The immune system can be viewed as having two inter-connected branches, namely the innate and adaptive immune responses 5. The innate immune system is an ancient, highly conserved response, being present in some form in all metazoan organisms. This response is activated by components of the wall of invading micro-organisms, such as lipopolysaccharide (LPS) or peptidoglycan, following the binding of these pathogen-associated molecular patterns to pattern recognition receptors, such as the Toll-like receptors (TLRs) on tissue macrophages. The innate immune response is also activated by endogenous 'danger' signals, such as mitochondrial components 15, providing an elegant explanation for why non-septic insults can also lead to organ injury and dysfunction. An inflammatory cascade is then initiated, involving cytokine signaling activation of phagocytes that kill bacteria, as is activation of the (later) adaptive immune response.

Hypercapnic acidosis has been demonstrated to inhibit multiple components of the host innate immune responses. Activation of the innate immune response initiates a conserved signaling cascade that culminates in the activation of transcription factors, such as nuclear factor kappa-B (NF-κB) 5. These transcription factors drive the expression of multiple genes that activate and regulate the pro-inflammatory and repair processes. Increasing evidence suggests that hypercapnic acidosis directly inhibits the activation of NF-κB 16. Intriguingly, this effect of hypercapnic acidosis may be a property of the CO₂ rather than its associated acidosis 171819. If confirmed, this finding suggests the presence of a molecular CO₂ sensor in mammalian cells. This mechanism of action of hypercapnic acidosis has been demonstrated to underlie some of the anti-inflammatory effects of hypercapnia 16, and to be a key mechanism by which hypercapnia - whether buffered or not - reduces pulmonary epithelial wound healing 18.

Hypercapnic acidosis also interferes with coordination of the innate immune response by reducing cytokine signaling between immune effector cells. Hypercapnic acidosis reduces neutrophil 20 and macrophage 21 production of pro-inflammatory cytokines such as tumor necrosis factor (TNF)-α, interleukin (IL)-1β, IL-8 and IL-6. Hypercapnic acidosis reduced endotoxin stimulated macro-phage release of TNFα and IL-1β in vitro 21. Peritoneal macrophages incubated under hypercapnic conditions demonstrated a prolonged reduction in endotoxin-stimulated TNF-α and IL-1β release 22. In contrast, a recent study reported rapid onset and rapid reversibility of IL-6 inhibition by hypercapnia in mature macrophage stimulated with LPS 19. The mechanism underlying hypercapnic acidosis-mediated inhibition of cytokine and chemokine production appears to be mediated at least in part via inhibition of activation of NF-κB.

Neutrophils and macrophages are important effectors of the innate immune response in the setting of bacterial infection. Neutrophils rapidly migrate from the blood-stream to areas of infection, and rapidly phagocytose invading microorganisms. Tissue macrophages and their blood borne monocyte counterparts are activated by bacterial products such as endotoxin, and coordinate the activation of the adaptive immune response in the setting of infection by presenting foreign antigen to lymphocytes and secreting chemokines. Both monocytes and macrophages phagocytose and kill pathogens by similar mechanisms but at a slower rate than neutrophils.

Hypercapnic acidosis may impact on the cellular immune response via both direct and indirect mechanisms. Hypercapnic acidosis inhibits neutrophil expression of the chemokines, selectins and intercellular adhesion molecules 1620, which facilitate neutrophil binding to the endothelium and migration out of the vascular system. The potential for hypercapnic acidosis to inhibit neutrophil chemotaxis and migration to the site of injury has been confirmed in vivo, where hypercapnic acidosis inhibits pulmonary neutrophil infiltration in response to endotoxin instillation 11. Hypercapnic acidosis directly impairs neutrophil phagocytosis in vitro 23. This inhibitory effect appears to be a function of the acidosis per se, with buffering restoring neutrophil phagocytosis 24. Hypercapnic acidosis also inhibits phagocytosis of opsonized polystyrene beads by human alveolar macrophages, although the levels of CO₂ utilized to demonstrate this effect were well beyond the range encountered clinically 19.

Neutrophils and macrophages kill ingested bacteria by producing free radicals such as superoxide, hydrogen peroxide, and hypochlorous acid, and releasing these into the phagosome. This is a pH-dependent process, with free radical production decreased at low pH 25. Hypercapnic acidosis inhibits the generation of oxidants such as superoxide by unstimulated neutrophils and by neutrophils stimulated with opsonized Escherichia coli or with phorbol esters 20. In contrast, hypocapnic alkalosis stimulates neutrophil oxidant generation 20. Inhibition of the intracellular pH changes with acetazolamide attenuated these effects. More recently, hypercapnic acidosis has been demonstrated to reduce oxidative reactions in the endotoxin injured lung by a mechanism involving inhibition of myeloperoxidase-dependent oxidation 26. The potential for hypercapnic acidosis to reduce free radical formation, while beneficial where host oxidative injury is a major component of the injury process, may be disadvantageous in sepsis, where free radicals are necessary to cause bacterial injury and death.

Neutrophil apoptosis following phagocytic activity generally occurs within 48 hours of release into the circulation. Conversely, neutrophil death via necrosis causes release of intracellular contents, including harmful enzymes, which can cause tissue destruction. Neutrophils appear to have an increased probability of dying by necrosis following intracellular acidification during phagocytosis 27. Hypercapnic acidosis may, therefore, increase the probability of neutrophil cell death occurring via necrosis rather than apoptosis.

The initial host response to invading pathogens is dominated by neutrophil activation, migration to the infective site, and phagocytosis and killing of bacteria. Compartmentalized release of neutrophil proteolytic enzymes and myeloperoxidase-dependent oxygen radicals results in effective pathogen destruction. However, excessive release of these potent mediators into the extracellular space results in damage to host tissue and worsening ALI. Consistent with this is the finding that recovery of neutrophil count in neutropenic patients worsens the severity of ALI 35. Hypercapnic acidosis may reduce the potential for damage to host tissue during the response to infection, by reducing lung neutrophil recruitment 10, adherence 16, intracellular pH regulation 12, oxidant generation 8, and phagocytosis 23. These mechanisms are considered to underlie some of the protective effects of hypercapnic acidosis in nonsepsis induced ALI 7. However, these effects of hypercapnic acidosis may be detrimental in sepsis, given the central role of neutrophil mediated phagocytosis of microbial pathogens and activation of the cytokine cascade to the host response to infection. In this context, defects in neutrophil function are associated with increased sepsis severity and worse outcome 36.

The effects of this hypercapnic acidosis-induced immune modulation may vary depending upon the stage of the infective process. The anti-inflammatory properties of hypercapnic acidosis may reduce the intensity of the initial host response to infection, thus attenuating tissue damage (Figure 2). However, the mechanisms whereby bacteria mediate tissue injury are complex and not limited to the contribution from an excessive host response. In late or prolonged pneumonia, in which tissue injury from direct bacterial spread and invasion makes a significant contribution, hypercapnic acidosis might impair bactericidal host responses. In the absence of effective antibiotic therapy, this may lead to enhanced bacterial spread and replication leading to more severe tissue destruction and lung and systemic organ injury (Figure 2).

Hypercapnic acidosis has been demonstrated to retard the repair process following lung cell and tissue injury. Hypercapnia slowed resealing of stretch-induced cell membrane injuries 37 and inhibited the repair of pulmonary epithelial wounds 18 by a mechanism involving inhibition of the NF-κB pathway. These findings raise the potential that hypercapnic acidosis could lead to increased bacterial translocation through defects in the pulmonary epithelium, while also delaying the recovery process following a septic insult.

Recent studies in relevant preclinical models have significantly advanced our understanding of the effects of hypercapnic acidosis in both pulmonary and systemic sepsis-induced ALI/ARDS. These studies reveal the importance of severity, site, and stage of the infective process, the need for antibiotic therapy, and the utility of buffering the hypercapnic acidosis in this setting.

The effect of hypercapnic acidosis on pneumonia-induced ALI appears to depend on the stage and severity of the infection. In an acute severe bacterial pneumonia-induced lung injury, hypercapnic acidosis improved physiological indices of injury 38. Intriguingly, these protective effects were mediated by a mechanism independent of neutrophil function. In contrast, hypercapnic acidosis did not alter the magnitude of lung injury in a less severe acute bacterial pneumonia 39. Importantly these in vivo studies showed no increase in bacterial count in animals exposed to hypercapnic acidosis, a reassuring finding given concerns regarding retardation of the host bactericidal response and potential bacterial proliferation.

In the clinical setting, many critically ill patients will have established infection at the time of presentation. Thus animal models of established bacterial pneumonia, in which hypercapnic acidosis was introduced several hours following induction of infection with E. coli, more closely resemble the clinical setting. In an established pneumonia model, hypercapnic acidosis induced after the development of a significant pneumonia-induced lung injury reduced physiological indices of lung injury 40. Of importance, these protective effects of hypercapnic acidosis were enhanced in the presence of appropriate antibiotic therapy 40. Again, reassuringly, lung bacterial loads were similar in the hypercapnic acidosis and normocapnia groups 40.

In an animal model of prolonged untreated pneumonia, sustained hypercapnic acidosis worsened histological and physiological indices of lung injury, including compliance, arterial oxygenation, alveolar wall swelling and neutrophil infiltration 23. Of particular concern to the clinical setting, hypercapnic acidosis was associated with a higher lung bacterial count. The mechanism underlying this effect appeared to be inhibition of neutrophil function, as evidenced by impaired phagocytotic ability in neutrophils isolated from hypercapnic rats 23. Of importance to the clinical context, the use of appropriate antibiotic therapy abolished these deleterious effects of hypercapnia, reducing lung damage and lung bacterial load to levels comparable to those seen with normocapnia.

These findings have been confirmed and considerably expanded in a recent study of hypercapnia in the fruit fly 41. Helenius et al., in a series of elegant in vivo studies, found that prolonged hypercapnia decreased expression of specific anti-microbial peptides in Drosophilia melanogaster 41. Hypercapnia decreased bacterial resistance in adult flies exposed to pathogens as evidenced by increased bacterial loads and increased mortality in flies inoculated with E. faecalis, A. tumefaciens, or S. aureus 41. The previously demonstrated suppressive effects of hypercapnic acidosis on the NF-κB pathway appeared to underlie the decreased resistance to infection 41. These findings raise significant concerns regarding the safety of hypercapnia in the setting of prolonged pneumonia, particularly in the absence of effective antibiotic therapy.

Hypercapnic acidosis reduces the severity of early septic shock and lung injury induced by systemic sepsis. In a rodent model of peritoneal sepsis induced by cecal ligation and puncture, hypercapnic acidosis slowed the development of hypotension, preserved central venous oxygen saturation, and attenuated the rise in serum lactate compared to control conditions, in the first 3 hours post injury 42. The severity of early lung injury was reduced as evidenced by a decrease in the alveolar-arterial oxygen gradient, and reduced lung permeability, compared to normocapnia. Alveolar neutrophil concentration was reduced by hypercapnic acidosis but IL-6 and TNF-α were unchanged 42. Of importance, there were no differences in bacterial loads in the lung, blood, or peritoneum in the hypercapnia group.

Using an ovine model of fecal peritonitis, Wang et al compared the effects of hypercapnic acidosis with those of dobutamine 43. Over an 18-hour study period, hypercapnic acidosis resulted in improved hemodynamics of a magnitude comparable to that of dobutamine. Compared with normocapnia, both hypercapnic acidosis and dobutamine raised cardiac index and systemic oxygen delivery and reduced lactate levels. In addition, hypercapnic acidosis attenuated indices of lung injury, including lung edema, alveolar-arterial oxygen partial pressure difference and shunt fraction. Hypercapnic acidosis did not decrease survival time compared to normocapnia in this setting 43. In a more prolonged systemic sepsis model, Costello et al. demonstrated that sustained hypercapnic acidosis reduced histological indices of lung injury compared with normocapnia in rodents following cecal ligation and puncture 42. Reassuringly there was no evidence of an increased bacterial load in the lung, blood, or peritoneum of animals exposed to hypercapnia.

Direct intra-abdominal administration of CO₂ - by means of a pneumoperitoneum - reduces the severity of abdominal sepsis-induced lung and systemic organ injury. Insuation of CO₂ into the peritoneal cavity prior to laparotomy for endotoxin contamination increased animal survival 44. Most recently, CO₂ pneumoperitoneum has been demonstrated to increase survival in mice with polymicrobial peritonitis induced by cecal ligation and puncture (Figure 3) 31. These protective effects of intraperitoneal carbon dioxide insufflation appear be due to the immunomodulatory effects of hypercapnic acidosis, which include an IL-10 mediated downregulation of TNF-α 44. Importantly, these effects appear to be mediated by the localized peritoneal acidosis, rather than by any systemic effect.

In rodent models of acute pneumonia induced by intra-tracheal E. coli and by endotoxin, buffered hypercapnia worsened lung injury 24. Compared with normocapnic controls, buffered hypercapnia increased multiple indices of lung injury including arterial oxygenation, lung compliance, pro-inflammatory pulmonary cytokine concentrations, and measurements of structural lung damage. In these experiments, buffered hypercapnia was established in the animals by exposure to hypercapnic conditions until renal buffering to normal pH had occurred, thus avoiding the confounding effects of exogenous acid or alkali administration. This contrasts with the protective effects of hypercapnic acidosis in similar models 1138. Of note, buffered hypercapnia did not reduce the phagocytic capacity of neutrophils, and did not increase lung bacterial load in these studies 24.

In a study designed to assess the contribution of acidosis versus hypercapnia to the effects of hypercapnic acidosis on the lung and hemodynamic profile in systemic sepsis, Higgins et al. exposed rats to environmental hypercapnia until renal buffering had restored pH to the normal range 45. Both buffered hypercapnia and hypercapnic acidosis reduced the severity of early shock and attenuated the increase in serum lactate compared with normocapnia. In contrast, buffered hypercapnia did not attenuate physiologic or histologic indices of lung injury in these studies 45. Reassuringly, there was no evidence to suggest that buffered hypercapnia worsened the degree of lung injury compared to normocapnia, and buffered hypercapnia did not increase the bacterial load in the lungs or the bloodstream 45.