Journal of Critical Care (2011) 26, 550–555

Dynamic changes of matrix metalloproteinases and theirtissue inhibitors in severe sepsis☆

Diána Mühl PhD, MDa,⁎,1, Bálint Nagy MDa,1, Gábor Woth MDa, Boglárka Falusi MDa,Lajos Bogár PhD, MDa, György Weber PhD, MDb, János Lantos PhD, MDb

aDepartment of Anesthesiology and Intensive Therapy, University of Pécs, Pécs, HungarybDepartment of Surgical Research and Techniques, University of Pécs, Pécs, Hungary

0d

Keywords:Matrix metalloproteinases;Tissue inhibitors ofmetalloproteinases;

Severe sepsis

AbstractPurpose: Little is known about the dynamic changes of matrix metalloproteinases (MMPs) and theirinhibitors (TIMPs) in sepsis. Our aim was therefore to investigate the time course of MMPs and theirinhibitors in patients experiencing severe sepsis.Methods: Our prospective controlled analysis included 38 patients with severe sepsis. Plasma levels ofMMP-2, MMP-9, TIMP-1, and TIMP-2 were measured daily at a 5-day-long period with enzyme-linkedimmunosorbent assay. Seventeen healthy volunteers were invited as controls.Results: MMP-2 showed no difference compared to controls, whereas significantly elevated MMP-9levels were detected on admission (P b .005). Significantly elevated but declining TIMP-1 levels weremeasured during the whole trial (P b .002-.004). Except for the second day, TIMP-2 levels weresignificantly lower than controls (P b .05-.009). MMP2/TIMP-1 ratios were significantly lower in septicpatients (P b .03-.006), whereas MMP-2/TIMP-2 ratios were elevated throughout our study (P b .03-.006). MMP-9/TIMP-1 ratios were significantly lower at the first 3 days (P b .05-.008). MMP-9/TIMP-2was significantly elevated on admission (P b .006).Conclusions: Our research is the first follow-up study dealing with MMPs, TIMPs, and their ratios insevere sepsis. Our results indicate that MMPs and TIMPs may play a crucial role in severe sepsis,especially TIMP-1, MMP-9, and possibly TIMP-2, after an extensive study.© 2011 Elsevier Inc. All rights reserved.

1. Introduction The use of biomarkers may improve early diagnosis and

Severe sepsis and septic shock are major causes ofmortalityin intensive care units worldwide. As the incidence of sepsisrises with soaring medical expenses, mortality has furtherrisen to 80% with the complication of 4 organ failures [1-3].

☆ The authors declare no conflicts of interest.⁎ Corresponding author. Tel.: + 36 72 536 440; fax: + 36 72 536 441.E-mail address: drdianamuhl@gmail.com (D. Mühl).1 These authors contributed equally to this work.

883-9441/$ – see front matter © 2011 Elsevier Inc. All rights reserved.oi:10.1016/j.jcrc.2011.02.011

therapy in critically ill patients and may consecutivelyinfluence the morbidity and mortality of these patients.Metalloproteinases and their inhibitors may represent apromising new class of biomarkers for the prognosis ofsevere sepsis patients [4]. Matrix metalloproteinases(MMPs) are a family of zinc-containing metalloendopepti-dases which are classified by their substrates into gelatinases(MMP-2 and MMP-9), collagenases (MMP-1, MMP-8, andMMP-13), stromelysins (MMP-3, -10, and -11), elastases(MMP-7 and -12), and membrane-type MMPs [5-7]. Amurine aorta aneurysm model showed that MMP-9 is mostly

Table 1 Demographic and characteristic data of septic patients

Septic patients

Number of patients 38

551MMPs and their tissue inhibitors in severe sepsis

macrophage-derived in inflammation while majority ofMMP-2 is of mesenchymal origin [8]. Tissue inhibitors ofmatrix metalloproteinases (TIMPs) are a group of proteins(TIMP-1, TIMP-2) responsible for the regulation of theactivity of MMPs. Recently published data suggests thatbesides their already described role in extracellular matrixprotein degradation and tissue remodelling, MMPs couldbe considered signalig proteases. MMPs have a role inintracellular protein cleavage, atherosclerosis and tumorgrowth. Vascular hyporeactivity to vasoconstrictors, a signof sepsis has been also addressed to MMPs in a rat in vivomodel [9,10]. The prognostic importance of these enzymeshas been shown in several diseases, such as atherosclero-sis, lymphomas, and in skin fibrosis of severely burntpatients [11,12]. Under the condition of infection, MMPspromote the extravasation and migration of leukocytes byproteolysis of basement membrane and by modulating theinflammatory response [13-15]. The MMP-2 is implicatedin various aspects of inflammation including homing ofinflammatory cells, tissue injury healing, and remodelling[16]. MMP-9 is secreted by human neutrophils in responseto endotoxin, tumor necrosis factor α, and IL-8 [17].TIMP-1 can activate normal human granulocyte andprotect them from apoptosis during inflammation [18].Several studies have investigated the levels of MMPs andTIMPs and their prognostic value of mortality in severesepsis [4,7,16,19,20], but the alteration of these proteinsand their inhibitors during the therapy of severe sepsisremains unclear.

The aim of our study was to investigate the time course ofmatrix metalloproteinases and their inhibitors, as well as theratio of MMPs and TIMPs in septic patients.

Age (y) 59 ± 13Sex (male/female) 24:14Survivor/nonsurvivor During study period: 30:8

Signs of sepsisNumber of organ failures 3 ± 1SAPS II 38 ± 9SOFA1st day 9.2 ± 3.23rd day 8.2 ± 3.75th day 6.5 ± 2.8MODS1st day 7.2 ± 3.73rd day 6.5 ± 3.95th day 4.8 ± 3.5Conscious disorder 16Vasopressor and/or inotropes 33ARDS 11Renal failure 27Hepatic failure 14CRP (1st day) 225 ± 111PCT (1st day) 18 ± 16Lactate (1st day) 1.9 ± 1

Data are presented as mean ± standard error of mean. ARDS indicatesacute respiratory distress syndrome; CRP, C-reactive protein; PCT,procalcitonin.

2. Methods

2.1. Patient selection and data collection

The study protocol fulfilled the ethical guidelines of the2003 Declaration of Helsinki, and permission was obtainedfrom the Institutional Scientific and Human Research EthicsCommittee of the University of Pécs (2406/2005).

Patients (n = 38) treated for severe sepsis were included inour study. After being informed clearly about the details ofthe study and blood sampling, patients or next of kinprovided a written informed consent. The treatment of organfailures, volume resuscitation, and supportive therapy forsepsis were based on currently applied guidelines [3]. Fluidresuscitation with crystalloids and colloids, early broadspectrum antibiotic treatment (most commonly carbapenemswith or without aminoglycosides), norepinephrine with orwithout dobutamine in case of hemodynamic shock, forglucose level control infusion of actrapid, in oliguria oranuria furosemide with hemodialysis, stress ulcer prophy-laxis with pantoprazole or famotidine, thrombosis prophy-

laxis by enoxaparine or fraxiparine were administered.Morphine and propofol were used for analgesia and sedationduring mechanical ventilation at appropriate doses. None ofthe patients received drotrecogin alfa and antiplatelet drugs.Blood samples were collected daily via a radial arterycannula for the first 5 days.

Seventeen patients scheduled for outpatient ophthalmol-ogy examination were invited as age-adopted healthycontrols. Age and sex parameters showed no significantdifference compared to the severe septic group. Bloodsampling was carried out once from every volunteer.

The diagnosis of severe sepsis was assessed by the NewSimplified Acute Physiology Score (SAPS II), SequentialOrgan Failure Assessment (SOFA) score, and the MultipleOrgan Dysfunction Score (MODS), and the AmericanCollege of Chest Physicians/Society of Critical CareMedicine consensus was considered as well. Our inclusioncriteria were defined by 2 or more organ dysfunctions,hypoperfusion abnormality, or sepsis-induced hypotensionand a higher than 5 ng/mL serum procalcitonin level [21].

2.2. MMP-TIMP assays

Blood samples were collected on intensive care unitadmission (day 1) and on the 4 consecutive mornings (days

552 D. Mühl et al.

2 to 5) thereafter. Plasma was isolated from heparinanticoagulated blood samples by low-speed centrifugationat 4 °C and stored at −80 °C until analyzed in a single batchat the end of the study. The matrix metalloproteinases(MMP-2, MMP-9) and tissue inhibitors of metalloprotei-nases (TIMP-1, TIMP-2) were determined by the quantita-tive sandwich enzyme-linked immunosorbent assaytechniques according to the manufacturer's instructions(R&D Systems, Inc, Minneapolis, Minn). In comparisonwith standard MMP and TIMP curves, the concentrations ofMMPs and TIMPs in plasma were determined spectropho-tometrically (Multiskan Ascent microplate photometer,Type 354, Thermo Electron Corporation, Waltham, MAUSA) by reading the absorbance at 450 nm and wereexpressed as entire amounts in the plasma (ng/mL).

2.3. Statistical analysis

Data were expressed as minimum, maximum, median,and interquartile range (standard 25th-75th percentile). Incase of normal distribution, Student's t test was applied;otherwise, Wilcoxon signed rank test or Mann-WhitneyU test was used for the analysis of nonparametric data.Values of P b .05 were considered significant.

Fig. 1 Plasma levels of MMP-2, MMP-9, TIMP-1, and TIMP-2 on admData are expressed as minimum, maximum, median, and interquartile ranhealthy control results. Asterisks indicate statistical differences (P b .05)significant difference (P b .05) in severe septic patients compared to con

3. Results

The main demographical data, MODS, and SOFA scorescalculated on the first, third, and fifth admission day and themost commonly occurring organ failures of our studypopulation are shown in Table 1.

MMP-2 levels were elevated although not significantly bydays 2 and 3, but by day 4 they decreased to the controllevels (Fig. 1A). The plasma MMP-9 levels were signifi-cantly higher (P b .005) on admission and reducedsignificantly by day 3 (P b .009). At the consecutive days,the MMP-9 remained at this nonsignificantly elevated level(Fig. 1B). The plasma level of TIMP-1 (Fig. 1C) wassignificantly elevated (P b .002-.004) during the whole studyperiod compared to controls. It decreased significantly fromday 3 (P b .006). Except for day 2, the plasma level of TIMP-2 was significantly (P b .05-.009) lower in septic patients(Fig. 1D). The MMP-2/TIMP-1 ratio (Fig. 2A) wassignificantly decreased during the whole observation periodcompared to controls (P b .004-.002) and elevatedsignificantly from day 3 (P b .001). MMP-2/TIMP-2 ratiowas significantly elevated during all 5 days in septic patients(P b .03-.006) (Fig. 2B). MMP-9/TIMP-1 ratio wassignificantly lower in septic patients compared to healthy

ission (day 1) and on the 4 consecutive days thereafter (days 2-5).ge. Shaded boxes represent severe septic patients; white boxes showinside the septic group compared to day 1. The “#” symbols showtrols.

Fig. 2 MMP-2/TIMP-1, MMP-2/TIMP-2, MMP-9/TIMP-1, and MMP-9/TIMP-2 ratios in plasma samples on admission (day 1) and on the4 consecutive days thereafter (days 2–5). Data are expressed as minimum, maximum, median, and interquartile range. Shaded boxes representsevere septic patients; white boxes show healthy control results. Asterisks indicate statistical differences (P b .05) inside the septic groupcompared to day 1. The “#” symbols show significant difference (P b .05) in severe septic patients compared to controls.

553MMPs and their tissue inhibitors in severe sepsis

controls on the first 3 days (P b .05-.008) (Fig. 2C). MMP-9/TIMP-2 ratio was significantly (P b .006) elevated onadmission in septic patients and remained above the normalvalues during the 5-day-long investigation (Fig. 2D).

4. Discussion

MMPs and TIMPs serve as potentially new biomarkersfor the prognosis of sepsis. The novelty of our study is thedetermination of the time course of circulating MMPs andTIMPs in severe sepsis with serial clinical measurements.The lack of these serial measurements in the work of Lorenteet al is mentioned by Hoffman et al as a limitation of formertrials [4,7]. In accordance with previous studies [16,18,19],we found significantly elevated MMP-9 levels on admissionin septic patients. Lalu et al. found a significant increase ofcirculating MMP-9 and negative correlation with bloodpressure in a rat model. The MMP-9 elevation couldcontribute to the proteolytic state and may become a usefulmarker of the severity of sepsis [22]. The only multicentralanalysis by Lorente et al [7] could not support our findings;however, they observed a nonsignificantly higher MMP-9. Inour investigation, MMP-9 levels decreased on the third day.Lorente et al [7] found significantly lower MMP-9 levels innonsurvivors, so the decrease we observed could be due to

progression of sepsis. Albert et al [23] showed that MMP-9,not MMP-2, is released after lipopolysaccharide administra-tion. While MMP-9 has a pivotal role in the mobilization ofendothelial progenitor cells from the bone marrow [24],Rafat et al. described the increase of circulating endothelialprogenitor cells in peripheral blood and their correlation withimproved sepsis survival [25]. Therefore the increase ofMMP-9 may have beneficial effects during the course ofsepsis. Although the increase in MMP-2 and MMP-9 wasnoticeable during the first days of our study, Fortuna et al.realized that MMP-9 has a role in acute pulmonary embolisminduced pulmonary hypertension which doxycycline couldprevent, thus MMP-9 inhibition may help in hemodynamicderangements. [26]. In accordance with Albert et al [23] andHoffmann et al [16], we could not prove any remarkablechanges in the levels of MMP-2 in septic patients. Cena et al.found the protective effect of MMP-2 inhibition on a rat aortaring contractility model in the 6th study hour [27]. Lalu et al.presented a significant MMP-2 activity and content decreaseafter 6 hours of LPS administration but a steady increase ofboth parameters until the 24th hour in a septic ratmyocardium model [22]. The MMP-2 decrease may have acardioprotective effect [27]. We can not support theobservation of plasma MMP-2 decrease during the firstday, as infection may already existent for more than 24 hoursbefore inclusion. TIMP-1 is an inducible protein and plays a

554 D. Mühl et al.

vital role in severe sepsis as reflected by several studieswhich have shown higher TIMP-1 levels in non-survivors[7,16,18,19]. Because of the small number of nonsurvivors,the statistical power was questionable and therefore we didnot analyze a statistical difference between survivors andnonsurvivors. In our study, TIMP-1 levels were significantlyelevated on admission and started to decrease significantly inseptic patients from the third day onward. In contrast toTIMP-1, TIMP-2 is a constitutive protein. Although theabsolute values of TIMP-2 were higher in our measurementsthan in the Hoffmann et al [16] study, we found significantlylower TIMP-2 levels in septic patients compared to controls.Whereas Hoffmann et al [16] used citrated blood samples,we used heparin as an anticoagulant. Although Gerlach RF,et al. found similar MMP activity patterns in plasma samplesusing EDTA, citrate or heparin as anticoagulant, citrateplasma presented lower MMP activities [28]. MMP-TIMPratios may mirror the balance between the proteolytic activityof MMPs and their tissue inhibitors.

Lorente et al [7] suggested the importance of reducedMMP-9/TIMP-1 ratio as a predictive biomarker of severity inseptic patients. Our findings firmly support their opinion aswe also measured significantly reduced MMP-9/TIMP-1ratios. The initially reduced MMP-9/TIMP-1 ratios did notchange significantly until day 5, suggesting a balancebetween the enzyme and the inhibitor despite dynamic andsignificant changes between them. The decreased MMP-9/TIMP-1 levels upon admission could predict the severity andmortality of sepsis. The usefulness of the ratio of MMP-2/TIMP-1 and MMP-2/TIMP-2 is doubtful in severe sepsisbecause MMP-2 is hardly changed in 5 consecutive days.Therefore the significant changes seen in the ratio werecaused only by the alterations in the inhibitor levels (TIMP-1and TIMP-2). Multiple studies discussed the MMP inhibitoryeffect of various antioxidants and medications [29,30]. Asrecent sepsis guidelines improved patient survival signifi-cantly we strictly followed these recommendations duringpatient care. Our opinion is that measurements in studysettings following the same guidelines are comparable withour results.

5. Conclusion

Sepsis remains a major cause for concern in intensivecare. Besides other biomarkers, the exact role played by thediverse MMP-TIMP system in sepsis is still unclear. Our aimwas to investigate the time course of circulating MMPs andTIMPs in severe septic patients. The dynamic changeshighlight the importance of MMPs and TIMPs. Our findingssupport the hypothesis that TIMP-1, MMP-9, and possiblyTIMP-2 may play a key role in inflammatory processes. Theresults with TIMP-1 and MMP-9/TIMP-1 ratio as predictivebiomarkers are also promising, but further trials on largersamples are required to clarify and confirm the outcome.

Considering recent evidence of the beneficial effect of MMP-9 in increasing circulating endothelial progenitor cells thepossible therapeutic use of MMP inhibitors like doxycyclineshould be considered based on a large double-blind study.Our investigation shows that the efficiency of the MMP-2/TIMP-1 and MMP-2/TIMP-2 ratios can be debated insevere sepsis.

Acknowledgment

The authors express their gratitude to Dr. Jose Tanus-Santos for giving inspiration to this work. They also wouldlike to thank the laboratory assistants and the nurses of theintensive care unit at the Department of Anaesthesia andIntensive Therapy for their enormous contribution.

References

[1] Martin GS, Mannino DM, Eaton S, et al. The epidemiology of sepsis inthe United States from 1979 through 2000. N Engl J Med 2003;348:1546-54.

[2] Linde-Zwirble WT, Angus DC. Severe sepsis epidemiology: sampling,selection, and society. Crit Care 2004;8:222-6.

[3] Dellinger RP. Cardiovascular management of septic shock. Crit CareMed 2003;31:946-55.

[4] Hoffmann U, BrueckmannM, Borggrefe M.Matrix metalloproteinasesand their inhibitors: promising novel biomarkers in severe sepsis? CritCare 2009;13:1006.

[5] Woessner JF. The matrix metalloproteinase family. In: Parks WC,Mecham RP, editors. MatrixMetalloproteinases. San Diego: AcademicPress; 1998. p. 1-14.

[6] Hijova E. Matrix metalloproteinases: their biological functions andclinical implications. Bratisl Lek Listy 2005;106(3):127-32.

[7] Lorente L,Martin MM, Labarta L, et al. Matrix metalloproteinase-9, -10,and tissue inhibitor of matrix metalloproteinases-1 blood levels asbiomarkers of severity and mortality in sepsis. Crit Care 2009;13:R158.

[8] Longo GM, Xiong W, Greiner TC, et al. Matrix metalloproteinases 2and 9 work in concert to produce aortic aneurysms. J Clin Invest2002;110(5):625-32.

[9] Brinckerhoff CE, Matrisian LM. Matrix metalloproteinases: a tail of afrog that became a prince. Nat Rev Mol Cell Biol 2002;3:207-14.

[10] Chow AK, Cena J, Schulz R. Acute actions and novel targets of matrixmetalloproteinases in the heart and vasculature. Br J Pharmacol2007;152(2):189-205.

[11] Ulrich D, Noah EM, von Heimburg D, et al. TIMP-1, MMP-2, MMP-9,and PIIINP as serum markers for skin fibrosis in patients followingsevere burn trauma. Plast Reconst Surg 2003;111(4):1423-31.

[12] Pennanen H, Kuittinen O, Turpeenniemi-Hujanen T. Plasma MMP-2-TIMP-2 complex levels measured during follow-up predict a risk ofrelapse in patients with malignant lymphoma. Eur J Haematol 2008;80(1):46-54.

[13] Galis ZS, Khatri JJ. Matrix metalloproteinases in vascular remodelingand atherogenesis the good, the bad, and the ugly. Circ Res 2002;90(3):251-62.

[14] Elkington PT, O'Kane CM, Friedland JS. The paradox of matrixmetalloproteinases in infectious disease. Clin Exp Immunol 2005;142:12-20.

[15] Opdenakker GP, Van den Steen HE, Van Damme J. Gelatinase B: atuner and amplifier of immune functions. Trends Immunol 2001;22:571-9.

555MMPs and their tissue inhibitors in severe sepsis

[16] Hoffmann U, Bertsch T, Dvortsak E, et al. Matrix-metalloproteinasesand their inhibitors are elevated in severe sepsis: prognostic value ofTIMP-1 in severe sepsis. Scand J Infect Dis 2006;38:867-72.

[17] Pugin J, Widmer MC, Kossodo S, et al. Human neutrophils secretegelatinase B in vitro and in vivo in response to endotoxin andproinflammatory mediators. Am J Respir Cell Mol Biol 1999;20:458-64.

[18] Chromek M, Tullus K, Lundahl J, et al. Tissue inhibitor ofmetalloproteinase 1 activates normal human granulocytes, protectsthem from apoptosis, and blocks their transmigration duringinflammation. Infect Immun 2004;72:82-8.

[19] Yassen KA, Galley HF, Webster NR. Matrix metalloproteinase-9concentrations in critically ill patients. Anaesthesia 2001;56:729-32.

[20] Nakamura T, Ebihara I, Shimada N, et al. Modulation of plasmametalloproteinase-9 concentrations and peripheral blood monocytemRNA levels in patients with septic shock: effect of fiber-immobilizedpolymyxin B treatment. Am J Med Sci 1998;316:355-60.

[21] Bone RC, Balk RA, Cerra FB, Dellinger RP, et al. Definitions forsepsis and organ failure and guidelines for the use of innovativetherapies in sepsis. The ACCP/SCCM Consensus ConferenceCommittee. American College of Chest Physicians/Society of CriticalCare Medicine. Chest 1992;101(6):1644-55.

[22] Lalu MM, Csont T, Schulz R. Matrix metalloproteinase activities arealtered in the heart and plasma during endotoxemia. Crit Care Med2004;32(6):1332-7.

[23] Albert J, Radomski A, Soop A, et al. Differential release of matrixmetalloproteinase-9 and nitric oxide following infusion of endo-

toxin to human volunteers. Acta Anaesthesiol Scand 2003;47:407-10.

[24] Hristov M, Erl W, Weber PC. Endothelial progenitor cells:mobilization, differentiation, and homing. Arterioscler Thromb VascBiol 2003;23(7):1185-9.

[25] Rafat N, Hanusch C, Brinkkoetter PT, et al. Increased circulatingendothelial progenitor cells in septic patients: correlation with survival.Crit Care Med 2007;35(7):1677-84 [Erratum in: 1994. JAMA 271(17):1321].

[26] Fortuna GM, Figueiredo-Lopes L, Dias-Junior CA, et al. A role formatrix metalloproteinase-9 in the hemodynamic changes followingacute pulmonary embolism. Int J Cardiol 2007;114(1):22-7.

[27] Cena JJ, Lalu MM, Cho WJ, et al. Inhibition of matrix metallopro-teinase activity in vivo protects against vascular hyporeactivity inendotoxemia. Am J Physiol Heart Circ Physiol 2010;298(1):H45-51.

[28] Gerlach RF, Uzuelli JA, Souza-Tarla CD, et al. Effect of antic-oagulants on the determination of plasma matrix metalloproteinase(mmp)-2 and mmp-9 activities. Anal Biochem 2005;344:147-9.

[29] Ceron CS, Castro MM, Rizzi E, et al. Spironolactone andhydrochlorothiazide exert antioxidant effects and reduce vascularmatrix metalloproteinase-2 activity and expression in a model ofrenovascular hypertension. Br J Pharmacol 2010;160(1):77-87.

[30] Castro MM, Rizzi E, Rodrigues GJ, et al. Antioxidant treatmentreduces matrix metalloproteinase-2-induced vascular changes inrenovascular hypertension. Free Radic Biol Med 2009;46(9):1298-307.