In the present study we investigated whether administration of exogenous HMW HA could be used as a therapy for sepsis-induced acute lung injury with mechanical ventilation. We demonstrated that pretreatment with HMW HA (1,600 kDa) inhibited inflammatory cell infiltration, cytokine production, and lung injury with mechanical ventilation. LMW HA (35 kDa) inhibited lung neutrophil infiltration and cytokine production, but did not inhibit lung injury or lung monocyte infiltration. HMW HA used in a therapeutic manner 1 hour after LPS infusion inhibited LPS-induced lung inflammation and lung injury in a dose-dependent manner.
HMW HA is an effective treatment in a variety of disease conditions. HMW HA has been shown to be a beneficial treatment for osteoarthritis. HMW HA can downregulate proinflammatory cytokines including IL-8, TNFα, and inducible nitric oxide synthase in fibroblast-like synoviocytes [18
]. HMW HA prevented acute liver injury by reducing plasma MIP-2, TNFα, and IFNγ in a T-cell-mediated liver injury mouse model [11
]. HMW HA has been shown to be protective in animal models of emphysema, can decrease the number of acute infections in chronic bronchitis in humans, can block group A streptococcus colonization in mice, can block pancreatic elastase-induced bronchoconstriction and neutrophil elastase-induced airway responses in sheep, can decrease peritoneal permeability secondary to infection in rats, and can reduce exercise-induced airway hyperreactivity in humans [19
]. Beneficial effects in sepsis, however, have not been previously demonstrated.
Our findings are consistent with previous observations that mice overexpressing HMW HA are protected from bleomycin-induced lung injury [9
]. Both HAS 1 and HAS 2 produced HMW HA. HASs are located on the cell surface and secrete the chains of HA into the extracellular matrix [1
]. We have found in the normal lung that HA is of the HMW HA form; however, in high-tidal-volume-induced lung injury in rats we found both HMW HA and LMW HA.
To establish whether HMW HA could potentially have beneficial effects in the treatment of sepsis we initially used pretreatment with an intraperitoneal injection of 35% HMW HA prior to LPS injection. This concentration, given intraperitoneally 18 hours before liver injury, has been shown to be absorbed into the systemic circulation and to prevent concavalin-induced liver injury [11
]. We then explored the use of HMW HA as a treatment for sepsis-induced lung injury. We had in previous studies found that HMW HA given intraperitoneally after smoke inhalation failed to protect lung injury (data not shown), probably related to delayed absorption, and 35% HMW HA was too viscous for intravenous injection. We therefore used continuous infusion of 0.025%, 0.05% and 0.1% HMW HA, concentrations that allowed intravenous infusion, starting 1 hour after injection of LPS.
We used intra-arterial LPS rather than the more conventional intravenous route of injection. We designed our model to produce lung injury over a 4-hour period that did not result in death but in a lung injury that was increased by high-tidal-volume ventilation over this period. We studied both venous and arterial injection. With arterial injection we found that there was increased neutrophil infiltration with arterial injection (61 × 103
± 10 cells/ml BAL fluid) than with venous injection (23 × 103
± 1 cells/ml BAL fluid, P
< 0.05). Based on this response we chose the intra-arterial route. Intra-arterial injection has been used in other models of sepsis [24
The difference in effects on acute lung injury scores between HMW HA and LMW HA may have been related to the different effects of the two molecular weights. HMW HA may block the effects of LMW HA produced in lung injury. The breakdown of HMW HA causes HA fragments to increase quickly and markedly in response to endotoxin [26
], and elevated levels of plasma HA fragments have been detected in patients with septicemia [27
]. LMW HA (200 kDa) isolated from the serum of patients with acute lung injury stimulated cytokine production in macrophages [9
]. LMW HA mediates bleomycin-induced lung injury [9
In previous studies, we demonstrated that de novo
synthesis of LMW HA by HAS 3 was induced in lung fibroblasts exposed to cyclic stretch via tyrosine kinase signaling pathways [7
]. In vivo
, very-high-tidal-volume ventilation (30 ml/kg) induced LMW HA production, was dependent on HAS 3, and resulted in increased neutrophil infiltration in the lungs of mice [8
]. Alternatively, the beneficial effects of HMW HA inhibition on inflammation may have been secondary to an increase in the ratio of HMW HA to LMW HA, thereby maintaining the level of HMW HA in the extracellular matrix and maintaining the integrity of the extracellular matrix [4
HA receptors include CD44, RHAMM, Toll-like receptor 2 and Toll-like receptor 4 [9
]. LMW HA induces cytokine production by binding to HA cell surface receptors. LMW HA (200 kDa) isolated from the serum of patients with acute lung injury stimulated cytokine production by binding to Toll-like receptor 2 and Toll-like receptor 4. LMW HA binding to Toll-like receptor 2 and Toll-like receptor 4 initiates mRNA expression by activation of the JNK pathways and through MyD88 activation [9
]. HMW HA has been shown to block the action of LMW HA by competing LMW HA binding to its receptors [35
]. The beneficial effects of HMW HA in this model of sepsis may have been secondary to HMW HA blocking the binding of LPS or LMW HA to Toll-like receptors, which mediate inflammation.
Surprisingly, infusion of LMW HA – at the size (35 kDa) and concentration (up to 1%) used in the present study – did not cause increased inflammation, and actually inhibited lung inflammation. LMW HA has been found to be proinflammatory by many authors, but not in all studies. Other authors have found that it is the protein and DNA contamination found in the LMW HA that is proinflammatory, and not the LMW HA itself [10
]. One explanation of the lack of proinflammatory effects of the HA used in this study is the purity of the compound. The HA used in the present study has <0.1% protein and <0.1 absorbance units of neucleotides. We cannot rule out longer exposures or higher concentrations of LMW HA or other sizes of LMW HA causing inflammation. Since the 35 kDa LMW HA failed to prevent acute lung injury on pathology in our pretreatment studies, we did not use LMW HA in the postinjury studies.
Both HMW HA and LMW HA inhibited MIP-2 production in the BAL and inhibited infiltration of neutrophils and monocytes into the lung. The inhibition of MIP-2 most probably accounts for this effect. It has been previously shown that a gradient of chemokines between the alveoli and the circulation is needed to induce migration into the alveolar space [15
]. We have previously shown that neutralization of MIP-2 in the airways prevents lung inflammatory cell infiltration in ventilator-induced lung inflammation [13
Alternatively the beneficial effects were not secondary HMW HA inhibition of LMW HA, but secondary to an increase in the ratio of HMW HA to LMW HA – thereby maintaining the level of HMW HA in the extracellular matrix and maintaining the integrity of the extracellular matrix [4
], and preventing the influx of inflammatory cytokines into the alveoli. This maintenance of the extracellular matrix may be an important mechanism in the prevention of acute lung injury by HMW HA.
Infusion of CMC, a carbohydrate with a positive charge similar size to the HMW HA, was used to control for the effects of charge. CMC blocked inflammation and lung injury – but not to the same extent as HMW HA. Since LMW HA and CMC partially blocked lung inflammation, the positive charge of HA may play a role in preventing lung injury by binding negatively charged inflammatory proteins.
One limitation of our study comparing LMW HA and HMW HA is the difference in molarity between the two infusions. We were unable to match molarity with the two infusions, since the high concentration of LMW HA that would be necessary to match the molarity between the solutions was not soluble. We cannot rule out that the beneficial effects of HMW HA may be due to the higher molarity of the solution being a better method of fluid resuscitation. We used additional saline infusions, however, to maintain the systolic blood pressure above 70 mmHg to eliminate hypotension as a confounding factor. The systolic blood pressure did not differ between groups.
An important part of our model is the use of mechanical ventilation with LPS infusion. The management of acute respiratory failure requires the use of positive-pressure mechanical ventilation to provide adequate ventilation and oxygenation. But mechanical ventilation with a high tidal volume leads to ventilator-induced lung injury by alveolar overdistention coupled with repeated collapse and reopening during mechanical ventilation, which initiates a cascade of proinflammatory cytokines. Even mechanical ventilation with moderate tidal volumes can augment the sepsis-induced lung injury by synergistically increasing lung cytokines, and may play a pivotal role in the development of acute lung injury in patients with sepsis [37
]. The augmentation of acute lung injury by high tidal volumes has been termed ventilator-associated lung injury.