoxemia.

Methods Animals Mice lacking Mmp14, generated in a C57BL6/ Sv129 background [11], and their wildtype counterparts were used. Animals were kept in specific pathogen-free conditions, with 12-h light/dark cycles, and with free access to food and water until their inclusion in the study. Genotypes were determined by PCR using genomic DNA obtained from the mice tails. All the experiments were performed following the Spanish law on animal research. The protocol was evaluated and approved by the Animal Research Ethics Committee at the Universidad de Oviedo.

J Mol Med (2017) 95:487–497

Experimental model and tissue collection Newborn animals (5–8 days after birth) of both genotypes were randomized to receive an intraperitoneal dose of 10 mg/kg of lipopolysaccharide (LPS) in isotonic saline or only saline (final volume 50 μl), and followed up to 5 days to study survival (n = 8 per genotype and treatment). In separate experiments, injected animals (n = 6 per group) were anesthesized with intraperitoneal ketamine and xylazyn, tracheostomized, and the lungs, heart, kidney, and spleen removed. Samples from these organs were stored at −80 °C for biochemical assays and included in paraffin for histological studies. A bronchoalveolar lavage was performed in additional animals (n = 6 per group), injected with either LPS or saline as previously described. Briefly, mice were anesthesized, a tracheostomy performed, and three aliquots of sterile isotonic saline were injected (300 μl each). The recovered fluid (BALF) was centrifuged and the supernatants collected and stored at −80 °C. Histological studies Lung injury was assessed in hematoxylin-eosin stained histological preparations. Tissue damage was quantified using a semiquantitative scale ranging from 0 to 4 (0: normal lung, 1: capillary congestion, 2: alveolar wall thickening and inflammatory infiltrates, 3: intraalveolar flooding, 4: massive disruption of the lung structure). In separate immunohistological preparations, neutrophils were detected using an antibody against myeloperoxidase (MPO). The number of MPO-positive cells was averaged in three random ×40 fields. Western blot Tissues were homogenized in standard RIPA buffer containing protease and phosphatase inhibitors. Samples were then centrifuged at 13,000 rpm for 15 min at 4 °C. Both the supernatants and the pellets were stored at −80 °C. The protein content of each sample was measured using the bicinchoninic acid technique (BCA protein assay, Pierce, USA). Specific protein abundance was quantified by Western blotting. Equal amounts of protein of each sample were loaded in standard SDS-PAGE gels, electrophoresed, and transferred to methanol-activated PVDF membranes. These membranes were blocked with non-fat milk in TBS-T buffer and incubated overnight with antibodies against matrix metalloproteinase-14 (MMP-14, Abcam, UK), S100 calciumbinding protein A8 (S100A8, R&DSystems, USA), and S100 calcium-binding protein A9 (S100A9, ProtEra). Actin (Santa Cruz Biotechnology SC-1616, USA) was used as loading control. The appropriate peroxidase-linked secondary antibodies were used for detection using a Chemidoc Imaging system (UVP, USA), and the intensity of each band quantified using the ImageJ software (NIH, USA). Quantitative PCR RNA was extracted from the different tissues using the TRIzol reagent (Sigma, Poole, UK), according

J Mol Med (2017) 95:487–497

489

to the manufacturer’s protocol. Then, cDNA was synthetized from 1 μg of RNA using the Superscript II kit (Invitrogen Life Technologies) and random hexamers. Gene expression was quantified by quantitative PCR using 100 ng of cDNA, the SYBR Green PCR master mix (Applied Biosystems), and 10 μM of specific primers (Table 1, all from Sigma-Aldrich, USA) in an ABI Prism 7700 real-time PCR system. Expression of Gadph was used as loading control. The relative expression of the analyzed genes was calculated as 2ΔCT(gene of interest)–ΔCT(GAPDH) .

Gelatin zymography Levels of MMP-2 and MMP-9 were measured by gelatin zymography, as previously described [12]. Lung explants Wildtype and Mmp14−/− animals (5 days old, n = 6 per group) were anesthetized and the heart-lung block removed. Lung slices (1 mm thick) were rinsed in PBS with penicillin and streptomycin and incubated in DMEM-F12 medium supplemented with penicillin, streptomycin, and glutamine for 4 h. Then, LPS (at a final concentration of 100 ng/ml) or the equivalent volume of sterile saline, were added to the medium. In additional wells, activated MMP-2 (300 ng/ml, Sigma-Aldrich) was added. After 17 additional hours, tissue samples were collected and stored at −80 °C. Table 1

Oligonucleotide sequences used for qPCR assays

Gene

Oligonucleotide

Sequence 5′-3′

Mmp8

Forward Reverse Forward

CCACACACAGCTTGCCAATGCC GTGGCATTCCTCGAAGACCGGA CTTTAGAGGGAGAAAATTCTGG

Reverse Forward Reverse Forward

CATCATCATAACTCCACACG GAGATCTTCTTCTTCAAGGAC AATAGACCCAGTACTCATTTC CTTCAAAGGAGATAAGCACTG

Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse

AGAAGTAGGTCTTCCCATTG ACCACTTCACAAGTCGGAGG TGCAAGTGCATCATCGTTGT ATCCAGAGCTTGAGTGTGACG GTTAGCCTTGCCTTTGTTCAG TGCAAGTCAGAGACGTGGGG GATCGAGTGTCCACGACGGT CTGTTTCCATTGGGGACACTT CAAGTGTGGCCAGCCTTAGA ATACAAGGAAATCACCATGC ATATTCTGCACAAACTGAGG GGCCAACAAAGCACCTTCTC GTTGCCAACTGTGCTTCCAC GTGCAGTGCCAGCCTCGTCC GCCAGACTGCAAATGGCAGCCC

Mmp13 Mmp2 Mmp14 Il6 Cxcl2 Tgfb Il10 S100a8 S100a9 Gapdh

Patients The protocol involving human samples was reviewed and authorized by the regional ethics committee (Comité Ético de Investigación Clínica del Principado de Asturias, Oviedo, Spain), and participants were included after next-of-kin’s informed consent. Critically ill patients, with or without sepsis, were included in the study during their first 48 h after admission in an intensive care unit. The eligibility criteria for septic patients were the following: age > 18 years old, suspected or confirmed infection and diagnosis of sepsis, or septic shock according to the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) [13]. Patients admitted to the ICU after major cardiac surgery were included in the non-septic group. The sequential organ failure assessment (SOFA) score [14] was calculated and, when available, procalcitonin and C-reactive protein levels in serum were collected from the hospital database. Exclusion criteria included pregnancy or lactation, HIV infection, history of bone marrow or solid organ transplantation, patients undergoing chemotherapy, radiotherapy or immunotherapy, and history of solid cancer or hematological disease. For non-septic patients, the exclusion criteria included in addition any suspected or confirmed infection and diagnosis of a septic process. Main clinical data of this sample is shown in Table 2. A blood sample was drawn and serum obtained. MMP-14 levels were measured in serum by western blotting. Serum MMP-2 was quantified by gelatin zymography. Statistics Data are expressed as mean ± SEM. Variables were compared using a two-way analysis of the variance, including genotype and treatment (saline or LPS) as factors. Post-hoc tests were done when appropriate using Bonferroni’s correction. Comparisons between two groups were done using t tests. A p value lower than 0.05 was considered significant (p < 0.05).

Results Survival and lung injury Wildtype and MMP14-deficient pups were injected either with LPS or saline and followed for 5 days. We first studied changes in lung MMP-14 levels in wildtype animals 16 h after LPS injection. Interestingly, the levels of this protease decreased in endotoxemic mice (Fig. 1a). Wildtype animals showed a 50% survival after LPS administration. However, all deficient mice died within 80 h after LPS injection (Fig. 1b, log-rank test p = 0.02). No deaths in the saline-treated groups were observed. Lung injury was assessed 16 h after LPS administration. Mice lacking Mmp14 developed a more severe lung injury than their wildtype counterparts, as demonstrated by their higher histological scores also showing an enlargement of the distal airways and alveoli and the accumulation of

490

J Mol Med (2017) 95:487–497

Table 2 Clinical data of the patients included in the study. CRP: C-reactive protein. PCT: Procalcitonin

Age

Sex

SOFA

CRP

PCT

Main diagnosis

(mg/L)

(ng/mL)

8 2

225 153

102.8 0.2

Abdominal sepsis Sepsis secondary to pneumonia

Septic-1 Septic-2

79 65

Male Male

Septic-3 Septic-4

63 78

Female Male

5 11

368 163

0.15 1.23

Abdominal sepsis Sepsis secondary to pneumonia

Septic-5

51

Female

4

207

40.54

Sepsis secondary to pneumonia

Septic-6 Septic-7

67 59

Female Female

9 12

304 139

6.46 0.57

Urinary sepsis Sepsis secondary to pneumonia

Septic-8

82

Male

Septic-9 Septic-10

50 76

Female Male

Control-1

80

Male

4

NA

NA

Major cardiac surgery

Control-2 Control-3

65 75

Female Male

3 2

NA NA

NA NA

Major cardiac surgery Major cardiac surgery

Control-4 Control-5 Control-6

61 79 51

Male Female Male

5 7 2

NA NA NA

NA NA NA

Major cardiac surgery Major cardiac surgery Major cardiac surgery

Fig. 1 Role of MMP-14 in endotoxemia. MMP-14 levels in lung homogenates decreased in wildtype mice after LPS injection (a). The survival rate was higher in wildtype mice if compared with mice lacking Mmp14, who all died in the first 80 h after LPS injection (b). To assess the severity of lung damage 16 h after LPS injection, histological sections were scored (c–d), the abundance of proteins in the

8

22

312.8

Sepsis of unknown origin

14 16

291 302

14.03 71.86

Abdominal sepsis Abdominal sepsis

bronchoalveolar lavage fluid (BALF, E) and neutrophilic infiltration, expressed as the number of myeloperoxidase (MPO)-positive cell count in lung sections (f–g) were studied. Knockout mice showed a more severe lung damage, with higher injury scores, increased alveolar permeability and neutrophilic infiltrates. *p < 0.05 in post-hoc tests. Scale bars represent 100 μm

J Mol Med (2017) 95:487–497

extravasated blood cells in the lung of Mmp14−/− animals (Fig. 1c–d). In line with this finding, mutant animals had higher protein abundance in BALF (Fig. 1e) and higher myeloperoxidase-positive cell counts (Fig. 1f–g), suggesting an increased inflammatory response within the lung. Effects of Mmp14 deletion on other matrix metalloproteinases To test if some compensatory mechanisms occur in animals lacking Mmp14, expression of the other major collagenases (Mmp8 and Mmp13) was studied in lung homogenates 16 h after LPS injection. There was an increase in Mmp8 expression after LPS injection. This response was more accentuated in Mmp14−/− mice (Fig. 2a). There were no differences in Mmp13 expression after treatment with LPS or between genotypes (Fig. 2b). As MMP-14 is essential for pro-MMP-2 activation, quantitative PCR and gelatin zymography gels were performed to study the differences between genotypes in this process. There Fig. 2 Deficiency of Mmp14 alters the expression of other metalloproteases. qPCR analysis of lung homogenates showed that absence of Mmp14 increased Mmp8 expression after LPS injection (a), with no changes in Mmp13 (b) or Mmp2 expression (c). As MMP-14 is involved in MMP-2 activation, gelatin zymography gels were performed. Activation of MMP-2 was decreased in mutant mice, both after saline or LPS treatment (d). There were no changes in the other gelatinase MMP-9 (e). Panel f shows a representative zymography. *p < 0.05 in post-hoc tests.

491

were no differences in lung Mmp2 gene expression between genotypes or treatments (Fig. 2c). Quantification of zymographies revealed no significant changes in active MMP-2 levels in wildtype mice after LPS injection. However, absence of Mmp14 lead to a constitutive defect in MMP-2 activation regardless of the treatment received (Fig. 2d, f). Additionally, gelatin zymography demonstrated no differences in MMP-9 activity (Fig. 2e–f).

Inflammatory response As we documented an increased inflammatory response in Mmp14−/− animals suggested by the increased lung damage and neutrophil content, we measured the expression of different mediators in lung homogenates to address the underlying mechanisms. The expression of some mediators such as Il6 or Cxcl2 increased after LPS injection whereas the antiinflammatory cytokine Il10 decreased. However, there were no differences in the expression of Il6, Cxcl2, Il10, or Tgfb between wildtype and knockout animals (Fig. 3a–d).

492

J Mol Med (2017) 95:487–497

Matrix metalloproteinases can cleave a number of alarmins, resulting in inactivation of the inflammatory response. Therefore, gene expression and protein levels of alarmins S100A8 and S100A9 were studied. As expected, expression of both genes increased after LPS administration, with no differences between genotypes (Fig. 3e–f). In line with these findings, protein levels of S100A8 were similar in wildtype and knockout mice. However, mice lacking Mmp14 showed higher levels of S100A9 (Fig. 3g–i) in lung homogenates.

structural damage (Fig. 4a) or neutrophilic infiltration (Fig. 4b) in these organs. Moreover, levels of active MMP-2 (Fig. 4c) and S100A9 (Fig. 4d) in these tissues were similar in Mmp14+/+ and Mmp14−/−, suggesting that the lung may be responsible for the differences observed between wildtype and Mmp14-deficient animals.

The lung is the main injured tissue in Mmp14−/− deficient mice

Our results show that absence of MMP-14 is related to an increase in S100A9 that may explain the increased lung inflammatory response observed after LPS injection. As S100A9 is a known substrate of MMP-2, and MMP-14 is required for MMP-2 activation, we tested whether the damage observed in Mmp14-deficient mice could be due to an impaired MMP-2 activation. To do so, cultured lung explants

To test if the differences in survival could be caused by differences in other tissues, liver, heart, spleen, and kidneys from wildtype and knockout animals were harvested after LPS injection. There were no differences between genotypes in

Fig. 3 Inflammatory response. Expression of Il6, Cxcl2, Tgfb, Il10 S100a8, and S100a9 increased after LPS injection was analyzed by qPCR with no differences between genotypes (a–f). There were no significant differences in the S100A8 protein levels between genotypes (g). However, mice lacking

Activated MMP-2 rescues the phenotype of Mmp14−/− lung explants

Mmp14 showed higher levels of S100A9 after LPS injection (h). A representative Western blot of these alarmins is shown (i). *p < 0.05 in post-hoc tests.

J Mol Med (2017) 95:487–497

493

Fig. 4 Impact of Mmp14 deficiency in other organs. Different tissues from wildtype (white bars) and Mmp14 deficient (black bars) mice were obtained and sections were stained with hematoxylin-eosin. No differences in tissue structure (a) or in myeloperoxidasepositive cells (as a marker of neutrophilic infiltration, b) were observed between genotypes in liver, heart, spleen, and kidney. Scale bars represent 100 μm. Additionally, there were no differences between genotypes in activated MMP-2 (c) or S100A9 (d) levels in any of these organs

from wildtype and mutant animals were challenged with LPS, and activated MMP-2 was added. As shown in Fig. 5a, b, addition of activated MMP-2 to the culture medium increased its activity in gelatin zymographies. Moreover, MMP-2 supplementation decreased S100A9 levels in Mmp14−/− animals, but not in Mmp14+/+ (Fig. 5c–d). These results confirm that MMP-2 activation by MMP-14 is essential for S100A9 inactivation. MMP-14 expression is reduced in human sepsis To extend our observations beyond the animal model, blood samples were drawn from critically ill patients, with or without sepsis (ten and six cases, respectively). Clinical data of these patients are shown in Table 2. Soluble MMP-14 can be detected in serum after release from tissues [15]. Abundance of MMP-14 in serum was decreased in patients with sepsis compared to controls (Fig. 6a). In line with our findings,

MMP-2 activity was decreased in serum from septic patients (Fig. 6b). Figure 6c shows a representative Western blot and zymography of these samples.

Discussion Our results show that MMP-14 is decreased during endotoxemia and its absence is related to an enhanced lung inflammatory response and increased mortality after neonatal endotoxemia. The impaired activation of MMP-2 in these mutant mice could lead to an accumulation of alarmins, thus causing a more severe lung injury. Furthermore, data from patients showed a similar response in sepsis. Collectively, these data suggest that MMP-14 could play a role in the activation of a proteolytic cascade aimed to the resolution of acute inflammation.

494

J Mol Med (2017) 95:487–497

Fig. 5 Exogenous MMP-2 restores S100A9 levels in lung explants from Mmp14−/− mice. Lung explants were cultured in presence of LPS, with or without exogenous MMP-2. As expected, levels of this protease increased after supplementation (a–b). Interestingly, levels of S100A9 in the tissues remained constant in explants from wildtype animals (white bars), and increased in explants from knockout mice (black bars) after LPS treatment. Exogenous MMP-2 completely abolished this increase (c). Panel d shows a representative Western blot. *p < 0.05 in post-hoc tests.

MMPs in endotoxemia and sepsis The role of MMPs in sepsis and endotoxic shock is still largely unknown. Nevertheless, it is widely accepted that there is a close interplay between presence of MMPs and activation of the inflammatory response, as their expression is stimulated by the presence of cytokines and the NFκB pathway [16, 17].

Circulating levels of MMP-3, -7, -8, or -9 are elevated in septic patients and some of them correlate to patient’s outcome [18]. MMPs seem to play multitask roles during the different stages of the systemic inflammation, aimed to fight the pathogen invasion by increasing the pro-inflammatory response, but also trying to keep a negative feedback in order to later restore homeostasis. Interaction with the damage associated molecular patterns (DAMP) systems by shedding RAGE from membrane surfaces [19] and by cleaving chemotactic mediators such as alarmins S100A8 and S100A9 [6, 20] follows this strategy. Moreover, they regulate migration of immune cells into the tissue by controlling cytokine bioavailability (i.e., releasing membrane-bound TNF-α or TGF-β deposits in the ECM), chemokine activity (i.e., by processing IL-8), or by simply processing ECM components therefore facilitating cell infiltration [21]. MMP-14

Fig. 6 MMP-14 and -2 in patients with sepsis. MMP-14 was decreased in serum from critically ill patients with sepsis (a). There was also a decrease in active MMP-2 (b). Panel c shows a representative Western blot and a zymography of the samples. *p < 0.05

Genetic ablation of individual MMP genes has allowed the study of their individual contribution in a variety of conditions. Interestingly, most of the animals lacking an Mmp gene have normal phenotypes in baseline conditions, suggesting that there is a substantial overlap in the functions of these enzymes. MMP-14 is an exception to this rule. Absence of Mmp14 gene is related to a severe phenotype that results in death after 3–4 weeks post-partum [11]. Although the ultimate reasons for this premature death are unknown, mutant mice show severe skeletal abnormalities and impaired growth. These abnormalities are accompanied by a strong accumulation of type I collagen in hearts that might contribute to the premature death of these mice, although it still remains

J Mol Med (2017) 95:487–497

unknown. Regarding the lungs, these animals have an abnormal lung development, with a decreased number of alveoli and septa, and emphysema-like changes [22, 23]. These differences are only mild after birth and become evident after the first postnatal week. The experiments described herein were performed before this time point to minimize the impact of developmental defects on the outcomes, although we cannot discard that these underlying abnormalities play a role in the observed outcomes. However, although the structural abnormalities observed in mutant mice could have influenced the differences in survival, they cannot explain the differences in inflammation identified. It has been described that absence of MMP-14 polarizes macrophages towards a pro-inflammatory phenotype by non-proteolytic mechanisms [24]. Similarly, Mmp14−/− mice show an increase in plasma levels of IL-6 15 days after birth [11]. However, we did not find any difference in our experiments. These differences could be due to the measurement at an organ level, and not in cell cultures, and in an earlier time point. Although the endotoxemic model used involves a systemic response, we focused on lung injury, as this was the only organ showing significant differences between genotypes. This could be due to a preferential role of MMP-14 on lung tissue, but also to the specific responsiveness of this organ to RAGE agonists [25]. Moreover, we cannot discard the involvement of other organs, such as the brain, in which a pathogenetic role of MMPs, and specifically MMP-8, has been described [26, 27]. Due to the lack of tissue samples, we measured serum MMP-14 in patients. It has been demonstrated that MMP14 can be released from the cell membrane by different mechanisms including shedding [15] and exocytosis [28], and increased levels of serum MMP-14 have been identified in patients with different conditions [29, 30]. The decreased levels observed in the septic population correlates with the tissue decrease observed in the animal model. Based on these results, MMP-14 could be used as a biomarker of sepsis, although its accuracy remains to be clarified.

MMPs as a mechanism of resolution of inflammation MMPs are released at the sites of inflammation and have been traditionally seen as mediators of damage. In this sense, MMP inhibition has been proposed as a therapeutic strategy to limit injury in inflammatory diseases [31]. However, there is increasing evidence that individual MMPs may play opposite roles, and there is emerging evidence on the contribution of proteases to the resolution of inflammation. In order to develop successful treatments, the specific contributions of individual MMPs must be elucidated, to select specific targets to inhibit or

495

potentiate. This situation is similar to the different, opposed roles of MMPs in cancer [32]. Different MMPs may cleave inflammatory mediators. Although in some cases this proteolytic processing may result in an increased activity, in other cases leads to inactivation of the molecule. In this sense, Greenlee et al. first demonstrated that MMP-2 plays an essential role in the resolution of allergic airway inflammation by cleavage of S100A8 and S100A9 [7]. Additional studies have shown that other collagenases such as MMP-1 and MMP-8 can directly inactivate these alarmins [6, 20]. However, it has been reported that MMP-14 does not directly cleave S100A8 or S100A9 [33]. Our results suggest that the known role of MMP-14 as an activator of a proteolytic cascade involving other MMPs is also relevant in endotoxemia. The downregulation of MMP-14 after LPS administration and in septic patients could represent a switch towards a proinflammatory state, as S100A9 cleavage by MMP-2 requires its activation by MMP-14. This mechanism opens the possibility of new anti-inflammatory strategies in sepsis based on MMP-14 or active MMP-2. For instance, growth factors such as VEGF, with beneficial effects in lung injury [34, 35], promote MMP-14 expression [36]. However, this proposal must be viewed with caution due to the limitations of our work: First, our experimental design shows only a strong association between the increased S100A9 levels and the subsequent increased lung injury, but a firm causality relationship cannot be established or the involvement of other molecules excluded. Second, endotoxemia and sepsis are different phenomena, and although we have identified some common features regarding low MMP-14 levels, the underlying mechanisms could be different. Third, as these proteases have pleiotropic effects, their therapeutic use could lead to unintended consequences, including cancer dissemination or additional tissue disruption. Further experimental studies with more complex models should be carried out to explore this therapeutic strategy. In conclusion, we have explored the role of MMP-14 in endotoxemia, and found that this enzyme is required to trigger the anti-inflammatory effects of MMP-2. The findings that MMP-14 levels are decreased in both the mouse model and in septic patients points to a potential role of this protease as a biomarker and a pathogenetic mechanism in septic shock. Acknowledgements The authors thank Carlos López-Otín for his help during the development of this work, and all the personnel at the Intensive Care Unit and the Clinical Laboratory in Hospital Valle del Nalón and Hospital Universitario Central de Asturias for their help during collection of samples from patients. Funding This is supported by Fundación para el Fomento en Asturias de la Investigación Científica aplicada y la Tecnología (FICYT,

496

J Mol Med (2017) 95:487–497

GRUPIN14-089). Guillermo M Albaiceta is the recipient of a grant from Instituto de Salud Carlos III (INT15-002). Compliance with ethical standards 15. Conflict of interest The authors have no conflicts of interest. 16.

References 1. 2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

Gotts JE, Matthay MA (2016) Sepsis: pathophysiology and clinical management. BMJ 353:i1585 Hotchkiss RS, Monneret G, Payen D (2013) Sepsis-induced immunosuppression: from cellular dysfunctions to immunotherapy. Nat Rev Immunol 13:862–874 Matute-Bello G, Downey G, Moore BB, Groshong SD, Matthay MA, Slutsky AS, Kuebler WM (2011) An official American Thoracic Society workshop report: features and measurements of experimental acute lung injury in animals. Am J Respir Cell Mol Biol 44:725–738 Aschner Y, Zemans RL, Yamashita CM, Downey GP (2014) Matrix metalloproteinases and protein tyrosine kinases: potential novel targets in acute lung injury and ARDS. Chest 146:1081–1091 Tester AM, Cox JH, Connor AR, Starr AE, Dean RA, Puente XS, Lopez-Otin C, Overall CM (2007) LPS responsiveness and neutrophil chemotaxis in vivo require PMN MMP-8 activity. PLoS One 2:e312 Gonzalez-Lopez A, Aguirre A, Lopez-Alonso I, Amado L, Astudillo A, Fernandez-Garcia MS, Suarez MF, Batalla-Solis E, Colado E, Albaiceta GM (2012) MMP-8 deficiency increases TLR/RAGE ligands S100A8 and S100A9 and exacerbates lung inflammation during endotoxemia. PLoS One 7:e39940 Greenlee KJ, Corry DB, Engler DA, Matsunami RK, Tessier P, Cook RG, Werb Z, Kheradmand F (2006) Proteomic identification of in vivo substrates for matrix metalloproteinases 2 and 9 reveals a mechanism for resolution of inflammation. J Immunol 177:7312–7321 Strongin AY, Collier I, Bannikov G, Marmer BL, Grant GA, Goldberg GI (1995) Mechanism of cell surface activation of 72kDa type IV collagenase. Isolation of the activated form of the membrane metalloprotease J Biol Chem 270:5331–5338 Atkinson JJ, Holmbeck K, Yamada S, Birkedal-Hansen H, Parks WC, Senior RM (2005) Membrane-type 1 matrix metalloproteinase is required for normal alveolar development. Dev Dyn 232:1079–1090 Holmbeck K, Bianco P, Caterina J, Yamada S, Kromer M, Kuznetsov SA, Mankani M, Robey PG, Poole AR, Pidoux I et al (1999) MT1-MMP-deficient mice develop dwarfism, osteopenia, arthritis, and connective tissue disease due to inadequate collagen turnover. Cell 99:81–92 Gutierrez-Fernandez A, Soria-Valles C, Osorio FG, Gutierrez-Abril J, Garabaya C, Aguirre A, Fueyo A, Fernandez-Garcia MS, Puente XS, Lopez-Otin C (2015) Loss of MT1-MMP causes cell senescence and nuclear defects which can be reversed by retinoic acid. EMBO J 34:1875–1888 Gonzalez-Lopez A, Astudillo A, Garcia-Prieto E, FernandezGarcia MS, Lopez-Vazquez A, Batalla-Solis E, Taboada F, Fueyo A, Albaiceta GM (2011) Inflammation and matrix remodeling during repair of ventilator-induced lung injury. Am J Physiol Lung Cell Mol Physiol 301:L500–L509 Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, Bellomo R, Bernard GR, Chiche JD, Coopersmith CM et al (2016) The third international consensus definitions for sepsis and septic shock (sepsis-3). JAMA 315: 801–810 Vincent JL, Moreno R, Takala J, Willatts S, De Mendonca A, Bruining H, Reinhart CK, Suter PM, Thijs LG (1996) The SOFA

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

(Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on SepsisRelated Problems of the European Society of Intensive Care Medicine. Intensive Care Med 22:707–710 Osenkowski P, Toth M, Fridman R (2004) Processing, shedding, and endocytosis of membrane type 1-matrix metalloproteinase (MT1-MMP). J Cell Physiol 200:2–10 Soria-Valles C, Gutierrez-Fernandez A, Osorio FG, Carrero D, Ferrando AA, Colado E, Fernandez-Garcia MS, BonzonKulichenko E, Vazquez J, Fueyo A, Lopez-Otin C (2016) MMP25 metalloprotease regulates innate immune response through NFkappaB signaling. J Immunol 197:296–302 Fanjul-Fernandez M, Folgueras AR, Cabrera S, Lopez-Otin C (2010) Matrix metalloproteinases: evolution, gene regulation and functional analysis in mouse models. Biochim Biophys Acta 1803:3–19 Yazdan-Ashoori P, Liaw P, Toltl L, Webb B, Kilmer G, Carter DE, Fraser DD (2011) Elevated plasma matrix metalloproteinases and their tissue inhibitors in patients with severe sepsis. J Crit Care 26: 556–565 Yamakawa N, Uchida T, Matthay MA, Makita K (2011) Proteolytic release of the receptor for advanced glycation end products from in vitro and in situ alveolar epithelial cells. Am J Physiol Lung Cell Mol Physiol 300:L516–L525 Fanjul-Fernandez M, Folgueras AR, Fueyo A, Balbin M, Suarez MF, Fernandez-Garcia MS, Shapiro SD, Freije JM, Lopez-Otin C (2013) Matrix metalloproteinase Mmp-1a is dispensable for normal growth and fertility in mice and promotes lung cancer progression by modulating inflammatory responses. J Biol Chem 288:14647– 14656 Gueders MM, Foidart JM, Noel A, Cataldo DD (2006) Matrix metalloproteinases (MMPs) and tissue inhibitors of MMPs in the respiratory tract: potential implications in asthma and other lung diseases. Eur J Pharmacol 533:133–144 Irie K, Komori K, Seiki M, Tsuruga E, Sakakura Y, Kaku T, Yajima T (2005) Impaired alveolization in mice deficient in membranetype matrix metalloproteinase 1 (MT1-MMP). Med Mol Morphol 38:43–46 Oblander SA, Zhou Z, Galvez BG, Starcher B, Shannon JM, Durbeej M, Arroyo AG, Tryggvason K, Apte SS (2005) Distinctive functions of membrane type 1 matrix-metalloprotease (MT1-MMP or MMP-14) in lung and submandibular gland development are independent of its role in pro-MMP-2 activation. Dev Biol 277:255–269 Shimizu-Hirota R, Xiong W, Baxter BT, Kunkel SL, Maillard I, Chen XW, Sabeh F, Liu R, Li XY, Weiss SJ (2012) MT1-MMP regulates the PI3Kdelta.Mi-2/NuRD-dependent control of macrophage immune function. Genes Dev 26:395–413 Guo WA, Knight PR, Raghavendran K (2012) The receptor for advanced glycation end products and acute lung injury/acute respiratory distress syndrome. Intensive Care Med 38:1588–1598 Folgueras AR, Fueyo A, Garcia-Suarez O, Cox J, Astudillo A, Tortorella P, Campestre C, Gutierrez-Fernandez A, FanjulFernandez M, Pennington CJ et al (2008) Collagenase-2 deficiency or inhibition impairs experimental autoimmune encephalomyelitis in mice. J Biol Chem 283:9465–9474 Vandenbroucke RE, Dejonckheere E, Van Lint P, Demeestere D, Van Wonterghem E, Vanlaere I, Puimege L, Van Hauwermeiren F, De Rycke R, Mc Guire C et al (2012) Matrix metalloprotease 8dependent extracellular matrix cleavage at the blood-CSF barrier contributes to lethality during systemic inflammatory diseases. J Neurosci 32:9805–9816 Hakulinen J, Sankkila L, Sugiyama N, Lehti K, Keski-Oja J (2008) Secretion of active membrane type 1 matrix metalloproteinase (MMP-14) into extracellular space in microvesicular exosomes. J Cell Biochem 105:1211–1218

J Mol Med (2017) 95:487–497 29.

30.

31.

32.

Kaimal R, Aljumaily R, Tressel SL, Pradhan RV, Covic L, Kuliopulos A, Zarwan C, Kim YB, Sharifi S, Agarwal A (2013) Selective blockade of matrix metalloprotease-14 with a monoclonal antibody abrogates invasion, angiogenesis, and tumor growth in ovarian cancer. Cancer Res 73:2457–2467 Ren XH, Peng XD, Wu XP, Liao EY, Sun ZQ (2008) Association between serum soluble membrane type matrix metalloproteinase-1 (MT1-MMP) levels and bone mineral density, and biochemical markers in postmenopausal women. Clin Chim Acta 390:44–48 Vandenbroucke RE, Libert C (2014) Is there new hope for therapeutic matrix metalloproteinase inhibition? Nat Rev Drug Discov 13:904–927 Overall CM, Lopez-Otin C (2002) Strategies for MMP inhibition in cancer: innovations for the post-trial era. Nat Rev Cancer 2:657–672

497 33.

34.

35.

36.

Stegemann C, Didangelos A, Barallobre-Barreiro J, Langley SR, Mandal K, Jahangiri M, Mayr M (2013) Proteomic identification of matrix metalloproteinase substrates in the human vasculature. Circ Cardiovasc Genet 6:106–117 Song J, Lu H, Zheng X, Huang X (2015) Effects of vascular endothelial growth factor in recovery phase of acute lung injury in mice. Lung 193:1029–1036 Koh H, Tasaka S, Hasegawa N, Yamada W, Shimizu M, Nakamura M, Yonemaru M, Ikeda E, Adachi Y, Fujishima S et al (2007) Protective role of vascular endothelial growth factor in endotoxininduced acute lung injury in mice. Respir Res 8:60 Ding BS, Nolan DJ, Guo P, Babazadeh AO, Cao Z, Rosenwaks Z, Crystal RG, Simons M, Sato TN, Worgall S et al (2011) Endothelial-derived angiocrine signals induce and sustain regenerative lung alveolarization. Cell 147:539–553