Matrix metalloproteinase (MMP)-9 is crucial in atherosclerotic plaque rupture and tissue remodeling after a cardiac event. The balance between MMP-9 and endogenous inhibitor, tissue inhibitors of matrix metalloproteinase 1 (TIMP-1), is important in acute coronary syndrome (ACS). This is an age- and gender-matched case-control study of ACS (N = 669). Patients (45.7%) were resampled after recovery, and all were followed up for 6 years. The molecular forms of MMP-9 were investigated by gelatin zymography. Diagnostically, MMP-9 and the MMP-9/TIMP-1 molar ratio were associated with ACS (OR 5.81, 95% CI 2.65–12.76, and 4.96, 2.37–10.38). The MMP-9 concentrations decreased 49% during recovery (p < 0.001). The largest decrease of these biomarkers between acute and recovery phase (ΔMMP-9) protected the patients from major adverse cardiac events, especially the non-fatal events. The fatal events were associ- ated with in vitro activatable MMP-9 levels (p = 0.028). Serum MMP-9 and the MMP-9/TIMP-1 molar ratio may be valuable in ACS diagnosis and prognosis. High serum MMP-9 activation potential is associated with poor cardiovascular outcome. . . . . . Keywords Atherosclerosis Coronary artery disease Serum biomarker Cardiovascular diseases Plaque rupture Inflammation Introduction Atherosclerosis is a chronic inflammatory process of arteries Associate Editor Craig Stolen oversaw the review of this article [1, 2]. Matrix metalloproteinases (MMPs) destabilize athero- Electronic supplementary material The online version of this article sclerotic plaques by degrading extracellular matrix (ECM), (https://doi.org/10.1007/s12265-018-9789-x) contains supplementary material, which is available to authorized users. especially in the shoulder regions. This may lead to plaque rupture and a fatal acute coronary syndrome (ACS) event. * Laura Lahdentausta Inflammatory and oxidative mediators increase the amounts firstname.lastname@example.org of MMPs . MMPs enable leukocytes and inflammatory mediators to migrate across tissues , accelerating the devel- opment of pathogenic atherosclerotic plaques. Department of Oral and Maxillofacial Diseases, University of Helsinki and Helsinki University Hospital, Helsinki, Finland Matrix metalloproteinase-9 (MMP-9), also known as Department of Periodontology, Operative and Preventive Dentistry, gelatinase B, is an enzyme that degrades mainly type IV col- University Hospital Bonn, Bonn, Germany lagen and elastin . MMP-9 is secreted by various cell types, Division of Periodontology, Department of Dental Medicine, such as neutrophils, macrophages, endothelial cells, and Karolinska Institutet, Huddinge, Sweden smooth muscle cells. Interactions with specific tissue inhibi- Skåne University, Skåne, Sweden tors of matrix metalloproteinases (TIMPs) determine the func- tion of MMP-9 [6, 7] by binding to MMP at a molar equiva- Department of Paediatrics, Division of Paediatric Cardiology, Skåne University Hospital, Lund, Sweden lence . Inactive, latent pro-form MMP-9 may be activated J. of Cardiovasc. Trans. Res. (2018) 11:210–220 211 by reactive oxygen species (ROS), trypsin, chymotrypsin, or Patient history and use of medications were registered, and bacterial proteases . In addition, MMP-9 can be activated blood samples were drawn. chemically with APMA (p-aminophenylmercuric acetate) Cases were diagnosed with typical symptoms, laboratory in vitro . measurements, and electrocardiogram (ECG). AMI and UAP Significantly elevated plasma MMP-9 concentrations have were diagnosed according to prevailing criteria in 2000. AMI been previously reported in ACS patients [11–13]. Elevated was diagnosed if the patient had two of the following criteria: serum MMP-9 concentration was associated with plaque rup- chest pain related to exercise lasting over 20 min or changes in ture when compared to stable angina pectoris patients . ECG, such as ST-elevations followed by T-wave inversion or However, there is no information available, if the MMP-9 new Q-waves, or an increase in Creatine kinase-MB concentrations decrease during the recovery or if the recovery (CK-MB) to more than twice the upper limit of the normal levels associate with recurrent cardiovascular (CV) events. In value (> 5 μg/l). UAP was diagnosed if two of following general, only a few follow-up studies have been published. In criteria were fulfilled: continuous chest pain, ST-segment de- one study, ST-elevation myocardial infarction (STEMI) pa- pression in the ECG (<1 mm), or elevation of CK-MB (5 < tients showed higher serum MMP-9 and MMP-9/TIMP-1 CK-MB < 10 μg/l) or troponin T (0.05 < TnT < 0.10 μg/l). values compared to subjects without coronary artery disease The controls (N = 326) did not have previous coronary (CAD), but they had no prognostic value in a 2-year follow-up heart disease, stroke, or angina-like chest pain and were not . In another study with a longer follow-up (mean taking medication for dyslipidemia, hypertension, or diabetes. 4.1 years), elevated MMP-9 was associated with an adverse They were selected from the same suburbs as the patients, and event . the groups were matched for gender and age ± 2 years. The In the present study, we determined serum MMP-9 and control samples were collected and stored similarly as those of TIMP-1 concentrations and calculated their molecular ratio. the cases. The determinations were done on controls at baseline and on All subjects of the study were followed up on average for ACS patients in the acute and in the recovery phase. The 6 years (range 4.56–7.13) to the end of the study or to a major patients were followed up for 6 years. Using these serum adverse cardiac event (MACE), i.e., cardiovascular death or concentrations, our aims were to investigate (1) their associa- hospitalization for an ACS event. During the follow-up, 150 tion and diagnostic value in ACS in a cross sectional setting, patients suffered MACE, including 61 fatal and 89 non-fatal (2) the difference between acute and recovery phase (i.e., Δ events. Among the patients, whose recovery samples were values), (3) their prognostic value by using the follow-up data, available, 63 patients suffered a MACE, including 14 fatal and (4) the relevance of MMP-9 activation degree in ACS. and 49 non-fatal events. A subsample of patients who suffered a non-fatal (N =23) or fatal (N = 7) MACE and whose both acute and recovery phase samples were available were randomly selected for gel- Materials and Methods atin zymography. Zymographies were run also on a set of controls (N = 28) that were matched for age and gender to Subjects and Diagnosis the cases on a group level. The total number of samples in these analyses was 58. The patients in this subsample differed Data was collected between March 1999 and April 2002 as from the rest of the cases only as regards to systolic blood described earlier [17–19]. Cases were ACS patients admitted pressure (150 vs. 130 mmHg, p =0.033). in the heart intensive care unit at Lund University Hospital. The 343 patients included 108 unstable angina pectoris (UAP) Laboratory Determinations and 235 acute myocardial infarction (AMI) patients. Forty-eight patients, who were invited to the study, chose Serum without activators was collected from the patients with- not to participate and 21 patients died before the appointment. in 24 h after the diagnosis. The research nurse collected 70 ml After a 6-month recovery period (median 350 days, IQR of blood from controls. The blood samples were transferred on 434 days), 157 (45.8%) patients agreed to participate in the ice to the central laboratory for centrifugation. All samples resampling. were frozen (− 20 °C) until laboratory analyses. In the The inclusion criteria for both cases and controls were age in-house laboratory quality control, no significant effect of under 80 years, no cognitive intellectual disability, and no storage time on biomarker levels has been observed . operations or chemotherapy within the previous 4 weeks. TIMP-1-ELISA (R&D Systems, Minneapolis, MN, USA) Twelve patients were excluded from the study because of and MMP-9-ELISA (GE Healthcare UK Limited, Amersham aortic aneurysm, pulmonary embolism, myocarditis, pericar- Place, UK) were performed according to manufacturer’sin- ditis, unspecific precordial pain, or atrial fibrillation. Data on structions on diluted samples (1:10 in TIMP-1 and 1:20 in demographic factors were collected by a questionnaire. MMP-9). The inter-assay CV% of TIMP-1 was 8.2% (N = 212 J. of Cardiovasc. Trans. Res. (2018) 11:210–220 12) and for MMP-9 9.2% (N = 12). The detection limits were non-parametric Mann-Whitney U test or Wilcoxon 0.08 and 0.05 ng/ml, respectively. TIMP-1 can form a com- signed-rank test. The diagnostic sensitivity and specificity of plex with MMP-9 in a 1:1 stoichiometry with a high affinity MMP-9 and MMP-9/TIMP-1 were calculated by receiver op- . For calculation of MMP-9/TIMP-1, stoichiometric mo- erating characteristics (ROC) from logarithmically trans- lecular ratio molecular weights of 92,000 g/mol (MMP-9) and formed values. Two different multivariate logistic regression 28,000 g/mol (TIMP-1) were used. The recovery phase levels models were used to determine the association of MMP-9 and of MMP-9 and MMP-9/TIMP-1 were subtracted from those MMP-9/TIMP-1 molar ratio with ACS. The first model was of the acute phase to calculate the Δ levels. stratified for age and sex; the second model was stratified for Gelatin zymography was performed using a previously de- age, sex, and adjusted for C-reactive protein (CRP), cholester- scribed technique  with minor modifications to explore ol, and smoking. For the follow-up data, we used Cox regres- activatable MMP-9 in vitro. The incubation time (16 h) and sion model adjusted for age and sex. MACE and its sub- sample volume (1 μl) were based on pilot zymography anal- groups, fatal and non-fatal MACE, were used as endpoints. yses with self-cast gels using fluorescent gelatin substrate. Correlation analyses were performed by Spearman correla- Gelatin zymography analyses of selected samples were per- tion. The analyses were performed using IBM SPSS formed with commercial gels (BIO-RAD 10% Ready Gel® Statistics 22. Zymogram Gel, 10 well, 50 μl, #161-1167, CA, USA). Experiments were conducted with serum samples as such and with samples processed with APMA for in vitro activation Results of MMP-9. Serum was incubated at a final concentration of 1 mM APMA at + 37 °C for 1 h. Subsequently, all samples Acute Phase were incubated in Laemmli buffer at room temperature for 1 h. After incubation, the samples were applied on gels and elec- Characteristics of the cases and controls are presented accord- trophoresis was performed in + 4 °C for 2 h. After electropho- ing to acute phase serum MMP-9 quartiles (Table 1). CRP and resis, the gels were incubated at room temperature in the first smoking status differed statistically significantly between the wash buffer (50 mM Tris-HCl, pH 7.5, containing 2.5% quartiles in both groups. No other significant differences were Tween 80 and 0.02% NaN ) for 30 min, in the second buffer observed. More detailed characteristics of the subjects have (50 mM Tris-HCl, pH 7.5, containing 2.5% Tween 80 and been presented previously . The mean (SD) age of the 0.02% NaN and supplemented with 1 μM ZnCl ,5 mM cases and controls was 63.3 (9.2) and 63.0 (9.2) years and 3 2 CaCl ) at room temperature for further 30 min, and in the last they included 21.3% and 22.1% women, respectively. Of the buffer (the same as the second buffer but without 2.5% Tween cases, 235 patients (68.5%) were diagnosed with AMI and 80) at 37 °C for 16 h. The gels were stained with 0.1% 108 (31.5%) with UAP. Coomassie Brilliant Blue and destained with 20% methanol/ The serum MMP-9 and MMP-9/TIMP-1 molar ratio were 10% acetic acid. Low-range prestained SDS-PAGE standards significantly higher in ACS patients than in controls (BIO-RAD, #161-0305, CA, USA) were used in every gel. (p < 0.001) (Table 2; Supplement 1). The same difference MMP-9 (Proteaimmun 100 ng/μl) was used as a positive con- wasseeninboth AMI (p < 0.001) and UAP (p <0.001). trol in every gel, and the molecular weights of the gelatinolytic Both MMP-9 (p < 0.001) and MMP-9/TIMP-1 (p = 0.023) zones were compared to the positive control. The gels were were higher in AMI compared to UAP. The association of scanned with LI-COR ODYSSEY (Lincoln, NE, USA), and serum levels with ACS was investigated with multivariate data was analyzed with Image Studio software. The data was logistic regression models, which are presented in Table 3. presented as arbitrary scanning units of intensities. Both MMP-9 concentration and MMP-9/TIMP-1 were strongly associated with ACS and its subgroups. Statistics Diagnostic Ability of Serum MMP-9 The distribution of variables was tested before statistical anal- and MMP-9/TIMP-1 ysis. Normally distributed continuous variables are presented as means and standard deviations (SD). The statistical signif- In ROC analyses, both MMP-9 and the MMP-9/TIMP-1 icance of the differences between the groups was tested by molar ratio distinguished ACS from the controls with an Student’s t test or ANOVA of logarithmically transformed AUC (95%) of 0.742 (0.704–0.781, p < 0.001) and 0.702 values. Categorical variables were tested by chi-square test. (0.662–0.742, p < 0.001), respectively. Smoking de- The ELISA measurements of MMP-9 and TIMP-1 as well as creased the diagnostic value of serum MMP-9 in ACS; the gelatin zymography results displayed a skewed distribu- in the ROC analysis, the sensitivity and specificity were tion and are presented as medians and interquartile ranges higher in non-smokers than smokers with an AUC of (IQR). Statistical significance was tested by using 0.765 (0.722–0.808, p < 0.001) and 0.659 (0.560–0.758, J. of Cardiovasc. Trans. Res. (2018) 11:210–220 213 Table 1 Baseline characteristics Quartilesofserum MMP-9concentrations of cases (acute phase) and controls in quartiles of serum 1st 2nd 3rd 4th MMP-9 concentrations Mean (SD) p Age (years) Cases 63.7 (9.4) 63.2 (9.7) 62.7 (8.9) 63.7 (8.6) NS Controls 62.5 (9.6) 62.8 (8.6) 63.7 (9.6) 63.1 (9.2) NS Cholesterol (mmol/l) Cases 5.5 (1.1) 5.2 (1.6) 5.3 (1.3) 5.2 (1.0) NS Controls 5.6 (1.0) 5.8 (1.1) 5.9 (1.0) 5.8 (1.1) NS CRP (mg/l) Cases 12.6 (24.7) 14.6 (18.2) 23.1 (31.3) 53.3 (62.9) <0.001 Controls 1.8 (1.7) 2.1 (2.2) 2.7 (2.9) 2.5 (2.5) 0.035 N (%) p Sex (% men) Cases 65 (76.5) 65 (75.6) 66 (76.7) 74 (86.0) NS Controls 57 (70.4) 61 (74.4) 69 (84.1) 67 (82.7) NS Current smoker Cases 12 (16.2) 7 (9.1) 14 (20.0) 22 (31.4) 0.006 Controls 7 (9.5) 12 (15.0) 14 (17.5) 33 (41.3) <0.001 Diabetic Cases 11 (13.1) 11 (13.3) 11 (13.4) 14 (16.5) NS Controls –––– – Lipid-lowering medication Cases 15 (17.9) 22 (26.5) 23 (28.0) 17 (20.0) NS Controls –––– – MACE in follow-up Cases 30 (35.3) 39 (45.3) 43 (50.0) 38 (44.2) NS Controls 8 (9.9) 5 (6.1) 9 (11.1) 9 (11.1) NS Fatal Cases 9 (10.6) 12 (14.6) 18 (20.9) 22 (25.6) NS Controls 4 (4.9) 3 (3.7) 2 (2.5) 4 (4.9) NS Non-fatal Cases 21 (24.7) 27 (31.4) 25 (29.1) 16 (18.6) NS Controls 4 (4.9) 2 (2.4) 7 (8.5) 7 (8.6) NS Significant values are in italics NS not significant Numbers of cases and controls in quartiles: 1st 85 and 81, 2nd 86 and 82, 3rd 86 and 82, and 4th 86 and 81 ANOVA of log-transformed values Chi-square test p = 0.003), respectively. Serum MMP-9 had a significant Follow-Up and Prognostics correlation with CRP (r = 0.453, p < 0.001), CK-MB (0.278, p < 0.001), and troponin T (0.291, p < 0.001). The cases were followed up for 6 years, and MACE was registered. Cases experiencing MACE during follow-up had significant differences in CRP between the acute phase Recovery Phase MMP-9 quartiles (CRP increasing in a dose-dependent man- ner). No other significant differences were observed in the Recovery phase samples were collected from ACS patients characteristics between the quartiles (data not shown). not earlier than 6 months after the acute event. We evaluated if serum MMP-9 and MMP-9/TIMP-1 molar Characteristics of patients in recovery phase and according ratio or their Δ values predict MACE. Elevated acute phase to the difference of MMP-9 concentration between acute MMP-9 was a significant predictor of fatal events with a HR and recovery phase, i.e., ΔMMP-9, are presented in the 2.88 (1.32–6.27, p = 0.025, Q4 vs. Q1), but not non-fatal ones. supplementary material (Supplements 2 and 3). In the re- Elevated recovery phases MMP-9 and MMP-9/TIMP-1 were covery phase relative to the acute phase, serum MMP-9 significant predictors of MACE with HRs 4.15 (1.87–9.23, concentrations and MMP-9/TIMP-1 decreased 49 and p = 0.006) and 3.32 (1.48–7.42, p =0.009) (Fig. 1), especially 34% (p < 0.001), respectively (Table 2; Supplement 1). non-fatal events [6.05 (2.27–16.12), p = 0.004 and 6.89 (2.52– The recovery phase MMP-9 concentrations did not differ 18.83), p =0.001]. significantly from those of controls, but the MMP-9/ The highest ΔMMP-9 values presented protective predic- TIMP-1 ratio remained significantly higher than the levels tive value of MACE with a HR 0.44 (0.20–0.97, p =0.003) observed in controls (Table 2; Supplement 1). (Fig. 1) and especially non-fatal events [0.24 (0.09–0.59), p = 214 J. of Cardiovasc. Trans. Res. (2018) 11:210–220 Table 2 Median serum MMP-9 concentrations in the recovery phase relative to the acute phase MMP-9 (ng/ml) Median (IQR) p value N Compared Compared Compared a b c to controls to acute phase to Bno endpoint^ Controls 326 150.2 (189.4) Cases Acute phase ACS 343 343.5 (298.7) <0.001 –– UAP 108 302.4 (278.4) <0.001 –– AMI 235 375.0 (330.4) <0.001 –– MACE in the follow-up No endpoint 193 322.2 (318.2) <0.001 –– Non-fatal 89 327.6 (234.8) 0.011 – <0.001 Fatal 61 419.3 (266.5) <0.001 – <0.001 Recovery phase ACS 157 174.6 (225.3) NS <0.001 – UAP 56 178.2 (255.4) NS 0.013 – AMI 101 172.9 (184.9) NS <0.001 – MACE in the follow-up No endpoint 94 182.4 (226.6) NS <0.001 – Non-fatal 49 167.6 (214.0) NS <0.001 NS Fatal 14 138.4 (295.9) NS 0.003 NS The statistically significant p values are in italics NS not significant Mann-Whitney test Wilcoxon signed-rank test 0.001]. Also, high ΔMMP-9/TIMP-1 was protective from APMA; the difference was significant for all comparisons non-fatal MACE with a HR 0.29 (0.11–0.73, p =0.019). (p ≤ 0.001 for all). Other gelatinolytic active MMP, MMP-2, was observed only in pro-form (72 kDa), but not in proteolyt- Activatable MMP-9 ically activated form (64 kDa) (Fig. 2b). High molecular size (> 100 kDa) gelatinolytic species were also detected. Zymography was utilized to explore activatable MMP-9 Medians of active and APMA-activatable MMP-9 were in vitro. The serum MMP-9 concentrations obtained by compared between cases and controls. When cases were ana- ELISA correlated significantly with the total intensity units lyzed as one group (endpoint being MACE), there were no obtained by gelatin zymography (r = 0.397, p < 0.001) statistically significant differences in the active and activatable (Fig. 2a). In all samples, MMP-9 could be activated with MMP-9 between the cases and controls in acute phase. Table 3 The association of serum MMP-9 and MMP-9/TIMP-1 quartiles with ACS at baseline OR (95% CI) p 1st 2nd 3rd 4th ACS MMP-9 Model 1 1 1.92 (1.20–3.08) 5.62 (3.48–9.08) 10.37 (6.20–17.35) < 0.001 Model 2 1 1.64 (0.81–3.30) 4.84 (2.41–9.72) 5.81 (2.65–12.76) < 0.001 MMP-9/TIMP-1 Model 1 1 1.78 (1.12–2.82) 5.22 (3.25–8.37) 6.29 (3.87–10.22) < 0.001 Model 2 1 1.79 (0.87–3.67) 5.34 (2.61–10.91) 4.96 (2.37–10.38) < 0.001 UAP MMP-9 Model 1 1 1.81 (0.61–2.30) 3.72 (2.00–6.94) 4.49 (2.28–8.85) < 0.001 Model 2 1 0.89 (0.34–2.31) 3.09 (1.29–7.37) 3.81 (1.39–10.47) 0.004 MMP-9/TIMP-1 Model 1 1 1.29 (0.66–2.52) 3.76 (1.98–7.13) 3.75 (1.93–7.26) < 0.001 Model 2 1 1.08 (0.40–2.89) 3.18 (1.28–7.89) 3.89 (1.55–9.74) 0.005 AMI MMP-9 Model 1 1 2.70 (1.50–4.86) 7.65 (4.25–13.77) 16.82 (9.13–30.98) < 0.001 Model 2 1 3.01 (1.10–8.27) 9.05 (3.32–24.62) 10.97 (3.72–32.35) < 0.001 MMP-9/TIMP-1 Model 1 1 2.14 (1.23–3.70) 6.28 (3.62–10.87) 8.34 (4.76–14.61) < 0.001 Model 2 1 3.55 (1.26–10.00) 11.46 (4.09–32.06) 8.73 (3.03–25.10) < 0.001 Model 1 stratified for age and sex (N = 654); model 2 stratified for age, sex, and adjusted for CRP, cholesterol concentration, and smoking (N =504) J. of Cardiovasc. Trans. Res. (2018) 11:210–220 215 Fig. 1 Cumulative survival according to the serum MMP-9, MMP-9/TIMP-1, ΔMMP-9, and ΔMMP-9/TIMP-1 quartiles in ACS patients, endpoint event being MACE. MACE includes both fatal and non-fatal endpoints. ΔMMP-9 and ΔMMP-9/TIMP-1 refer to difference between acute and recovery phase (i.e., acute— recovery phase values). The survival was investigated by Cox regression model adjusted for age and sex However, both molecular forms of MMP-9 decreased in the Additionally, values were calculated separately in cases with recovery phase (p = 0.179 for active and p = 0.003 for different endpoints both in the acute and recovery phase activatable) below the levels observed in controls. (Fig. 3). In cases with different endpoints, active and Fig. 2 Gelatin zymography results. a Scatter plot of serum MMP-9 levels measured by ELISA and total MMP-9 intensities analyzed by gelatin zymography. The correlation coefficient and p value are shown. b Representative gelatin- zymography of ACS serum samples. Lane 1 is the molecular weight standard. Lanes 2, 4, and 6 are serum samples of patients with ACS without MMP-9 activating pretreatment. Lanes 3, 5, and 7 are the same serum samples, respectively, with 1 mM APMA pretreatment, which activates the pro-form of MMP-9. The gels are 10%, and the bands were visualized by Coomassie Brilliant Blue staining. Pro- MMP-2 bands are seen at 72 kDa. No proteolytically activated MMP-2 was observed (64 kDa) 216 J. of Cardiovasc. Trans. Res. (2018) 11:210–220 activatable MMP-9 decreased from the acute phase to the follow-up correlated positively with initial stroke severity and recovery phase, but the difference was significant only for outcome. In this study during the short follow-up, the MMP-9 activatable MMP-9 in cases suffering a non-fatal MACE concentrations did not reach the levels observed in the controls (p =0.018) (Fig. 3). Patients with fatal MACE during and the Δ values were not as valuable as in our study. follow-up had significantly higher activatable MMP-9 In our study, increased serum MMP-9 concentrations were zymography intensities (p = 0.028) and MMP-9 ELISA con- associated with both AMI and UAP, which confirms previous centrations (p < 0.001) in acute phase than controls (Fig. 3). In findings . In patients with STEMI, ACS, and non-STEMI the zymography subpopulation, MMP-9 concentrations ob- , the MMP-9 concentrations elevate earlier than the clas- tained by ELISA differed statistically significantly between sical myocardial damage marker, high-sensitivity troponin T, acute and recovery phase in cases with non-fatal (p <0.001) presumably reflecting plaque rupture. MMP-9 concentrations and fatal MACE (p =0.028) (Fig. 3). were highest after admission to hospital in patients with Proteolytically activated MMP-2 lanes (64 kDa) were not non-STEMI but decreased rapidly during the following 48 h observed in our zymography analysis, only MMP-9 was pro- . MMP-9 had a significant correlation with teolytically activated (Fig. 2b). Medians of pro-MMP-2 myocardium-specific markers CK-MB and troponin T, sug- (72 kDa) were significantly higher in cases with a fatal gesting that ischemic myocardial tissue is a source of systemic MACE when compared to controls both without (p =0.013) MMP-9 after an ACS event. Myocardial damage can induce and with (0.022) APMA-activation and when compared to enhanced MMP expression and activation . In our study, non-fatal MACE without (p = 0.016) and with (p =0.025) elevated MMP-9 levels were found also during the recovery APMA-activation (Supplement 4). The corresponding values phase, which suggests that serum MMP-9 may derive also did not differ significantly between non-fatal MACE and from other source than plaque rupture. Cytokines and other controls. pro-inflammatory mediators increase the synthesis and secre- tion of MMP-9 from inflammatory cells . CRP increases MMP-9expressioninsmoothmusclecells ina Discussion dose-dependent manner and correlates with MMP-9 levels in ACS patients . This correlation was also observed in the Serum MMP-9 concentrations were elevated in the acute present study. Serum MMP-9 may reflect a systemic inflam- phase of ACS and they generally decreased during the recov- matory state , but genetic variation may also contribute to ery. The high levels during the acute phase predicted fatal the MMP-9 levels and activity  and thereby to the risk of MACE. Patients with high MMP-9 concentrations or coronary artery disease . MMP-9/TIMP-1 ratio that persisted during the course of re- Serum MMP-9 and MMP-8 share common pathways in the pathophysiology of ACS. We have published earlier the re- covery were at increased risk for MACE, especially for non-fatal events. As novel findings, the largest decrease of sults on serum MMP-8 and MMP-8/TIMP-1 in the present MMP-9 concentrations between acute and recovery phase population, which offers a chance to compare these two bio- (Δ values) was protective against MACE, especially markers. The MMP-9 concentrations and MMP-9/TIMP-1 non-fatal events. When examining the molecular forms of molar ratio differed between cases and controls similar to MMP-9, the cases with a fatal outcome had during the acute MMP-8 and MMP-8/TIMP-1 and had strong associations phase the highest activatable MMP-9 values. with ACS . Both MMP-9 and MMP-8 decreased in a The association of MMP-9 with different cardiac outcomes similar manner in the course of recovery . However, the has been previously reported; plasma MMP-9 is associated prognostic value of MMP-9 and MMP-8 differed from each with CVD mortality in patients with CAD at baseline [16, other markedly: the acute phase MMP-9 was a predictor of 23]. Our results are in agreement with these findings; acute cardiovascular death and the recovery MMP-9 of hospitaliza- phase MMP-9 concentration predicted especially CVD death tion for an ACS event, while MMP-8 had no prognostic value in the 6-year follow-up. Earlier studies have also suggested that for these end points in the present population . elevated MMP-9 is associated with higher risk of death due to Upregulated TIMP-1 may suppress MMP-8 and MMP-9 any cause  and with CVD risk factors and total cardiovas- , thus providing an important regulatory step in physio- cular risk in subjects without symptoms of CAD . In our logical circumstances. In our previous study, acute and recov- study, the recovery phase MMP-9 concentrations and MMP-9/ ery phase TIMP-1 concentrations were associated with cardio- TIMP-1 ratio were better predictors of future outcome than the vascular death with hazard ratios of 4.31 (p < 0.001) and 4.69 levels measured at the baseline. Importantly, the more these (p = 0.037), respectively . In the present study, MMP-9/ levels decreased after the acute phase, the better was the pre- TIMP-1 molar ratio determined in the recovery phase predict- dictable outcome, especially regarding non-fatal events. ed MACE. Therefore, the balance between MMPs and TIMPs Similar approach was used in a 30-day follow-up of stroke may be crucial for the destabilization of atherosclerotic patients , where the serum MMP-9 concentration in the plaques in acute phase , since MMP inhibitors have been J. of Cardiovasc. Trans. Res. (2018) 11:210–220 217 Fig. 3 Active and APMA-activatable MMP-9 analyzed by gelatin the subjects selected for zymography. All measurement points are zymography presented as arbitrary units of intensities (Y1-axis) and presented as dots, and group medians are presented with a line. corresponding MMP-9 concentrations measured by ELISA (Y2-axis) in Statistically significant differences are presented below the plot shown to reduce myocardial infarct size . Synthetic MMP active MMP-9 analyzed by gelatin zymography did not differ inhibitors, such as ilomastat, could be cardioprotective, espe- significantly between cases and controls in the acute phase. cially in preventing reperfusion injury . However, there On the other hand, in vitro conversion was observed in all are challenges in optimizing the pharmacological targeting, samples when serum MMP-9 was processed chemically with since MMP-9 is required also in the healing processes . APMA. In gelatin zymography, the functional activity of Investigation of MMP-9 activation and activation potential MMP-9 under physiological conditions cannot be thoroughly may offer new insights in MMP-9 diagnostics and prognos- addressed, as it is not clear whether SDS disrupts the normal tics, even if challenges remain. Gelatin zymography is a wide- physiological interactions between MMP-9 and TIMP-1 . ly used technique for detecting active MMP-9 based on mo- MMP-2 is other gelatinolytic MMP, which is related to lecular weight . MMP-9 is expressed as a 92 kDa inflammation in myocardium . MMP-2 is usually activat- pro-form, whose molecular weight upon activation usually ed with proteolysis and molecular weight drops from 72 to decreases by 10 kDa . High molecular size (> 100 kDa) 64 kDa . In our small zymography subpopulation, we gelatinolytic species represent MMP-9 clustered together with noticed differences in the full size MMP-2 between cases the neutrophil gelatinase associated lipocalin (NGAL)  and controls, but proteolytically activated MMP-2 band was and MMP-9 linked with TIMP-1 . Patients with cardio- not present. Oxidative stress can also activate MMP-2 and vascular death had significantly increased acute phase MMP-9, and in a pilot study, serum nitrotyrosine, a marker activatable MMP-9 values when compared to controls. describing oxidative stress, correlated with active MMP-9, but Interestingly, the activatable MMP-9 decreased significantly not with active MMP-2 . The present study confirms and between the acute and recovery phase in cases suffering a further extends earlier findings regarding the importance of non-fatal MACE. Thus, the proportion of molecular forms MMP-9 in CAD. and Bactivation potential^ may be crucial for prognostics. The diagnostic efficacy and accuracy of serum MMP-9 needs MMP-9 can be activated by several proteolytic and to be further explored. In our study, diabetes did not associate non-proteolytic ways, and the alternative MMP-9 activation with serum MMP-9 levels. While this observation differs from does not necessarily involve the molecular weight change [7, earlier findings , the small number of diabetic subjects may have affected these results. Also, other systemic conditions such 40]. This may at least in part explain why the amounts of 218 J. of Cardiovasc. Trans. Res. (2018) 11:210–220 (LL), the Lund University (EP), the Lund University Hospital (EP), the as obesity  and metabolic syndrome  elevate serum Academy of Finland (grant 126650 to PJP), the Sigrid Juselius foundation MMP-9 concentrations. Smoking is a strong confounding factor (PJP), the Yrjö Jahnsson foundation (PJP), the Aulikki and Sakari in MMP-9 diagnostics; similar trend using serum MMP-8 in Sohlberg foundation (PJP), and the Helsinki University Hospital ACS-diagnostics was previously documented . Research Foundation (Grants TYH 2016251, TYH 2017251, Y1149SUL32) (TS). There are several limiting factors in our study. We cannot estimate the proportion of MMP-9 excreted from ischemic Compliance with Ethical Standards myocardium or ruptured atherosclerotic plaques from serum analysis. In addition, we do not have information on the in- Conflict of Interest The authors declare that they have no conflict of farct size; thus, the models cannot be adjusted by it. interest. Nevertheless, the hazard ratio for a non-fatal MACE measured from recovery phase samples strongly supports the hypothesis Ethical Approval The study was conducted according to the Declaration of Helsinki. The ethical committee of the Lund University approved the that the elevation of MMP-9 in these patients was mainly due study design. No animal studies were carried out by the authors for this to the cardiac events. In zymography analysis, we cannot cal- article. culate the total physiological MMP-9 activity. The serum sam- ples were collected in 1999–2002, when the use of statins was Informed Consent All participants provided informed consent. not common, and we have information only on lipid-lowering medication. Statins may affect MMP-9 concentrations , Open Access This article is distributed under the terms of the Creative albeit in meta-analyses statins had no significant effect on Commons Attribution 4.0 International License (http:// plasma MMP-9 . Plasma TIMP-1 concentrations de- creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give creased after use of statins . We could not adjust the appropriate credit to the original author(s) and the source, provide a link models with BMI and blood pressure, as we did not have this to the Creative Commons license, and indicate if changes were made. information on our control population. However, according to our inclusion criteria, the controls were not using anti-hypertensive medication. Novel biomarkers are needed to identify CVD risk patients, References early diagnostics of atherosclerotic plaque rupture or ische- mia, and CVD prognostics. Serum MMP-9 determination is 1. Ross, R. (1999). Atherosclerosis—an inflammatory disease. The sensitive and specific, thus being a good candidate as a bio- New England Journal of Medicine, 340(2), 115–126. marker in clinical practice. MMP-9 plays a role both in ath- 2. Libby, P., Ridker, P. M., Hansson, G. K., & Leducq Transatlantic Network on Atherothrombosis. (2009). Inflammation in atheroscle- erosclerotic plaque rupture and tissue destruction after a car- rosis: from pathophysiology to practice. Journal of the American diac event. The balance between MMP-9 and TIMP-1 may be College of Cardiology, 54(23), 2129–2138. crucial for disease progression. The difference of MMP-9 and 3. Packard, R. R., & Libby, P. (2008). Inflammation in atherosclerosis: MMP-9/TIMP-1 between acute and recovery phase provide from vascular biology to biomarker discovery and risk prediction. novel prognostic information of secondary cardiac events. Clinical Chemistry, 54(1), 24–38. 4. Tayebjee, M. H., Lip, G. Y., & MacFadyen, R. J. (2005). Matrix Additionally, the analysis of MMP-9 activation potential metalloproteinases in coronary artery disease: clinical and therapeu- may offer new insights into cardiac diagnostics and tic implications and pathological significance. Current Medicinal prognostics. Chemistry, 12(8), 917–925. 5. Van Doren, S. R. (2015). Matrix metalloproteinase interactions with collagen and elastin. Matrix Biology, 44-46,224–231. Clinical Relevance 6. Arpino, V., Brock, M., & Gill, S. E. (2015). The role of TIMPs in regulation of extracellular matrix proteolysis. Matrix Biology, 44- MMP-9 can be utilized as an early stage biomarker, because 46,247–254. 7. Bannikov, G. A., Karelina, T. V., Collier, I. E., Marmer, B. L., & its elevation reflects atherosclerotic plaque rupture and myo- Goldberg, G. I. (2002). Substrate binding of gelatinase B induces its cardial tissue destruction. Furthermore, MMP-9 has prognos- enzymatic activity in the presence of intact propeptide. The Journal tic value, which is important in secondary prevention and of Biological Chemistry, 277(18), 16022–16027. planning of personalized treatment. 8. Brew, K., Dinakarpandian, D., & Nagase, H. (2000). Tissue inhib- itors of metalloproteinases: evolution, structure and function. Acknowledgements The authors wish to thank Kati Hyvärinen, Ph.D. for Biochimica et Biophysica Acta, 1477(1–2), 267–283. her help in gelatin zymography and Eva Andsberg, BM, and research 9. Yabluchanskiy, A., Ma, Y., Iyer, R. P., Hall, M. E., & Lindsey, M. L. nurses Laura Darcy and Annica Maxedius for their excellent job in (2013). Matrix metalloproteinase-9: many shades of function in interviewing the subjects, collecting blood samples, and collecting and cardiovascular disease. Physiology (Bethesda), 28(6), 391–403. registering data. 10. Grierson, C., Miller, D., LaPan, P., & Brady, J. (2010). Utility of combining MMP-9 enzyme-linked immunosorbent assay and Funding Information The study was supported by grants from the MMP-9 activity assay data to monitor plasma enzyme specific ac- tivity. Analytical Biochemistry, 404(2), 232–234. Finnish Dental Society Apollonia (LL), the Orion Research Foundation J. of Cardiovasc. Trans. Res. (2018) 11:210–220 219 11. Kai, H., Ikeda, H., Yasukawa, H., Kai, M., Seki, Y., Kuwahara, F., 25. Garvin, P., Nilsson, L., Carstensen, J., Jonasson, L., & Kristenson, Ueno, T., Sugi, K., & Imaizumi, T. (1998). Peripheral blood levels M. (2008). Circulating matrix metalloproteinase-9 is associated of matrix metalloproteases-2 and -9 are elevated in patients with with cardiovascular risk factors in a middle-aged normal popula- acute coronary syndromes. Journal of the American College of tion. PLoS One, 3(3), e1774. Cardiology, 32(2), 368–372. 26. Abdelnaseer, M. M., Nervana, M. E., Esmail, E. H., Kamal, M. M., 12. Inokubo, Y., Hanada, H., Ishizaka, H., Fukushi, T., Kamada, T., & & Elsawy, E. H. (2017). Matrix metalloproteinase-9 and recovery Okumura, K. (2001). Plasma levels of matrix metalloproteinase-9 of acute ischemic stroke. Journal of Stroke and Cerebrovascular and tissue inhibitor of metalloproteinase-1 are increased in the cor- Diseases, 26,733–740. onary circulation in patients with acute coronary syndrome. 27. Kobayashi, N., Hata, N., Kume, N., Yokoyama, S., Shinada, T., American Heart Journal, 141(2), 211–217. Tomita, K., Kitamura, M., Shirakabe, A., Inami, T., Yamamoto, 13. Derosa, G., D’Angelo, A., Scalise, F., Avanzini, M.A., Tinelli, C., M., et al. (2011). Matrix metalloproteinase-9 for the earliest stage Peros, E., Fogari, E., & Cicero, A.F. (2007). Comparison between acute coronary syndrome. Circulation Journal, 75(12), 2853–2861. metalloproteinases-2 and -9 in healthy subjects, diabetics, and sub- 28. Guzel, S., Serin, O., Guzel, E. C., Buyuk, B., Yilmaz, G., & jects with acute coronary syndrome. Heart and Vessels, 22(6), 361– Güvenen, G. (2013). Interleukin-33, matrixmetalloproteinase-9, and tissue inhibitor of matrix metalloproteinase-1 in myocardial 14. Fukuda, D., Shimada, K., Tanaka, A., Kusuyama, T., Yamashita, infarction. The Korean Journal of Internal Medicine, 28(2), 165– H., Ehara, S., Nakamura, Y., Kawarabayashi, T., Iida, H., 173. Yoshiyama, M., et al. (2006). Comparison of levels of serum matrix 29. Peterson, J. T., Li, H., Dillon, L., & Bryant, J. W. (2000). Evolution metalloproteinase-9 in patients with acute myocardial infarction of matrix metalloprotease and tissue inhibitor expression during versus unstable angina pectoris versus stable angina pectoris. The heart failure progression in the infarcted rat. Cardiovascular American Journal of Cardiology, 97(2), 175–180. Research, 46(2), 307–315. 15. Tan, J., Hua, Q., Gao, J., & Fan Z. X. (2008). Clinical implications 30. Cimmino, G., Ragni, M., Cirillo, P., Petrillo, G., Loffredo, F., of elevated serum interleukin-6, soluble CD40 ligand, metallopro- Chiariello, M., Gresele, P., Falcinelli, E., & Golino, P. (2013). C- teinase-9, and tissue inhibitor of metalloproteinase-1 in patients reactive protein induces expression of matrix metalloproteinase-9: a with acute ST-segment elevation myocardial infarction. Clinical possible link between inflammation and plaque rupture. Cardiology, 31(9), 413–418. https://doi.org/10.1002/clc.20254. International Journal of Cardiology, 168(2), 981–986. 16. Blankenberg, S., Rupprecht, H. J., Poirier, O., Bickel, C., Smieja, 31. Opstad, T. B., Pettersen, A. A., Weiss, T. W., Akra, S., Øvstebø, R., M., et al. (2003). Plasma concentrations and genetic variation of Arnesen, H., & Seljeflot, I. (2012). Genetic variation, gene- matrix metalloproteinase 9 and prognosis of patients with cardio- expression and circulating levels of matrix metalloproteinase-9 in vascular disease. Circulation, 107(12), 1579–1585. patients with stable coronary artery disease. Clinica Chimica Acta, 17. Pesonen, E., El-Segaier, M., Persson, K., Puolakkainen, M., Sarna, 413(1–2), 113–120. S., Ohlin, H., & Pussinen, P. J. (2009). Infections as a stimulus for 32. Li, J., Lu, H., Tao, F., Zhou, H., Feng, G., He, L., & Zhou, L. coronary occlusion, obstruction, or acute coronary syndromes. (2013). Meta-analysis of MMP9-562C/T and the risk of coronary Therapeutic Advances in Cardiovascular Disease, 3(6), 447–454. heart disease. Cardiology, 124(1), 53–59. 18. Pussinen, P. J., Sarna, S., Puolakkainen, M., Öhlin, H., Sorsa, T., & 33. Vilmi-Kerälä, T., Lauhio, A., Tervahartiala, T., Palomäki, O., Pesonen, E. (2013). The balance of serum matrix Uotila, J., Sorsa, T., & Palomäki, A. (2017). Subclinical inflamma- metalloproteinase-8 and its tissue inhibitor in acute coronary syn- tion associated with prolonged TIMP-1 upregulation and arterial drome and its recurrence. International Journal of Cardiology, stiffness after gestational diabetes mellitus: a hospital-based cohort 167(2), 362–368. study. Cardiovascular Diabetology, 16(1), 49. 19. Lahdentausta, L., Sorsa, T., Pussinen, P. J., & Pesonen, E. (2013). 34. Hausenloy, D. J., Garcia-Dorado, D., Bøtker, H. E., Davidson, S. The effect of smoking on diagnostic value of serum matrix M., Downey, J., Engel, F. B., Jennings, R., et al. (2017). Novel metalloproteinase-8 in acute coronary syndrome. Journal of targets and future strategies for acute cardioprotection: Position Molecular Biomarkers and Diagnosis, S4,002. https://doi.org/10. paper of the European Society of Cardiology Working Group on 4172/2155-9929.S4-002 Cellular Biology of the Heart. Cardiovascular Research, 113(6), 20. Alfakry, H. (2014). Immune and proteolytic events associated with 564–585. the signs of periodontal and cardiovascular diseases and their 35. Bencsik, P., Pálóczi, J., Kocsis, G. F., Pipis, J., Belecz, I., Varga, Z. treatment. Doctoral dissertation, University of Helsinki. ISBN: V., Csonka, C., et al. (2014). Moderate inhibition of myocardial 978-952-10-9985-4. matrix metalloproteinase-2 by ilomastat is cardioprotective. 21. Nagase, H., & Brew, K. (2003). Designing TIMP (tissue inhibitor of Pharmacological Research, 80,36–42. metalloproteinases) variants that are selective metalloproteinase in- 36. Barkho, B. Z., Munoz, A. E., Li, X., et al. (2008). Endogenous hibitors. Biochemical Society Symposium, 70,201–212. matrix metalloproteinase (MMP)-3 and MMP-9 promote the differ- entiation and migration of adult neural progenitor cells in response 22. Sorsa, T., Salo, T., Koivunen, E., Tyynelä, J., Konttinen, Y. T., to chemokines. Stem Cells, 26,3139–3149. Bergmann, U., Tuuttila, A., Niemi, E., Teronen, O., Heikkilä, P., Tschesche, H., Leinonen, J., Osman, S., & Stenman, U. H. (1997). 37. Kupai, K., et al. (2010). Matrix metalloproteinase activity assays: Activation of type IV procollagenases by human tumor-associated importance of zymography. Journal of Pharmacological and trypsin-2. The Journal of Biological Chemistry, 272(34), 21067– Toxicological Methods, 61(2), 205–209. 38. Kjeldsen, L., Johnsen, A. H., Sengeløv, H., & Borregaard, N. 23. Kelly, D., Khan, S. Q., Thompson, M., Cockerill, G., Ng, L. L., (1993). Isolation and primary structure of NGAL, a novel protein Samani, N., & Squire, I. B. (2008). Plasma tissue inhibitor of associated with human neutrophil gelatinase. The Journal of metalloproteinase-1 and matrix metalloproteinase-9: novel indica- Biological Chemistry, 268(14), 10425–10432. tors of left ventricular remodelling and prognosis after acute myo- 39. Roy, R., Louis, G., Loughlin, K. R., Wiederschain, D., Kilroy, S. cardial infarction. European Heart Journal, 29(17), 2116–2124. M., Lamb, C. C., Zurakowski, D., & Moses, M. A. (2008). Tumor- 24. Hansson, G. (2005). Inflammation, atherosclerosis, and coronary specific urinary matrix metalloproteinase fingerprinting: identifica- artery disease. The New England Journal of Medicine, 352(16), tion of high molecular weight urinary matrix metalloproteinase spe- cies. Clinical Cancer Research, 14(20), 6610–6617. 1685–1695. 220 J. of Cardiovasc. Trans. Res. (2018) 11:210–220 40. Van Wart, H. E., & Birkedal-Hansen, H. (1990). The cysteine 45. Marx,N.,Froehlich,J.,Siam, L.,Ittner,J.,Wierse,G., Schmidt,A., Scharnagl, H., Hombach, V., & Koenig, W. (2003). Antidiabetic switch: a principle of regulation of metalloproteinase activity with potential applicability to the entire matrix metalloproteinase gene PPAR gamma-activator rosiglitazone reduces MMP-9 serum levels family. Proceedings of the National Academy of Sciences of the in type 2 diabetic patients with coronary artery disease. United States of America, 87(14), 5578–5582. Arteriosclerosis, Thrombosis, and Vascular Biology, 23(2), 283–288. 41. Ikeda, M., Maekawa, R., Tanaka, H., Matsumoto, M., Takeda, Y., 46. Andrade, V. L., Petruceli, E., Belo, V. A., Andrade-Fernandes, C. Tamura, Y., Nemori, R., & Yoshioka, T. (2000). Inhibition of M., Caetano Russi, C. V., Bosco, A. A., Tanus-Santos, J. E., & gelatinolytic activity in tumor tissues by synthetic matrix metallo- Sandrim, V. C. (2012). Evaluation of plasmatic MMP-8, MMP-9, proteinase inhibitor: application of film in situ zymography. TIMP-1 and MPO levels in obese and lean women. Clinical Clinical Cancer Research, 6(8), 3290–3296. Biochemistry, 45(6), 412–415. 42. DeCoux, A.,Lindsey,M. L.,Villarreal,F.,Garcia,R. A.,&Schulz,R. 47. Hopps, E., Lo Presti, R., Montana, M., Noto, D., Averna, M. R., & (2014). Myocardial matrix metalloproteinase-2: inside out and upside Caimi, G. (2013). Gelatinases and their tissue inhibitors in a group down. Journal of Molecular and Cellular Cardiology, 77,64–72. of subjects with metabolic syndrome. Journal of Investigative 43. Jacob-Ferreira, A. L., Kondo, M. Y., Baral, P. K., James, M. N., Medicine, 61(6), 978–983. Holt, A., Fan, X., & Schulz, R. (2013). Phosphorylation status of 48. Andrade,V.L., do Valle, I.B.,&Sandrim, V. C.(2013). 72 kDa MMP-2 determines its structure and activity in response to Simvastatin therapy decreases MMP-9 levels in obese women. peroxynitrite. PLoS One, 8(8), e71794. Journal of Clinical Pharmacology, 53(10), 1072–1077. 44. Bencsik,P.,Sasi,V.,Kiss,K.,Kupai,K.,Kolossváry,M., 49. Ferretti, G., Bacchetti, T., Banach, M., Simental-Mendía, L. E., & Maurovich-Horvat, P., Csont, T., Ungi, I., Merkely, B., & Sahebkar, A. (2016). Impact of statin therapy on plasma MMP-3, Ferdinandy, P. (2015). Serum lipids and cardiac function correlate MMP-9, and TIMP-1 concentrations. Angiology, 68,850–862. with nitrotyrosine and MMP activity in coronary artery disease pa- tients. European Journal of Clinical Investigation, 45(7), 692–701.
Journal of Cardiovascular Translational Research – Springer Journals
Published: Jan 18, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera