*2.6. Treatment with* α*-Ketoglutaric Acid Ameliorates Hypercapnia-Induced ER Dysfunction and Restores Na,K-ATPase Function*

Up to this point, our studies have suggested that hypercapnia enhances retention of ER-resident NKA-β due to increased protein oxidation in the ER. In a recent publication, it has been shown that elevated CO2 levels inhibit expression of isocitrate dehydrogenase 2 (IDH2), which is a key enzyme in the tricarboxylic acid (TCA) cycle. Thereby, it causes mitochondrial dysfunction and ATP depletion [24]. Furthermore, treatment with α-ketoglutaric acid (α-KG) compensated the loss of IDH2 and raised ATP levels in hypercapnia-exposed cells [24]. Thus, we hypothesized that the increased ER oxidation by elevated CO2 levels was due to mitochondrial dysfunction that might be rescued by administration of α-KG. Treatment of AEC with α-KG prevented the hypercapnia-induced increase in protein oxidation in the ER (Figure 7A).

Next, we determined the effects of α-KG treatment on ER-resident and PM-located NKA-β upon hypercapnia. AEC were exposed to normal or elevated levels of CO2 for 12 h in the presence or absence of α-KG and, subsequently, the ER faction and cell surface abundance of NKA-β were measured (Figure 7B,C). Our results revealed that α-KG treatment rescues the effects of hypercapnia on the levels of NKA-β in the ER, which was associated with an increased expression of NKA-β at the PM. Moreover, to determine whether an increased cell surface abundance of the NKA-β resulted in an elevation of NKA activity, we measured ouabain-sensitive ATPase activity in primary and cultured AEC (Figure 7D,E) and observed that treatment with α-KG partially restored NKA function upon hypercapnia. Lastly, treatment with α-KG partially restored the hypercapnia-induced reduction in total intracellular ATP levels (Figure 7F). Taken together, these studies suggest that rescuing mitochondrial function and administration of α-KG restore the hypercapnia-induced ER retention of NKA-β, which increases PM abundance and activity of NKA.

**Figure 7.** Treatment withα-KG reverses hypercapnia-induced protein oxidation and decreases ER-retained forms of Na,K-ATPase. (**A**) A549 cells were exposed to 40 (Ctrl) or 120 mmHg CO2 (CO2) with an extracellular pH = 7.4 for 12 h in the presence or absence of α-KG (10 mM) or vehicle. The ER fraction was isolated by ultracentrifugation and the level of protein oxidation was assessed as described above. Bars represent 2,4-DNPH/Coomassie ratio. Values are expressed as mean ± SD (*n* = 3, \* *p* < 0.05, n.s—non-significant). (**B**) A549 cells were exposed to 40 (Ctrl) or 120 mmHg CO2 (CO2) with an extracellular pH = 7.4 for 12 h in the presence or absence of α-KG or vehicle. NKA-β levels were measured by immunoblotting. Representative immunoblots are shown. Bars represent the ER-resident NKA-β/β-actin ratio. Values are expressed as mean ± SD (*n* = 5, \*\*\* *p* < 0.001, n.s. – non-significant). (**C**) A549 cells were exposed to 40 (Ctrl) or 120 mmHg CO2 (CO2) for 12 h in the presence or absence of α-KG or vehicle. NKA-β PM abundance was determined by biotin-streptavidin pull-down and immunoblotting. Representative immunoblots of NKA-α, NKA-β, and TfR at the PM are shown. Bars represent the NKA-β/TfR ratio. Values are expressed as mean ± SD (*n* = 5, \* *p* < 0.05, n.s—non-significant). (**D**) Primary rat ATII and (**E**) A549 cells were exposed to 40 (Ctrl) or 120 mmHg CO2 (CO2) with extracellular pH = 7.4 for 12 h in the presence or absence of α-KG or vehicle. The PM fraction was isolated by ultracentrifugation and ouabain-sensitive Na,K-ATPase activity was measured by using a colorimetric ATP bioluminescence assay kit. Values are expressed as mean ± SD (*n* = 5, \*\*\* *p* < 0.001, n.s. – non-significant). (**F**) A549 cells were exposed to 40 (Ctrl) or 120 mmHg CO2 (CO2) with an extracellular pH = 7.4 for 12 h in the presence or absence of α-KG or vehicle. Intracellular ATP levels were measured by using an ATP bioluminescence assay kit. Bars represent total ATP/total protein ratio. Values are expressed as mean ± SD (*n* = 3, \* *p* < 0.05, \*\* *p* < 0.01).

#### **3. Discussion**

In the present study, we show that, in alveolar epithelial cells, hypercapnia downregulates NKA function by promoting oxidation of NKA-β, which, thereby, impairs its interaction with NKA-α and retains the misfolded NKA-β in the ER. This decreases PM abundance and activity of the transporter. In addition, we demonstrate that treatment with α-KG, which is an active metabolite, that has been previously shown to rescue the impaired TCA cycle and ATP production in hypercapnia-exposed cells, attenuates the elevated CO2–induced NKA-β oxidation in the ER, which increases the abundance of the enzyme at the PM.

Effective alveolar gas exchange requires relatively "dry" airspaces achieved by an active vectorial Na<sup>+</sup> transport process across the alveolar epithelium [28], driven by the basolaterally-located NKA and the apical epithelial sodium channel, ENaC, which promotes alveolar fluid clearance [17,29,30]. It is well documented that impaired function/expression of these transporters have a deleterious impact on the resolution of lung edema in models of ALI [31,32]. In patients with ARDS, disruption of the alveolar-capillary barrier causes pulmonary edema that severely alters gas exchange and

leads to hypoxia and hypercapnia. Since patients with ARDS require ventilation with low tidal volumes to limit ventilator-induced lung injury, hypercapnia is often further aggravated in this patient group [11,12], which has been tolerated in the last two decades. This is a concept termed "permissive hypercapnia." Hypercapnia is a double-edged sword and the associated acidosis may exhibit advantageous anti-inflammatory effects [33]. However, recent reports suggest that, in patients with severe hypercapnia, elevated levels of CO2 are associated with higher complication rates, more organ failure, worse outcome, and increased risk of intensive care unit mortality [34]. A possible explanation, in addition to the recently established negative effects of elevated CO2 levels on innate immunity and host defense [35–37], might be the hypercapnia-driven impairment of alveolar edema resolution due to the high CO2-driven downregulation of PM abundance and function of NKA and ENaC [14,38].

We have previously reported that hypercapnia rapidly (within minutes) decreases NKA PM abundance by inducing phosphorylation of the α-subunit of the transporter at the Ser18 residue by protein kinase C- ζ that is activated by a CO2-specific signaling pathway including subsequent phosphorylation of extracellular signal-regulated kinase and AMP-activated protein kinase, whereas the c-Jun N terminal kinase drives NKA retrieval from the cell surface [14–16,38]. In contrast, potential effects of elevated CO2 levels on the regulatory β-subunit of NKA have not been previously reported. In the current study, we describe a novel mechanism by which sustained hypercapnia leads to misfolding of NKA-β in the ER. The NKA-β is a glycoprotein with a molecular mass depending on its glycosylation profile of ~35–55 kDa. Once being synthetized by the ribosomes, the nascent NKA-β protein is co-translationally transported to the ER, where high mannose N-glycan subunit forms are generated. In the ER, the NKA-β undergoes folding, posttranslational modifications, and assembling with the NKA-α. Subsequent maturation and further glycosylation of NKA-β results in the formation of complex-type N-glycan forms, which are located in the Golgi and at the PM [20]. The ER maturation steps play a pivotal role in the NKA-β glycosylation processing [19,21] as well as in the formation of the α:β complexes, which underline the pivotal role of NKA-β in trafficking and the correct membrane insertion of the NKA α-subunit [19,39].

We first investigated whether hypercapnia affects protein levels of NKA-β focusing on the abovementioned glycosylation states in AEC. We observed a marked time-dependent and dose-dependent increase in the ER-resident high mannose glycosylation forms of NKA-β peaking at 12 h upon exposure to elevated CO2 levels. Of note, these effects were independent of extracellular acidosis, which suggests that the observed effects were directly mediated by CO2. This increase in ER-resident NKA-β forms was paralleled by a decreased PM abundance and activity of NKA-α:β. Moreover, while the levels of NKA-β were markedly increased in the ER where there was an enhanced interaction of NKA-β with the ER-resident chaperons, calnexin and BiP was evident and we observed a significant decrease in Golgi-resident forms of NKA. This is in line with previous reports that show that mutations in the NKA-α:β interaction regions [19], lipid peroxidation, and activation of oxidative stress by cadmium [20], removal of NKA-β glycosylation sites [20], and overexpression of unassembled NKA β-subunits [40] result in ER retention and increased binding of NKA-β to ER-resident chaperones [20,40]. Since the catalytic NKA-α-subunit cannot leave the ER (and, thus, reach the cellular surface) without being assembled with the regulatory NKA-β-subunit at a 1:1 stoichiometric ratio [19,20] and hypercapnia disturbs normal folding of NKA-β in the ER, which prevents its subsequent further maturation, collectively, our data suggest that elevated CO2 levels disrupt the formation of the NKA-α:β complex in the ER, which impairs its delivery to the PM.

Changes in the ER environment or modifications of the protein structure may result in misfolding and protein retention in the ER [41]. It has been shown that high Ca2<sup>+</sup> and sufficient ATP levels as well as a tightly regulated oxidizing environment are pivotal determinants of proper ER protein folding, glycosylating, and interaction with chaperones [42,43]. Thus, we hypothesized that the ER retention of NKA-β was due to changes in the ER environment. Of note, we establish, in this case, that elevated CO2 alters the oxidizing environment of the ER and leads to protein carbonylation. Irreversible

attachment of carbonyl groups to proteins is characteristic for a variety of oxidative pathways [44,45] that may promote protein misfolding and lead to various disease states [46,47]. The NKA-β contains a specific cysteine residue (Cys46) in its transmembrane domain that has been described to be highly susceptible to oxidative stress induced by glutathionylation [48], which possibly explains our findings regarding the enhanced ER retention of NKA-β. In addition, carbonylation of ER-resident chaperons may disrupt normal protein folding [46,49]. Whether hypercapnia leads to carbonylation of NKA-β at Cys46 and/or drives carbonylation of the above-mentioned ER-resident chaperons will need to be further investigated in future research. Furthermore, a common cellular response to transcriptional or translation errors is activation of protein carbonylation that "tags" peptides and, thereby, promotes their degradation [50]. Whether hypercapnia induces degradation of carbonylated proteins and NKA-β is currently being investigated in our laboratory.

In line with our previous findings in a different setting showing that elevated CO2 reduces ATP production by inhibiting IDH2 in the TCA cycle and leads to mitochondrial dysfunction [24], our current study shows that sustained hypercapnia markedly decreases intracellular ATP levels. As mentioned above, normal mitochondrial function and sufficient ATP supply of the ER are essential for normal protein folding [23]. We next aimed to study the potential connection between NKA-β oxidation and ER retention upon hypercapnia and the metabolic status of the cell. To overcome the elevated CO2-induced suppression of the TCA cycle, we treated AEC with α-KG (a TCA intermediate metabolite) that we have previously shown to rescue ATP production and cellular proliferation in CO2 exposed cells [24]. We observed that hypercapnia-induced ER oxidation is significantly reduced after α-KG treatment, which leads to decreased ER-retained NKA-β and increased NKA cell surface abundance and function. Together with our previously published observation, we propose that α-KG treatment may rescue IDH2 activity and, thus, ATP production. Therefore, this decreases ER oxidation and stabilizes normal NKA trafficking. Of note, recent findings suggest that the mechanism by which α-KG operates is not limited to the TCA cycle but may involve various other metabolic and cellular signaling pathways regulating cellular energy supply donor, epigenetics, and prolyl hydroxylases activity [51–53]. Thus, elucidation of the exact mechanisms of α-KG action in the context of hypercapnia may need further studies.

Our study has some clear limitations. Although we were able to show that elevated CO2 levels induce ER retention of the NKA-β subunit for up to 24 h that may lead to decreased cell surface abundance and activity of the enzyme, it is also evident that, at later time-points (36–72 h of high CO2 exposure), elevated levels of NKA-β in the ER are not evident. Our preliminary data (not shown) suggest that this is due to degradation of ER-retained NKA-β. A further study addressing this point is currently a major focus of our laboratory. Additionally, while we demonstrate that, during hypercapnia, an increase in NKA-β is associated with a decrease of the levels of NKA-β in the Golgi apparatus and at the PM, to further strengthen our hypothesis, subsequent studies will be necessary to directly assess trafficking events of both NKA-β and NKA-α from the ER to Golgi and PM. Furthermore, although we demonstrate that ER retention of NKA-β is due to the increased oxidation, we do not know what the primary cause of these reactions is. In addition, despite showing that the formation of carbonyl groups plays a central role in the CO2-induced protein oxidation reaction, the potential involvement of other oxidation reactions such as thiol oxidation, glutathionylation, or aromatic hydroxylation has not been investigated. Lastly, it will be important that future studies assess the effects of hypercapnia in more complex systems, such as precision-cut lung slices or human alveolar organoids to further enhance the translational relevance of these findings.

Taken together, our study shows for the first time that sustained hypercapnia decreases NKA cell surface abundance and function by inducing ER retention of its β-subunit via oxidative modification, which prevents its assembly with NKA-α. Furthermore, treatment with α-KG that increases ATP production and rescues the high CO2-induced mitochondrial dysfunction attenuates protein oxidation in the ER and, thus, prevents ER retention of the NKA-β and enhances PM abundance and activity of NKA. Since NKA is both a key driver of alveolar fluid clearance and also a central cell adhesion

molecule in the alveolar epithelium that is pivotal for prevention of alveolar edema formation, these novel findings may lead to novel therapeutic options that improve resolution of alveolar edema and restores alveolar epithelial barrier function in patients with ARDS and hypercapnia.

### **4. Materials and Methods**
