Secondary Logo

Journal Logo

Basic Science Aspects


Steinberg, Jay*; Halter, Jeffrey*; Schiller, Henry*; Gatto, Louis; Carney, David*; Lee, Hsi-Ming; Golub, Lorne; Nieman, Gary*

Author Information
Shock 24(4):p 348-356, October 2005. | DOI: 10.1097/01.shk.0000180619.06317.2c
  • Free



In the past 22 years, there were in excess of 10 million cases of sepsis reported in the United States alone (1). In 2001, it was estimated that there were over 750,000 cases of severe sepsis, with 215,000 deaths (2). Mortality can result from cardiovascular failure from septic shock or multiple system organ failure (MSOF), in particular, early pulmonary dysfunction in the form of the acute respiratory distress syndrome (ARDS) (3). ARDS develops in 30% to 43% of patients with severe sepsis and its manifestations correlate with an increased mortality (4).

ARDS originates when a traumatic event such as sepsis causes the release of proinflammatory mediators, predominately cytokines, initiating a systemic inflammatory response syndrome (SIRS) that ultimately results in recruitment of neutrophils into the pulmonary vasculature (5). With further stimulation, these “primed” sequestered neutrophils can release proteases, including matrix metalloproteinases (MMPs) and neutrophil elastase (NE) that damage the alveolar-capillary basement membrane, resulting in the high-permeability pulmonary edema that is the hallmark of ARDS (6-8). This damage results in a further release of inflammatory mediators, perpetuating a vicious cycle of inflammation and tissue destruction (9). This complex inflammatory cascade, with multiple overlapping and redundant initiators, has been exceedingly difficult to treat (10). To date, there are no FDA-approved drugs for the prevention or treatment of ARDS and only one agent, activated protein C (Xigris®), approved to treat severe sepsis.

The redundancy of released inflammatory mediators in sepsis and ARDS has limited the efficacy of drug therapies that target one specific mediator. In two previous studies using acute animal models of pulmonary injury, we demonstrated that blocking the proteases (NE, MMP-2, and MMP-9) with a unique modified tetracycline, COL-3, prevented the increase in pulmonary vascular permeability and, ultimately, ARDS (9, 11). Parallel studies in a cecal ligation and puncture (CLP) rat model demonstrated that COL-3 improved survival in a dose-dependent fashion when given at the time of injury, but before the onset of symptoms and significantly reduced lung tissue MMP-2 and MMP-9 levels (12). Although these studies strongly supported our original hypothesis (blocking neutrophil-derived proteases would prevent ARDS), we still needed to test the efficacy of COL-3 in a clinically applicable, insidious onset animal model before moving to clinical trials.

To this end, we developed a “two-hit” porcine model of sepsis plus gut ischemia-reperfusion (I/R) injury that parallels the pathogenesis of septic shock and ARDS in humans over a 48-h time period (13). We used this model in the current study and hypothesized that COL-3 would prevent the development of ARDS in treated animals by inhibiting multiple inflammatory mediators, including the proteases NE, MMP-2, and MMP-9.


Animal preparation

Female Yorkshire pigs (25-35 kg) were premedicated with glycopyrulate (0.01 mg/kg, i.m.) 10 min before intubations and were then anesthetized with xylazine (2 mg/kg, i.m.) and thiopental (10 mg/kg, i.v.). After intubation, the animals were ventilated with oxygen (1-2 L) delivered via a Veterinary Anesthesia Ventilator (Hallowell EMC; Matrix Medical, Orchard Park, NY). Continuous inhalation anesthesia was maintained using 1% to 3% isoflurane.

This animal model and protocol have been described in detail elsewhere (13). Briefly, under sterile conditions, cutdowns were performed on the right internal jugular vein and right carotid arterial line. Transducers leveled with the right atrium and recorded on a 16-channel PowerLab/16s (AD Instruments Pty Ltd., Milford, MA) and computer (Dell Dimension XPS R400; Dell Computers, Round Rock, TX) measured arterial pressure and was used for arterial blood gas sampling (models ABL 2 and OSM 3, Radiometer Inc., Paramus, NJ). The catheters were tunneled out through the subcutaneous tissue of the dorsolateral neck and were secured with 2-0 silk suture. The neck incision was closed in two layers using 3-0 chromic suture for the subcutaneous tissue and staples for the skin.

After placement of the catheters, baseline heart rate (HR), respiratory rate (RR), and temperature were obtained. Blood samples were obtained for blood culture, a complete blood count with differential, a lactate level, and cytokine and protease determination. Additionally, baseline arterial systemic pressure and an arterial blood gas were also obtained. A midline abdominal incision was made and the peritoneum was opened. A cystostomy was performed for insertion of a Foley catheter for monitoring continuous urine output. A Foley catheter was not inserted into the control pigs because, unlike the animals with superior mesenteric artery occlusion (SMA) + fecal blood clot (FC) that were lethargic for 48 h, the control animals were very active and would not tolerate the Foley.

The superior mesenteric artery was then clamped for 30 min to induce intestinal ischemia. The cecum was identified and eviscerated from the peritoneum and a 2-cm enterotomy was performed to obtain 0.5 mL/kg of feces that would be combined with 2 mL/kg of blood to create an FC. The enterotomy was allowed to remain patent. The clot was implanted into the lower portion of the abdominal cavity.

Experimental protocol

Animals were randomized into three groups the day before the surgery: experimental SMA + FC group (n = 7), consisting of gut ischemia/reperfusion injury in the form of cross-clamping of the superior mesenteric artery for 30 min followed by the intraperitoneal placement of an FC; 4-dedimethylaminosancycline or 6-demethyl-6-deoxy-4 dedimethylamino-tetracycline (COL-3) treatment group (SMA + FC + COL, n = 5) in which animals had COL-3 (200 mg/kg body weight) added to their food 12 h before surgery; and the control group (n = 3) was subjected to the identical surgery (except that the Foley catheter was not inserted) as groups 1 and 2 except that no placement of an FC and no clamping of the superior mesenteric artery. The route of COL-3 administration was necessary because COL-3 is currently only available in oral formulation and no additional COL-3 was given thereafter. COL-3 is a synthetic, nonantimicrobial derivative of the perhydronapthacene family of tetracycline-based compounds with a molecular weight of 371 (Fig. 1). During the surgical procedure, pigs received a fluid bolus of lactated Ringer's (25 mL/kg i.v. over 30 min) and broad-spectrum antibiotics (2 g of ampicillin i.v.; Bristol Myers Squibb, Princeton, NJ; and 500 mg of metronidazole i.v.; Baxter, Deerfield, IL).

FIG. 1:
Molecular structure of COL-3.

After full recovery from anesthesia, pigs were extubated and taken to an animal intensive care unit where they were followed continuously for 48 h. Twelve hours postsurgery, pigs received another dose of fluids (25 mL/kg i.v.) and antibiotics (2 g of ampicillin and 500 mg of metronidazole i.v.). This same regimen of fluids and antibiotics was given twice a day for the remainder of the experiment. Additional fluids (500 mL of lactated Ringer's i.v.) were administered if urine output decreased to less than 0.5 mL/kg/h or if mean arterial pressure (MAP) fell below 70 mmHg. The use of vasopressors was not permitted in this study. Pigs received pain medication in the form of i.v. buprenorphine HCL at 0.02 mg/kg every 6 h. Pigs were monitored 24 h a day with daily blood sampling for blood cultures, a complete blood count with differential, a lactate level, cytokine expression, and protease (NE, MMP-2, and MMP-9) determination. In addition, systemic arterial pressure, arterial blood gases, urine output, HR (EKG leads were placed on the pigs after surgery), RR, and temperature were monitored continuously and were recorded every 6 h along with a clinical assessment.

Development of ARDS

If the PaO2/FiO2 (P/F) ratio decreased to less than 250 mmHg, pigs were anesthetized (with i.v. pentobarbital) and placed on a mechanical ventilator (Galileo; Hamilton Medical Inc., Reno, NV) with volume-controlled ventilation (Vt 10 mL/kg), an RR necessary to maintain PCO2 between 35 and 40 torr and a positive end-expiratory pressure (PEEP) of 3 cm H2O. The mean FiO2 was 69.6% ± 6.2%. Pulmonary parameters (peak airway pressure, plateau airway pressure, static compliance, and airway resistance to inspiratory flow) were measured. Peak airway pressure by definition was the highest airway pressure measured during inspiration, plateau pressure was measured after an inspiratory pause of 5% of the total inspiratory time, and static compliance was calculated by dividing the plateau pressure by the tidal volume measured by the expiratory flow sensor. Anesthesia was maintained by a continuous infusion of pentobarbital (6 mg/kg/h) delivered intravenously by a Harvard infusion pump (model 907; Harvard Apparatus, Holliston, MA).

Once anesthetized and on the ventilator, a 7-French, flow-directed Swan-Ganz thermodilution catheter was passed through the right femoral vein into the pulmonary artery for determination of mixed venous blood gas (SVO2), arterial oxygen saturation (SaO2), mean pulmonary artery pressure, pulmonary artery wedge pressure, and cardiac output (CO). Alveolar-arterial PaO2 difference (A-a gradient) and pulmonary shunt fraction were determined using an Explorer Monitor (Baxter West Caldrun, NJ). EKG was measured with a pacemaker/defibrillator (Zoll Medical Gillette, NJ). Animals placed on the ventilator were followed and supported similar to unventilated animals until the end of the experiment (48 h). All pigs were euthanized with a pentobarbital overdose (150 mg/kg) 48 h postsurgery and the lungs and heart were removed en bloc with the lungs dissected free.

Plasma COL-3 measurement

The chemically modified tetracycline COL-3 (200 mg/kg body weight; CollaGenex Pharmaceuticals, Newtown, PA) was fed to animals the night before surgery. To assess for in vivo concentration of COL-3, 50-μL samples of plasma were obtained daily from the serum and incubated with 100 μL of precooled (−10°C) precipitating solution containing acetonitrile-methanol-0.5 M oxalic acid (60:30:10, v/v). The mixture was then centrifuged at 10,000 rpm for 5 min and the supernatant was collected for reverse phase HPLC analysis. COL-3 concentration was determined by injecting 25 μL of the supernatant into the HPLC system containing a C18 column (Waters System New Haven, CT) and eluted with acetonitrile:methanol (30:20) containing 5 mM oxalic acid at a flow rate of 1 mL/min. Final concentration was quantified by UV detection with peak area integration at 350 nm. The limit of detection in this system was 0.2 μg/mL. The assay described above was a modification of the technique described by us previously to measure different COLs in blood sera (14).

Bronchoalveolar lavage fluid (BALF) protein measurement

At necropsy, BALF was obtained for protein concentration in the following manner. The right middle lobe bronchus was cannulated and infused with 60 mL of saline as three aliquots of 20 mL each. Approximately 80% of the lavage fluid was recovered. Each aliquot was injected briskly and withdrawn slowly three times to obtain an optimal BALF specimen. The combined aliquots were centrifuged at 1000g for 10 min to remove cells, and the supernatant was frozen at −70°C for subsequent biochemical analysis. BALF protein analysis was based on the Bradford protein assay (Bio-Rad, Hercules, CA) with albumin as the standard. Standards ranged from 0 to 1000 μg/mL. Twenty milliliters of Coomassie Blue dye solution was diluted to 100 mL with saline. Ten microliters of standard solution or 10 mL of BALF was added to 5 mL of Coomassie Blue solution and the optical density was read at 590 nm in a spectrophotometer. The results were reported as micrograms of protein per milliliters of BALF.

MMP-2 and MMP-9 activity

The methods for purification of gelatinase and its assay by gelatin zymography have been fully described elsewhere (15). Briefly, zymography was performed on 2.5 μL of baseline and daily serum samples and BAL fluid as directed for precast NOVEX Zymogram (Invitrogen Corporation, Carlsbad, CA) gels to detect and characterize MMPs. These precast gels consist of a 10% SDS-polyacrylamide gel with 0.1% of gelatin incorporated as a substrate. The gels were run under nonreducing conditions with 1× Tris-Glycine SDS running buffer at 125V. The gels were then renatured with 2.5% Triton X 100 for 30 min, rinsed briefly, and developed overnight with 50 mM Tris/HCL buffer, pH 7.6, containing 10 mM CaCl2 at 37°C. Proteinases were easily identified as clear bands against a dark Coomassie Blue-stained background. Because of the effect of SDS, this zymography allows the determination of the molecular species of latent, activated, and complexed forms of gelatinases. The gelatin zymograms were then densitometrically scanned using a scientific imaging system (Eastman-Kodak, Rochester, NY) to determine the relative activity of MMP-9 and MMP-2. The data were presented as densitometric units.

Elastase activity

Elastase activity was determined in serum drawn at baseline and daily and in BALF obtained at necropsy. Specifically, elastase activity was determined by incubating 100 μL of serum or BALF and 400 μL of 1.25 mM methoxy succinyl-ala-pro-val-p-nitroanilide (specific synthetic elastase substrate) in a 96-well enzyme-linked immunosorbent assay (ELISA) plate at 37°C for 18 h. After incubation, the optical density was read at 405 nm. Data were expressed as micromoles elastase substrate degraded per milligram of protein per hour. These methods are described in full detail elsewhere (16).

Serum/BALF for TNF-α, IL-1, IL-6, IL-8, and IL-10

Serum was drawn at baseline and daily to determine serum levels of TNF-α, IL-1, IL-6, IL-8, and IL-10. Additionally, levels of all cytokines were determined in BALF. Serum and BAL levels of all cytokines were determined by ELISA, using a porcine cytokine ELISA kit (Endogen, Woburn, MA) according to the manufacturer's recommendations. Data was expressed in concentration as picograms per milliliter.


The lungs were assessed for pathologic injury in the following fashion. The left lower lobe bronchus was cannulated and the left lower lobe was inflated to a pressure of 25 cm H2O with 10% formalin. The cannula was clamped and the lung was stored in formalin at room temperature for 24 h. The tissue was blocked in paraffin, and serial sections were made for staining with hematoxylin and eosin. The lung tissue in each slide preparation was evaluated without knowledge of the treatment group from which it came. The slides were reviewed at low magnification for an overview to exclude sections containing bronchi, connective tissue, large blood vessels, and areas of confluent atelectasis, so that only regions reflecting the degree and stage of parenchymal injury would be evaluated. The areas of the slides that were not excluded were assessed at high magnification (×400) in the following manner. Five high-power fields (HPF) were randomly sampled. Features of alveolar wall thickening, intra-alveolar edema fluid, and the number of neutrophils were noted in each of the five HPF. Specifically, alveolar wall thickening, defined as greater than two cell layers thick, was graded as A0” (absent) or A1” (present) in each field. Intra-alveolar edema fluid, defined as homogenous or fibrillar proteinaceous staining within the alveoli, was graded as A0” (absent) or A1” (present) in each field. A total score/5HPF for alveolar wall thickening and intra-alveolar edema fluid was recorded for each animal. For example, in a given animal, if all five HPF evaluated demonstrated alveolar wall thickening and intra-alveolar edema fluid, the maximum score recorded would be 5/5HPF for each criteria. The total number of neutrophils was counted in each of the five HPFs and are expressed as the total number neutrophils/5HPF for each animal.

Pulmonary edema

To assess for evidence of pulmonary edema, representative tissue samples from the right lower and upper lobes and the left upper lobe were sharply dissected free of nonparenchymal tissue. Samples from the lung of each animal were placed in a dish and weighed, dried in an oven at 65°C for 24 h, and weighed again. This was repeated until there was no weight change over a 24-h period at which time the samples were determined to be dry. Lung water was expressed as a wet-to-dry weight ratio (W/D).


All data are expressed as means ± SE. A repeat analysis of variance (ANOVA) was used to test the difference between time points within each group for appropriate variables. Significant differences between the two groups at the end of each series of experiments for appropriate variables were determined by ANOVA. Whenever the F ratio indicated a significant difference, a Newman-Keul's post hoc test was used to identify the individual differences. Significance was assumed if the probability of the null hypothesis was less than 5% (P < 0.05).

Vertebrate animals

Vertebrate animal experiments described in this study were performed in accordance with the National Institutes of Health guidelines for the use of experimental animals in research. The protocol was approved by the Committee for the Humane Use of Animals at Upstate Medical University.



Animals remained unanesthetized until the P/F ratio fell below 250. At this point, the animal was anesthetized, a Swan-Ganz catheter was placed in the pulmonary artery, and the animal was placed on mechanical ventilation. Thus, in the following tables, the number of animals (n) changes for each parameter measured as animals reached criteria and were placed on mechanical ventilation. Thus, none of the control animals and only one of the COL-3-treated animals have any lung or pulmonary pressure data (Tables 1 and 2) until the 48-h reading because their P/F ratio did not fall below 250 throughout the study. When an animal met criteria (or at 48 h postsurgery), it was placed on volume-controlled ventilation (Vt 10 mL/kg) with an RR necessary to maintain PCO2 between 35 and 40 torr and a positive end-expiratory pressure (PEEP) of 3 cm H2O. No attempt was made to use protective modes of mechanical ventilation.

Table 1:
Hemodynamic and blood chemistry parameters
Table 2:
Pulmonary parameters


All animals that were subjected to SMA + FC, regardless if they were treated with COL-3, developed bacteremia. Bacteria cultured from the blood included Klebsiella pneumoniae, Serratia marcescens, Pseudomonas aeruginosa, Streptococcus, Aeromonas hydrophila, Escherichia coli, and Staphylococcus. There was a significant increase in white cell count (control, 12.8 ± 1.4 K/μL; SMA + FC, 19 ± 1.2* K/μL; SMA + FC + COL, 18.9 ± 1.7* K/μL; *P < 0.05 versus control, 48 h postinjury), HR, and temperature in both groups injured by SMA + FC (Table 1).

The increase in RR was blocked by COL-3 treatment (Table 1, 14 and 24 h postinjury, before animals were placed on the ventilator). Furthermore, COL-3 limited the platelet decrease seen in the SMA + FC group (control, 361 ± 12 K/μL, SMA + FC + COL, 250 ± 51 K/μL; SMA + FC, 95 ± 18 K/μL). COL-3 blood concentration was: day 1, 3.1 ± 0.3 μg/mL; day 2, 4.9 ± 1.0 μg/mL; and day 3, 3.1 ± 1.0 μg/mL. COL-3 concentration was zero in the other groups.

Septic shock

Septic shock in the untreated animals (SMA + FC) was evidenced by a fall in systemic blood pressure (Fig. 2). COL-3 treatment prevented the fall in blood pressure (Fig. 2) and also significantly reduced blood lactate levels (control, 1.5 ± 0.4 mmol/L; SMA + FC + COL, 2.0 ± 0.2 mmol/L; SMA + FC, 5.5 ± 1.3* mmol/L, *P < 0.05 versus both groups) and increased urine output (SMA + FC, 53 ± 8 mL/h; SMA + FC + COL, 251 ± 30* ml/h; *P < 0.05 versus SMA + FC) 48 h postinjury.

FIG. 2:
Mean arterial blood pressure. Data are mean ± SE. # P < 0.05 versus all groups.

In addition, COL-3 blocked the pH, base deficit, systolic and diastolic arterial blood pressure, pulmonary arterial pressure, and CO changes caused by SMA + FC (Table 1). There was no significant change in pulmonary artery wedge pressure in any group (Table 1). Thus, the COL-3-treated animals were bacteremic but did not develop septic shock.

Lung function

SMA + FC caused serious lung injury evidenced by a significant increase in plateau pressure, airway resistance, A-a gradient, pulmonary shunt and decrease in lung compliance, and P/F ratio (Table 2 and Fig. 3). COL-3 completely prevented all of the pathologic changes induced by SMA + FC in lung function (Table 2 and Fig. 3).

FIG. 3:
Arterial oxygenation expressed as P/F ratio (PaO2/FiO2). Data are mean ± SE. P < 0.05 versus all groups.

BALF cytokine and protease concentration

COL-3 (SMA + FC + COL) significantly blocked the increase in IL-6, IL-8, IL-10, and NE concentrations in BALF compared with the SMA + FC group (Table 3). IL-1 concentrations were not significantly different in any group (Table 3). TNFα levels were undetectable in BALF in all groups (data not shown). COL-3 reduced BALF MMP-2 by 63% and MMP-9 by 34% compared with the SMA + FC group, but these changes were not statistically significant (Table 3).

Table 3:

Plasma cytokine and protease concentration

There was no significant difference between groups in plasma NE, MMP-2, MMP-9, IL-8, or IL-10 levels in any group (Table 4). However, SMA + FC significantly increased plasma IL-1 and IL-6 concentrations on day 3. IL-1 was also increased on day 3 in the SMA + FC + COL group (Table 4). TNFα levels were undetectable in plasma at any time in all groups (data not shown).

Table 4:
Plasma proteases and cytokines


Treatment with COL-3 (SMA + FC + COL) significantly reduced interstitial alveolar wall thickness and intra-alveolar edema compared with the untreated group (SMA + FC; Table 5 and Fig. 4). Neutrophil recruitment into the lung was significantly increased in the SMA + FC groups compared with the control group, and COL-3 treatment (SMA + FC + COL) did not significantly reduce this influx of neutrophils (Table 5).

Table 5:
Pulmonary histology
FIG. 4:
(A) Control group demonstrating the thin alveolar walls and fully inflated alveoli typical of normal lungs. (B) SMA + FC group showing thickened and congested alveolar walls, intra-alveolar edema, and collapsed alveoli consistent with acute lung injury. (C) COL-3 treatment (SMA + FC + COL) prevented all of the pathologic changes seen in the SMA + FC group.

Pulmonary edema

Lung water assessed as the W/D was significantly increased in the SMA + FC group compared with the control and COL-3-treated groups (Fig. 5). Although the pulmonary artery pressure was significantly increased in the SMA + FC group, the pulmonary capillary wedge pressure was not. This plus the fact that BALF protein was significantly greater in the SMA + FC (Table 3) compared with the control or COL-treated group suggests a high-permeability edema typical of ARDS.

FIG. 5:
Pulmonary edema expressed as a wet-to-dry weight ratio. Data are mean ± SE. # P < 0.05 versus all groups.


The most important finding in this study is that prophylactic COL-3 completely prevented the lung injury associated with sepsis-induced ARDS and unexpectedly also prevented development of septic shock in spite of positive blood cultures. The protective effect of COL-3 was very dramatic and, in fact, most parameters in the COL-3-treated group were not significantly different from control. These data demonstrate that the combined injury of an FC plus SMA causes septic shock and ARDS in pigs in a time sequence and pathologic outcome analogous to septic shock and ARDS in humans. COL-3 not only prevents sepsis-induced ARDS, but it also prevents the deleterious systemic manifestations of septic shock. The gross photograph (Fig. 6) summarizes the near total protection offered the lung by COL-3.

FIG. 6:
Gross photos of lungs in the SMA + FC (A) and SMA + FC + COL (B) groups.

Importance of the sepsis model

An animal model that accurately duplicates the complex inflammatory and hemodynamic response that occurs in humans over several days during the development of septic shock and ARDS is critical to the understanding of sepsis pathophysiology (13). It has been shown that the success of clinical trials testing sepsis therapies depends upon sound preclinical data (17). Piper et al. (17) developed a paradigm to rate the results of preclinical sepsis treatments depending upon the animal models used. The highest recommendation for an animal model was “good evidence” that the results of the study would be similar in a clinical trial (17).

The importance of using a “good evidence” animal model was seen in the preclinical animal trials for the antiendotoxin, HA-1A (18). Initial studies demonstrated that HA-1A improved survival in a murine model of sepsis (19). However, in a phase III clinical trial, HA-1A did not show a survival benefit (20). However, when a chronic clinically relevant “good evidence” animal model of septic shock was used to test the efficacy of HA-1A, the results were very similar to the clinical trial and actually showed a decreased survival in the HA-1A-treated animals (18).

The consensus is that a chronic, insidious-onset large animal model that mimics the pathogenesis and time course of human septic shock and sepsis-induced ARDS is superior to small animal acute sepsis models as predictors of clinical efficacy (17). Our model contains all of the key components of a “good evidence” animal model, including a randomized, controlled study with supportive therapy (fluids and antibiotics) in a large animal, insidious-onset, chronic model of septic shock and ARDS (17). Our data clearly shows that animals progress from injury to SIRS to septic shock and finally to organ dysfunction (Figs. 2-6 and Tables 1-4). All pigs subjected to SMA + FC and not treated with COL-3 met the ARDS consensus conference criteria for acute lung injury (i.e., P/F < 250) (21) before the end of the study and by protocol were placed on mechanical ventilation. Histology and gross lung appearance were typical of that seen in human ARDS (Figs. 4 and 6). These data, combined with data from our initial work (13), strongly suggest that this model closely mimics human sepsis pathogenesis and thus, treatments that are effective in this model would very likely also be successful in human clinical trials.

Prophylactic treatment strategy

The ability to intervene before the onset of septic shock or ARDS is critical because it has been shown that pharmacologic intervention after disease onset is ineffective (10, 22). Prophylactic treatment strategies have been not been attempted clinically because of the high cost and risk of serious side effects with drugs such as anticytokines. However, COL-3 is an ideal candidate for prophylaxis treatment in patients at high risk for developing septic shock and/or ARDS (this includes patients with severe trauma, pancreatitis, major burns, or septicemia) because of its low toxicity and limited expense. COL-3 has already been studied in a phase I trial sponsored by the National Cancer Institution and has been reported to exhibit excellent pharmacokinetics in cancer patients and has shown efficacy in patients with Kaposi's sarcoma (23). Lastly, prophylactic treatment is not unique in trauma patients. Heparin and histamine blockers are given routinely to prevent deep vein thrombus and stress ulcers, respectively.

Proteases as a mechanism of sepsis-induced ARDS

Numerous animal studies have also demonstrated the importance of MMPs in the pathogenesis of sepsis-induced ARDS (9, 11, 12, 24, 25). Our laboratory has shown that MMP-2, MMP-9, and NE are increased in the lung tissue and BALF in several animal models of endotoxemia and sepsis (9, 11, 12). Martin et al. (24) demonstrated the importance of the MMP to TIMP ratio in the pathogenesis of sepsis-induced lung injury.They used TIMP-3 knockout mice and compared lung injury with wild-type mice in a CLP sepsis model. The TIMP-3 knockout mice had a significant increase in MMP-2 and MMP-9 and reduction in collagen and fibronectin levels. In addition, the collagen fibers that remained were disorganized, which suggests degradation by MMPs with a loss of structural integrity. Lois et al. (25) showed the importance of the antioxidant, glutathione, to MMP activation in a rat endotoxin plus fMLP model. Reduction of glutathione significantly increased MMP-2 and MMP-9 activity, which was associated with increased basement membrane degradation, measured by increased levels of the 7S fragment of type IV collagen in lung lavage. These findings strongly suggest that MMPs play a major role in the pathogenesis of ARDS. Elevated plasma NE levels are present in patients within minutes of serious trauma, and NE levels correlate with subsequent lung injury and progression to ARDS (26, 27).

COL-3 mechanism of action

COL-3 has been shown to inhibit MMP-2 and MMP-9 in several tissue types, including neutrophils and osteoclasts, cancer cells, and keratinocytes (for review, see Ref. 28). COL-3 appears to function as an MMP inhibitor by interfering with the mature active enzyme through competitive and noncompetitive chelation of the metal ions (Zn2+ and Ca2+), which are required for MMP activity, by interfering with the processing of the inactive precursors of MMPs into the mature enzymes, and by the down-regulating of MMP expression at the mRNA level (for review, see Ref. 28).

The impact of COL-3 on proteases in this study

Although the mechanism of COL-3 was not determined in this study, we nevertheless feel that NE and MMP inhibition by COL-3 play a major role. We have broken the inflammatory system down into initiators and effectors (9). Cytokines are the primary initiators and, although very important, do not directly damage lung tissue. NE and MMP are effectors working synergistically to cause tissue injury (2-9, 28). Also, it has been shown harmful to totally inhibit MMP (100% inactivation) because there is a baseline level of MMP activity necessary for normal physiologic processes. The beauty of COL is that it seems to reduce only the pathological activity of MMP-2 and MMP-9 (34% and 63%, respectively).

In this study, COL-3 significantly reduced NE concentrations in BALF and caused a significant trend toward reduction in MMP-2 (64% reduction, P = 0.291) and MMP-9 (34% reduction, P = 0.148) concentrations in BALF. NE was significantly elevated in the BALF of untreated animals and COL-3 significantly neutralized the increase BALF concentration of NE. These findings confirm those from our previous experiments (11).

MMP-2 was increased more dramatically in the BALF than was MMP-9. This could be explained by the fact that MMP-9 is only found in neutrophils and MMP-2 is found in multiple cell types, including fibroblasts, macrophages, and epithelial and endothelial cells. Because neutrophil recruitment into the lung was not significantly different between the COL-3-treated and untreated groups, one might not expect to find a significant difference between the two groups in terms of MMP-9 concentration. Increased MMP levels in the BALF were not as profound as those found in endotoxin (9, 11, 25) or CLP models (24). This may be because of the time postinjury that the plasma samples were harvested. It is known that the timing of MMP release is much different in acute animals models such as endotoxin injection compared with chronic clinically applicable models such as the one used in this study (10, 17). In the acute endotoxin studies (9, 11, 25), MMPs were measured no later than 4 h postendotoxin infusion (the length of the entire experiment), and in the CLP study (24), measurements were made 24 h after injury. In this study, MMP measurement in the BALF was 48 h after placement of the FC and SMA.

Lanchou et al. (29) measured MMP-2 and MMP-9 activity in ARDS patients for 12 days. They demonstrated that the activity of these MMPs in the BALF is significantly reduced within 4 days from the time that ARDS is first diagnosed (29). Measurements were only made every 4 days, therefore, it is unclear how soon the drop in MMP levels occurred. Thus, it is possible that our MMP levels were much higher at 24 h postinjury in the BALF in this study.

Another possible explanation for the relatively low MMP levels in the SMA + FC group could be because zymography fails to examine MMP activity in the presence of endogenous inhibitors (24, 29). Despite the fact that MMPs only rose moderately in the BALF in response to SMA + FC (see Table 3), COL-3 did affect a trend toward reduction in MMP-2 (64% reduction, P = 0.291) and MMP-9 (34% reduction, P < 0.148). NE was significantly elevated in the BALF of the SMA + FC group and was dramatically reduced with COL-3 treatment. Thus, inhibition of these proteases could be the mechanism by which COL-3 prevented ARDS in this study. However, COL-3 is pleiotropic, inhibiting multiple inflammatory mediators, and thus the exact mechanism of protection is unclear.

Other possible mechanisms of COL-3 protection

Although many biological agents, including TGF-β, interleukin 13, and dexamethasone, inhibit inducible nitric oxide synthase (iNOS) activity, few agents exert as profound an effect as COL-3 on iNOS expression and NO synthesis. COL-3 exhibited a time- and dose-dependent inhibition of NO production in cultured rat mesangial cells exposed to interferon-γ and endotoxin (30). In addition, COL-3 is a scavenger of reactive oxygen species that would be directly protective and also have a secondary protective effect by preventing HOCl from oxidatively activating pro-MMPs (31). Furthermore, COL-3 has been observed to inhibit two other mediators of inflammation, phospholipase A2 (32) and cyclooxygenase-2 (33).

It is known that tetracyclines can inhibit secretion of cytokines (15, 34, 35), but little is known about the impact of the modified tetracyclines (COL-3, etc.) on cytokine release. Recent studies have shown that COL-3 was much more effective than doxycycline in suppressing proinflammatory cytokine production by human monocytes in cell culture (H.-M. Lee, LM Golub, unpublished observations). Because all of the above mediators have been show to play a role in sepsis pathogenesis (36), it is possible that the mechanism by which COL-3 prevented septic shock and ARDS in this study was by blocking some or all of the these inflammatory mediators.

The impact of COL-3 on cytokines in this study

In this study, COL-3 effected a significant decrease in BALF IL-6, IL-8, and IL-10. In addition, COL-3 significantly reduced serum levels of IL-1 and IL-6 at 48 h post-SMA + FC. Preventing the increase in the inflammatory cytokines likely played a role, along with inhibition of NE and MMP, in the protective effect of COL-3 seen in these experiments. This study demonstrates that COL-3 has a dramatic impact on cytokine concentration in this sepsis + I/R model, which supports the data suggesting that tetracyclines in general and COL-3 specifically block cytokine increases in response to traumatic stimuli (15, 34, 35, 37). The mechanism of COL-3-induced reduction in cytokines concentrations during sepsis is unknown.


This study used a novel clinically applicable animal model in which the development of septic shock and ARDS was chronic and closely mirrored the pathogenesis of septic shock and ARDS seen in humans. Injury without COL-3 treatment caused septic shock and ARDS in all animals, whereas COL-3-treated animals were clinically indistinguishable from the control group. COL-3 is pleiotrophic, inhibiting numerous inflammatory mediators, making it difficult to pinpoint the exact mechanism of action; however, it may well be this pleiotrophic nature that is key to the effectiveness of COL-3. Because of the relatively low cost and toxicity profile of COL-3, we postulate that it can be delivered prophylactically at the time of trauma, similar to heparin for deep vein thrombus or histamine blocks for stress ulcers. This treatment strategy is directed at preventing the onset of septic shock and ARDS, which is preferable to treating these serious syndromes once they are developed. This work strongly suggests that COL-3 may reduce morbidity and mortality in patients at risk of developing sepsis or ARDS.


1. Martin GS, Mannino DM, Eaton S, Moss M: The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med 348:1546-1554, 2003.
2. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR: Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated cost of care. Crit Care Med 29:1303-1310, 2001.
3. Wheeler AP, Bernard GR: Treating patients with severe sepsis. N Engl J Med 340:207-214, 1999.
4. Gross CH, Brower RG, Hudson LD, Rubenfeld GD: ARDSnet. Incidence of acute lung injury in the United States. Crit Care Med 31:1607-1611, 2003.
5. Abraham E: Neutrophils and acute lung injury. Crit Care Med 31:S195-S199, 2003.
6. Gibbs DF, Shanley TP, Warner RL, Murphy HS, Varani J, Johnson KJ: Role of matrix metalloproteinases in models of macrophage-dependent acute lung injury. Evidence for alveolar macrophage as source of proteinases. Am J Respir Cell Mol Biol 20:1145-1154, 1999.
7. Moraes TJ, Chow C-W, Downey GP: Proteases and lung injury. Crit Care Med 31:S189-S194, 2003.
8. Lee WL, Downey GP: Leukocyte elastase. Physiological functions and role in acute lung injury. Am J Respir Crit Care Med 164:896-904, 2001.
9. Carney DE, McCann UG, Schiller HJ, Gatto LA, Steinberg J, Picone AL, Nieman GF: Metalloproteinase inhibition prevents acute respiratory distress syndrome. J Surg Res 99:245-252, 2001.
10. Zeni F, Freeman B, Natanson C: Anti-inflammatory therapies to treat sepsis and septic shock: a reassessment. Crit Care Med 25:1095-1100, 1997.
11. Carney DE, Lutz CJ, Picone AL, Gatto LA, Ramamurthy NS, Golub LM, Simon SR, Searles B, Paskanik A, Snyder K, Finck C, Schiller HJ, Nieman GF: Matrix metalloproteinase inhibitor prevents acute lung injury following cardiopulmonary bypass. Circulation 100:400-406, 1999.
12. Steinberg J, Halter J, Schiller HJ, Dasilva M, Landas S, Gatto LA, Maisi P, Sorsa T, Rajamaki M, Lee H-M, Nieman GF: Metalloproteinase inhibition reduces lung injury and improves survival after cecal ligation and puncture in rats. J Surg Res 111:185-195, 2003.
13. Steinberg JM, Halter JM, DaSilva M, Schiller H, Nieman G: The development of acute respiratory distress syndrome after gut ischemia/reperfusion injury followed by fecal peritonitis in pigs: a clinically relevant model. Shock 23:129-137, 2005.
14. Liu Y, Ramamurthy NS, Marecek J, Lee HM, Chen JL, Ryan ME, Rifkin BR, Golub LM: The lipophilicity, pharmacokinetics and cellular uptake of different chemically-modified tetracyclines (CMTs). Curr Med Chem 8:243-252, 2001.
15. Brown DL, Desai KK, Vakili BA, Nouneh C, Lee H-M, Golub LM: Clinical and biochemical results of the metalloproteinase inhibition with subantimicrobial dose of doxycycline to prevent acute coronary syndromes (MIDAS) pilot trial. Arterioscler Thromb Vasc Biol 24:733-738, 2004.
16. Ramamurthy NS, Golub LM: Diabetes increases collagenase activity in extracts of rat gingiva and skin. J Periodontal Res 17:455-462, 1983.
17. Piper RD, Cook DJ, Bone RC, Sibbald WJ: Introducing critical appraisal to studies of animal models investigating novel therapies in sepsis. Crit Care Med 24:2059-2070, 1996.
18. Quezado ZMN, Natanson C, Alling DW, Banks SM, Koev CA, Elin RJ, Josseini JM, Bacher JD, Danner RL, Hoffman WD: A controlled trial of HA-1A in a canine model of gram-negative septic shock. J Am Med Assoc 269:2221-2227, 1993.
19. Teng NNH, Kaplan HS, Herbert JM, et al.: Protection against gram-negative bacteremia and endotoxemia with human monoclonal IgM antibodies. Proc Natl Acad Sci USA 82:1790-1794, 1985.
20. Ziegler EJ, Fisher CJ, Sprung CL, Staube RC, Sadoff JC, Foulke GE, Wortel CH, Fink MP, Dellinger RP, Teng NN: Treatment of gram-negative bacteremia and septic shock with HA-1A human monoclonal antibody against endotoxin. N Engl J Med 266:1097-1102, 1991.
21. Bernard GR, Artigas A, Brigham KL, Carlet J, Falke K, Hudson L, Lamy M, Legall JR, Morris A, Spragg R: The American-European consensus conference on ARDS. Definitions, mechanisms, relevant outcomes, and clinical trials coordination. Am J Respir Crit Care Med 149:818-824, 1994.
22. Bernard GR: Sepsis trials: intersection of investigation, regulation, funding and practice. Am J Respir Crit Care Med 152:152-154, 1995.
23. Cianfrocca M, Cooley TP, Lee JY, Rudek MA, Scadden DT, Ratner L, Pluda JM, Figg WD, Krown SE, Dezube BJ: Matrix metalloproteinase inhibitor COL-3 in the treatment of AIDS-related Kaposi's sarcoma: a phase I AIDS malignancy consortium study. J Clin Oncol 20:153-159, 2002.
24. Martin EL, Moyer BZ, Pape MC, Starcher B, Leco KJ, Veldhuizen RAW: Negative impact of tissue inhibitor of metalloproteinase-3 null mutation on lung structure and function in response to sepsis. Am J Physiol Lung Cell Physiol 285:L1222-L1232, 2003.
25. Lois M, Brown LAS, Moss IM, Roman J, Guidot DM: Ethanol ingestion increases activation of matrix metalloproteinases in rat lung during endotoxemia. Am J Respir Crit Care Med 160:1354-1360, 1999.
26. Torri K, Iida K-I, Miyazaki Y, Saga S, Kondoh Y, Taniguchi H, Taki F, Takegi K: Higher concentrations of matrix metalloproteinases in bronchoalveolar lavage fluid of patients with adult respiratory distress syndrome. Am J Respir Crit Care Med 155:43-46, 1997.
27. Ricou B, Nicod L, Lacraz S, Welgus HG, Suter DM, Dayer JM: Matrix metalloproteinases and TIMP in acute respiratory distress syndrome. Am J Respir Crit Care Med 154:346-352, 1996.
28. Nieman GF, Zerler BR: A role for the anti-inflammatory properties of tetracyclines in the prevention of acute lung injury. Curr Med Chem 8:317-325, 2001.
29. Lanchou J, Corbel M, Tanguy M, Germain N, Boichot E, Theret N, Clement B, Lagentee V, Malledant Y: Imbalance between matrix metalloproteinases (MMP-9 and MMP-2) and tissue inhibitors of metalloproteinases (TIMP-1 and TIMP-2) in acute respiratory distress syndrome patients. Crit Care Med 31:536-542, 2003.
30. Trachtman H, Funerweit S, Greenwald R, Moak S, Singhal P, Franki N, Amin AR: Chemically modified tetracyclines inhibit inducible nitric oxide synthase expression and nitric oxide production in cultured rat mesangial cells. Biochem Biophys Res Commun 229:243-248, 1996.
31. Sorsa T, Ramamurthy NS, Vernillo AT, Zhang X, Konttinen YT, Rifkin BR, Golub LM: Functional sites of chemically modified tetracyclines: inhibition of the oxidative activation of human neutrophil and chicken osteoclast pro-matrix metalloproteinases. J Rheumatol 25:975-982, 1998.
32. Pruzanski W, Stefanski E, Vadas P, McNamara TF, Ramamurthy N, Golub LM: Chemically modified non-antimicrobial tetracyclines inhibit activity of phospholipase A2. J Rheumatol 25:1807-1811, 1998.
33. Patel RN, Attur MG, Patel IV, Abramson SB, Amin AR: A novel mechanism of action of chemically modified (CMTs): inhibition of COX-2 mediated PGE2 production. J Immunol 163:3459-3467, 1999.
34. Shapira L, Soskolne WA, Houri YM, Barak V, Halabi A, Stabholz A: Protection against endotoxic shock and lipopolysaccharide-induced local inflammation by tetracyclines: correlation with inhibition of cytokine secretion. Infect Immun 64:825-828, 1996.
35. Milano S, Arcoleo F, D'Agostino P, Cillari E: Intraperitoneal injection of tetracyclines protects mice form lethal endotoxemia downregulating inducible nitric oxide synthase in various organs and cytokine and nitrate secretion in blood. Antimicrob Agents Chemother 41:117-121, 1997.
36. Marshall JC: Such stuff as dreams are made on: mediator-directed therapy in sepsis. Nat Rev 2:391-405, 2003.
37. Lokeshwar BL, Escatel E, Zhu B: Cytotoxic activity and inhibition of tumor cell invasion by derivatives of a chemically modified tetracycline CMT-3 (COL-3). Curr Med Chem 8:271-279, 2001.

Modified tetracycline; COL-3; septic shock; ARDS; SIRS; matrix metalloproteinase; elastase

©2005The Shock Society