Carbon Dioxide sensing in Immune Cells

Carbon Dioxide sensing in Immune Cells


Million years ago carbon dioxide (CO2) levels used to be considerably higher compared to the measured levels today (Allègre and Schneider, 1994) (see Figure 1). Notably, the Mauna Loa Observatory in Hawaii repeatedly measured unprecedented inclines of CO2 levels over the last years with recent peaks of more than 400 ppm (i.e., atmospheric concentrations of 0.04%) (National Oceanic and Atmospheric Administration, 2019).

Figure 1. Graph of estimated atmospheric concentrations of oxygen (black lines; in %) and carbon dioxide (red lines, in ppm). While oxygen levels within our atmosphere continued to rise up to today 21%, carbon dioxide concentrations overall dropped down to 0.03-0.04% (i.e., 300-400 ppm). Note, that data concerning oxygen and carbon dioxide concentrations are derived from different models, and thus can only be seen as approximations (red and black shadowing). This graph is adapted from (Scott and Glasspool, 2006; dashed black lines), from (Falkowski et al., 2005; continuous black line) and from (Foster et al., 2017; red line).

One reason for the overall sharp decline in CO2 levels was the rise of photosynthesizing bacteria (i.e., cyanobacteria – green algae of the earth's oceans) and later plants that since then have been generating energy by using CO2 and producing oxygen (O2), thus, constantly enriching the earth’s atmosphere with O2 (“the oxidation event”) (Lyons et al., 2014). The fact that atmospheric CO2 levels markedly changed over time implicates that evolving organisms must have developed a distinct sensing mechanism that allows them to react and adapt to different levels of CO2. Indeed, plants, fungi and bacteria have developed unique CO2-sensing mechanisms (Cummins et al., 2014). Plants, for example, use one of the most abundant enzymes on earth ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO) to fix CO2 into biomass during the Calvin Cycle (Ellis, 1979).

There is, however, only limited knowledge on how mammalian and especially immune cells sense and react to high levels of CO2. The present mini-review will, therefore, highlight briefly how the human body reacts to different levels of CO2 and how immune cells might be able to sense CO2.


Carbon dioxide in water (H2O) quickly dissolves into bicarbonate (HCO3-) and proton molecules (H+), which at certain levels leads to an acidification (pH ↓) of an aqueous solution, such as blood (Lindskog and Coleman, 1973) (see Figure 2). In plants, bacteria and humans, this reaction is catalyzed by carbonic anhydrase (CA) and considered one of the fastest reactions in living organisms (Cummins et al., 2014; Lindskog and Coleman, 1973) (see Figure 2). Since CO2 is highly permissive and able to freely diffuse into cells over short distances (Missner and Pohl, 2009), mammals have several potent blood buffer systems including bicarbonates, phosphates and proteins functioning as either donors or acceptors of hydrogen ions or protons, thereby maintaining a pH equilibrium of approximately 7.35 to 7.45 (Elkinton, 1956). Changes in pH could, thus, be considered a surrogate parameter for different CO2 levels.

Figure 2. Enzymatic reaction in aqueous solutions. Reversible hydration of carbon dioxide (CO2) catalyzed by carbonic anhydrase (CA) producing bicarbonate (HCO3-) and protons (H+).

Next to enzymes, which use CO2 as a (co-)substrate (e.g., CA in humans or RuBisCO in plants), thus, representing putative CO 2-sensors, CO 2 on its own seems to directly act upon protein function via non-enzymatic carboxylation (also known as carbamylation) (Jimenez-Morales et al., 2014; Linthwaite et al., 2018). One of the most prominent examples of direct CO 2-mediated effects in human physiology is the impact of CO 2 on conformational changes of hemoglobin and thereby protein function (Bohr et al., 1904; Kilmartin and Rossi-Bernardi, 1969; Riggs, 1988). Carbon dioxide binds to hemoglobin, inducing accelerated O 2 dissociation from hemoglobin (“Bohr effect”) to improve critical tissue and organ oxygenation (Riggs, 1988). Intriguingly, estimations suggest that approximately 1-2% of all large proteins might be carboxylated due to elevated CO 2 levels (Jimenez-Morales et al., 2014).

In conclusion, it seems to be more likely that molecular CO 2-sensing is mediated by direct carboxylation of numerous distinct proteins, thus, affecting their protein function, rather than single cellular proteins functioning as an ubiquitously expressed “CO 2-sensor”.


Acute hypercapnia (i.e., increased levels of carbon dioxide; HC) and HC-driven acidosis (also called hypercapnic acidosis; HA) have immediate adverse effects on the human body ranging from dizziness and deficiency in motor performance to irreversible coma and even death (Case and Haldane, 1941). In addition to these central nervous effects, HC and HA significantly affect the cardio-vascular system and lead to an increase in heart rate and output, increased pulmonary and decreased systemic vascular tone, which overall improves tissue and organ oxygenation (Kavanagh and Laffey, 2006) (see Figure 3).

Figure 3. Schematic overview of CO2-mediated effects. High levels of CO2 (i.e., hypercapnia) impair neurological functions (left), affect the cardio-vascular system (middle) and mitigate lung injury due to overall immuno-suppressive effects (right). Note, that some of these effects are pH-dependent.

Central and peripheral nervous chemoreceptors detect hypercapnic blood levels within the human body and initiate a prompt stimulation of breathing rate and depth to normalize CO2 blood levels, thus, preventing troublesome hypercapnia and accompanied deleterious pH changes (Cummins et al., 2014). Interestingly, the breathing stimulus provided by chemoreceptors detecting high CO2 levels is significantly more sensitive compared to the stimulus induced by low O2 levels. While a decrease of 20-40 mmHg in partial pressure of oxygen (pO2) is necessary to induce a significant ventilatory response, an increase of only 10 mmHg in partial pressure of carbon dioxide (pCO2) is sufficient to induce the same response (Wilson et al., 2005; De Vito et al., 1998). This indicates that the physiological “window” of CO2 levels (i.e., blood levels of CO2 that can be well tolerated) is in fact considerably smaller than the one of changing O2 levels, suggesting an overall stricter control of CO2.

Patients experience HC and HA in a wide range of pathological states including chronic obstructive pulmonary disease and obstructive sleep apnea (Cummins et al., 2014). Intriguingly, analysis of CO2 levels revealed that the microenvironment of solid tumors is likewise characterized by elevated CO2 levels, indicating local HC within the tumor microenvironment (Gullino et al., 1965; Helmlinger et al., 2002).

The overall effects of HC on the immune system seem to be in part conflicting. On the one hand clinical studies could show that HC is associated with adverse clinical outcome in patients with sepsis due to pneumonia (Tiruvoipati et al., 2018) (see Figure 3). One explanation for the HC-associated increase in mortality in those patients might be that HC and HA significantly impair immune cell function of macrophages and neutrophils, thereby, leading to an overall immuno-suppressive phenotype (Casalino-Matsuda et al., 2015; Gates et al., 2013; Helenius et al., 2009) (see Figure 3). For this reason, there are currently numerous clinical trials investigating the beneficial effects of extra-corporal CO2 removal in critically ill patients with lung diseases (Tiruvoipati et al., 2018).

On the other hand animal studies could show that HA markedly reduces lung injury in several models of systemic abdominal sepsis (Curley et al., 2010) (see Figure 3). However, this effect was only evident upon additional antibiotic treatment of the underling bacterial infection (Curley et al., 2010). Moreover, the organ protective effects seem to be restricted to HA (and not HC), since buffered hypercapnia augments lung injury in models of pulmonary sepsis (Nichol et al., 2009). In addition, HC reduces proliferation of fibroblasts and alveolar epithelial cells due to mitochondrial dysfunction, which could overall reduce the wound healing capacity after organ injury in animals treated with HC alone (Vohwinkel et al., 2011) (see Figure 3). Mechanistically, the beneficial effects of HA on inflammatory cells seem to extend to the humoral response during systemic inflammation, as pro-inflammatory (potentially organ damaging) cytokines, such as interleukin 6 (IL6) and tumor necrosis factor (TNF) α, are significantly reduced after HA treatment during models of sepsis (Curley et al., 2010).

While the biological effects of CO2 on immune cells are well-described, it remains in part unclear whether all the above-mentioned effects on immune cell function are dependent on pH changes or solely mediated via CO2. Even more importantly, until today no ubiquitous molecular mechanism has been proposed on how immune cells sense and adapt to changing CO2 levels independent of pH changes. In this context, our research group could recently show that HC significantly attenuates pro-inflammatory NFκB signaling most probably by inducing nuclear translocation of IKK (inhibitor of κB kinase) α and RelB – both family members of the non-canonical NFκB pathway and important inhibitors of NFκB activity (Cummins et al., 2010; Keogh et al., 2017). Notably, these HC-sensitive effects were independent of pH changes and reversible, and could, therefore, represent one promising CO2-sensing mechanism of immune cells (Cummins et al., 2010; Keogh et al., 2017).

In conclusion, the immuno-modulatory (pro- versus anti-inflammatory) effects of CO2 seem to be highly context- and cell-type-specific (see Figure 3). Moreover, it is very likely that immune cells have evolved a molecular CO2-sensing mechanism independent of pH changes.


Investigating the effects of CO2 on gene and protein expression and the functional relevance concerning (immune) cell function remains challenging for the following reasons: (i) As outlined above, HC leads to profound changes in pH levels. Thus, discrimination between HC- and pH-mediated effects can be challenging. For this reason, many (not all) research groups adapt the supplemented buffer concentration of for example HCO3- within their cell culture medium to the applied CO2 concentration (Casalino-Matsuda et al., 2018; Cummins et al., 2010; Keogh et al., 2017). Alternatively, enzymes mediating pH changes, such as CA, can be pharmacologically inhibited with anti-diuretics, such as acetazolamide, to dampen HC-elicited pH changes (Cummins et al., 2010). Both strategies are used to determine biological effects that are only dependent on different levels of CO2. (ii) HC-mediated carboxylation of proteins (i.e., post-translational modifications) that might impact protein conformation or function are highly transient, and, therefore, difficult to detect via conventional mass spectrometry analysis (Linthwaite et al., 2018). Recently, Linthwaite and colleagues proposed that substrate trapping via triethyloxonium (TEO) treatment markedly increased the detection of protein carboxylation via mass spectrometry (Linthwaite et al., 2018). Using this remarkable approach the authors could confirm that hemoglobin is indeed non-enzymatically carboxylated at the known reaction sites due to high levels of CO2. Further small scale proteome analysis using extracts derived from the plant Arabidopsis thaliana could likewise reveal novel carboxylated proteins including peroxidase and iron superoxide dismutase 1 (Linthwaite et al., 2018). Notably, this proposed trapping strategy is of great interest, since it could potentially identify novel targets for protein carboxylation due to high levels of CO2 in any system.

Depending on the applied experimental conditions (e.g., buffered versus un-buffered media) the effects of different CO2 levels on cell function will greatly vary. This influencing factor needs to be considered when designing the experimental set-up. Furthermore, clinical translation of results derived from buffered cell culture experiments might be even more difficult than from un-buffered experiments, since one would expect that changing CO2 levels will eventually have an impact on pH both during physiological and patho-physiological conditions in “real-life”.


Profound changes of CO2 concentrations within our atmosphere over the last million years indicate that living organisms must have evolved a distinct CO2-sensing mechanisms to detect and adapt upon different CO2 levels. Indeed, plants, bacteria and mammals are able to detect different levels of CO2 via changes in pH on a cellular level. Recent findings using buffered experimental conditions, however, could show that CO2 directly affects the cell function of immune cells via decreased NFκB signaling independent of pH changes. Furthermore, CO2 elicits direct changes of protein conformation and function via non-enzymatic protein carboxylation (e.g., Bohr effect), suggesting CO2-adaptive pathways, which enable (immune) cells to sense and adjust upon different levels of CO2.

During inflammatory responses, such as sepsis or pulmonary infections, CO2 is highly immuno-modulatory with in part contrasting effects on the whole organism (organ protective) and individual immune cells (immuno-suppressive). A better understanding of these molecular CO2-sensing mechanisms will, therefore, have great therapeutic impact on all inflammation-driven diseases that are characterized by a hypercapnic micro-environment, including chronic lung and tumor diseases.


Allègre, C.J., Schneider, S.H., 1994. The Evolution of the Earth. Sci. Am. 271, 66–75.

Bohr, C., Hasselbach, K., Krogh, A., 1904. About a new biological relation of high importance that the blood carbonic acid tension exercises on its oxygen binding. Skand. Arch. Physiol., 16 402–412.

Casalino-Matsuda, S.M., Nair, A., Beitel, G.J., Gates, K.L., Sporn, P.H.S., 2015. Hypercapnia Inhibits Autophagy and Bacterial Killing in Human Macrophages by Increasing Expression of Bcl-2 and Bcl-xL. J. Immunol. Baltim. Md 1950 194, 5388–5396.

Casalino-Matsuda, S.M., Wang, N., Ruhoff, P.T., Matsuda, H., Nlend, M.C., Nair, A., Szleifer, I., Beitel, G.J., Sznajder, J.I., Sporn, P.H.S., 2018. Hypercapnia Alters Expression of Immune Response, Nucleosome Assembly and Lipid Metabolism Genes in Differentiated Human Bronchial Epithelial Cells. Sci. Rep. 8, 13508.

Case, E.M., Haldane, J.B.S., 1941. Human physiology under high pressure: I. Effects of Nitrogen, Carbon dioxide, and Cold. J. Hyg. (Lond.) 41, 225–249.

Cummins, E.P., Oliver, K.M., Lenihan, C.R., Fitzpatrick, S.F., Bruning, U., Scholz, C.C., Slattery, C., Leonard, M.O., McLoughlin, P., Taylor, C.T., 2010. NF-κB links CO2 sensing to innate immunity and inflammation in mammalian cells. J. Immunol. Baltim. Md 1950 185, 4439–4445.

Cummins, E.P., Selfridge, A.C., Sporn, P.H., Sznajder, J.I., Taylor, C.T., 2014. Carbon dioxide-sensing in organisms and its implications for human disease. Cell. Mol. Life Sci. 71, 831–845.

Curley, G., Contreras, M., Nichol, A.D., Higgins, B.D., Laffey, J.G., 2010. Hypercapnia and Acidosis in Sepsis: A Double-edged Sword? Anesthesiology 112, 462–472.

De Vito, E.L., Roncoroni, A.J., Berizzo, E.E., Pessolano, F., 1998. Effects of spontaneous and hypercapnic hyperventilation on inspiratory effort sensation in normal subjects. Am. J. Respir. Crit. Care Med. 158, 107–110.

Elkinton, J.R., 1956. Whole body buffers in the regulation of acid-base equilibrium. Yale J. Biol. Med. 29, 191–210.

Ellis, R.J., 1979. The most abundant protein in the world. Trends Biochem. Sci. 4, 241–244.

Falkowski, P.G., Katz, M.E., Milligan, A.J., Fennel, K., Cramer, B.S., Aubry, M.P., Berner, R.A., Novacek, M.J., Zapol, W.M., 2005. The Rise of Oxygen over the Past 205 Million Years and the Evolution of Large Placental Mammals. Science 309, 2202–2204.

Foster, G.L., Royer, D.L., Lunt, D.J., 2017. Future climate forcing potentially without precedent in the last 420 million years. Nat. Commun. 8, 14845.

Gates, K.L., Howell, H.A., Nair, A., Vohwinkel, C.U., Welch, L.C., Beitel, G.J., Hauser, A.R., Sznajder, J.I., Sporn, P.H.S., 2013. Hypercapnia impairs lung neutrophil function and increases mortality in murine pseudomonas pneumonia. Am. J. Respir. Cell Mol. Biol. 49, 821–828.

Gullino, P.M., Grantham, F.H., Smith, S.H., Haggerty, A.C., 1965. Modifications of the Acid-Base Status of the Internal Milieu of Tumors. JNCI J. Natl. Cancer Inst. 35, 857–869.

Helenius, I.T., Krupinski, T., Turnbull, D.W., Gruenbaum, Y., Silverman, N., Johnson, E.A., Sporn, P.H.S., Sznajder, J.I., Beitel, G.J., 2009. Elevated CO2 suppresses specific Drosophila innate immune responses and resistance to bacterial infection. Proc. Natl. Acad. Sci. U. S. A. 106, 18710–18715.

Helmlinger, G., Sckell, A., Dellian, M., Forbes, N.S., Jain, R.K., 2002. Acid Production in Glycolysis-impaired Tumors Provides New Insights into Tumor Metabolism. Clin. Cancer Res. 8, 1284–1291.

Jimenez-Morales, D., Adamian, L., Shi, D., Liang, J., 2014. Lysine carboxylation: unveiling a spontaneous post-translational modification. Acta Crystallogr. D Biol. Crystallogr. 70, 48–57.

Kavanagh, B.P., Laffey, J.G., 2006. Hypercapnia: permissive and therapeutic. Minerva Anestesiol. 72, 567–576.

Keogh, C.E., Scholz, C.C., Rodriguez, J., Selfridge, A.C., von Kriegsheim, A., Cummins, E.P., 2017. Carbon dioxide-dependent regulation of NF-κB family members RelB and p100 gives molecular insight into CO2-dependent immune regulation. J. Biol. Chem. 292, 11561–11571.

Kilmartin, J.V., Rossi-Bernardi, L., 1969. Inhibition of CO2 Combination and Reduction of the Bohr Effect in Haemoglobin chemically modified at its α-Amino Groups. Nature 222, 1243–1246.

Lindskog, S., Coleman, J.E., 1973. The Catalytic Mechanism of Carbonic Anhydrase. Proc. Natl. Acad. Sci. 70, 2505–2508.

Linthwaite, V.L., Janus, J.M., Brown, A.P., Wong-Pascua, D., O’Donoghue, A.C., Porter, A., Treumann, A., Hodgson, D.R.W., Cann, M.J., 2018. The identification of carbon dioxide mediated protein post-translational modifications. Nat. Commun. 9, 3092.

Lyons, T.W., Reinhard, C.T., Planavsky, N.J., 2014. The rise of oxygen in Earth’s early ocean and atmosphere. Nature 506, 307–315.

Missner, A., Pohl, P., 2009. 110 Years of the Meyer-Overton Rule: Predicting Membrane Permeability of Gases and Other Small Compounds. ChemPhysChem 10, 1405–1414.

National Oceanic and Atmospheric Administration, 2019. Monthly Average Mauna Loa CO2. URL (accessed 9.6.19).

Nichol, A.D., O’Cronin, D.F., Howell, K., Naughton, F., O’Brien, S., Boylan, J., O’Connor, C., O’Toole, D., Laffey, J.G., McLoughlin, P., 2009. Infection-induced lung injury is worsened after renal buffering of hypercapnic acidosis. Crit. Care Med. 37, 2953–2961.

Riggs, A.F., 1988. The Bohr Effect. Annu. Rev. Physiol. 50, 181–204.

Scott, A.C., Glasspool, I.J., 2006. The diversification of Paleozoic fire systems and fluctuations in atmospheric oxygen concentration. Proc. Natl. Acad. Sci. 103, 10861–10865.

Tiruvoipati, R., Gupta, S., Pilcher, D., Bailey, M., 2018. Hypercapnia and hypercapnic acidosis in sepsis: harmful, beneficial or unclear? Crit. Care Resusc. J. Australas. Acad. Crit. Care Med. 20, 94–100.

Vohwinkel, C.U., Lecuona, E., Sun, H., Sommer, N., Vadász, I., Chandel, N.S., Sznajder, J.I., 2011. Elevated CO(2) levels cause mitochondrial dysfunction and impair cell proliferation. J. Biol. Chem. 286, 37067–37076.

Wilson, D.F., Roy, A., Lahiri, S., 2005. Immediate and long-term responses of the carotid body to high altitude. High Alt. Med. Biol. 6, 97–111.

8th Jul 2020 M.J. Strowitzki, MD

Recent Posts