The 7-member mammalian FXYD protein family is expressed in a tissue-specific manner (1). Its members associate closely with the Na⁺/K⁺-ATPase, and most of them protect the Na⁺/K⁺-ATPase β1 subunit against glutathionylation, an oxidative modification in which the cytosolic tri-peptide glutathione (GSH) binds to a free SH group of a cysteine (Cys) residue in a protein. Glutathionylation of the β1 subunit compromises the structural integrity of the α1/β1 heterodimer of Na⁺/K⁺-ATPase and inhibits its function. Countering this, FXYD proteins can stabilize Na⁺/K⁺-ATPase structure and function (2).

A specific Cys residue bracketed by basic amino acids and conserved across most members of the FXYD protein family is critical for a FXYD protein’s protection against β1 subunit glutathionylation. The cellular expression of FXYD proteins that have the specific Cys residue is protective against oxidative stress-induced β1 subunit glutathionylation and Na⁺/K⁺-ATPase inhibition, while mutating the residue eliminates the protective effect of the expressed protein (2). The critical role of Cys residues is also evident from the effects of exposing cardiac myocytes to recombinant FXYD proteins with or without the Cys residue (2).

A brief (15 min) exposure of myocytes to recombinant FXYD3 protein displaces the native FXYD1 protein. Both the exogenous and the native protein have the Cys residue critical for the protective effect against β1 subunit glutathionylation, and the additional exogenous supply of FXYD3 counters β1 subunit glutathionylation induced by oxidative stress. In parallel, it counters sarcolemmal Na⁺/K⁺-ATPase inhibition. Exogenous recombinant FXYD1 has similar effects on Na⁺/K⁺-ATPase function. However, while a recombinant FXYD3 protein with Cys residues mutated to serine also displaces native FXYD1 in myocytes, it is not protective against β1 subunit glutathionylation or Na⁺/K⁺-ATPase inhibition (2).

Only the transmembrane domain and the extracellular PFXYD motif near the membrane leaflet have known bonds to α- and β Na⁺/K⁺-ATPase subunits (3, 4). Other than the PFXYD motif and the motif that includes the Cys residue critical for countering β1 subunit glutathionylation, extracellular and cytoplasmic domains are poorly conserved across the FXYD protein family (5). Since they are not expected to affect β1 subunit glutathionylation, we designed 36 amino acid peptide analogues of the Cys-preserved FXYD3 protein and its Cys-mutated (Cys →Ser) derivative without these domains. We refer to these peptides as FXYD3-pep CKCK and FXYD3-pep SKSK.

FXYD3 is often highly expressed in cancers, particularly in those of the pancreas (6), prostate (7), and breast (8). Suppression of FXYD3 with siRNA in cultured human breast cancer cells that overexpress FXYD3 augments cytotoxicity of doxorubicin (Dox) (9). Dox induces oxidative stress (10), and the augmentation of Dox cytotoxicity with downregulation of FXYD3 was associated with an increase in β1 Na⁺/K⁺-ATPase subunit glutathionylation (9). If augmentation of cytotoxicity with FXYD3 downregulation depends on the FXYD Cys residue that is critical for countering β1 subunit glutathionylation, Cys-mutated FXYD3-pep SKSK but not FXYD3-pep CKCK should reproduce the augmentation. We have examined the effects of FXYD3-pep SKSK or FXYD3-pep CKCK on Dox-induced cytotoxicity in breast and pancreatic cancer cells that do or do not overexpress FXYD3.

Similar to the distribution of overexpressed WT FXYD3 in resected pancreatic (6) and breast cancers (8), we detected TRITC-tagged FXYD3-pep SKSK intracellularly, predominately in a perinuclear distribution. Immunofluorescence signals for WT FXYD3 in MCF-7 cells partly overlap with signals for Na⁺/K⁺-ATPase α subunits (8), and a mostly perinuclear distribution of the β1 Na⁺/K⁺-ATPase subunits in cancer cells is also reported with the development of some colorectal cancers (22). We are not aware of studies reporting overexpression of the intact α/β/FXYD3 Na⁺/K⁺-ATPase molecular complexes in the intracellular compartment in cancers. However, all its constituents are reported intracellularly and in a shared distribution.

FXYD3 in MCF-7 cells is expressed in the plasmalemma as well as in intracellular membrane compartments (8). The overexpressed FXYD3 is confined to the perinuclear region because the stem cell-related transcription factor SOX9 promotes FXYD3 transcription, and the synthesized FXYD3 protein in turn confines SOX9 to the nucleus in a regulatory feedback loop for FXYD3 amplification (23). With experimentally induced synthesis of FXYD3 in Chinese Hamster Ovary (CHO) cells, detectable FXYD3 becomes distributed in the nuclear envelope and the endoplasmic reticulum (ER) but not in the plasmalemma (24). Here, it is unlikely that FXYD3-pep SKSK sensitized the cells to Dox because of a dependence on a high catalytic plasmalemmal Na⁺/K⁺-ATPase activity. The high expression of WT FXYD3 in parallel with the expression of α1 subunits in MCF-7 cells relative to MDA-MB-468 cells was not reflected by different maximally activated I_p. Furthermore, while MDA-MB-468 cells did not overexpress WT FXYD3, they did express the protein at levels comparable to those for non-transformed human breast cells, yet FXYD3-pep SKSK did not sensitize MDA-MB-468 cells to Dox.

Facilitated by bonds from FXYD proteins to α and β subunits seen in the crystal structure (4), the primary overexpression of FXYD3 in cancer cells might lead to an intracellular accumulation of FXYD3/α/β Na⁺/K⁺-ATPase complexes. That such assembly plausibly occurs is indicated by the spontaneous formation of FXYD/α/β Na⁺/K⁺-ATPase complexes with the in vitro exposure of human α1/β1 subunits to a molar excess of FXYD1, FXYD2, or FXYD3. FXYD/α/β Na⁺/K⁺-ATPase complexes also assemble when rat FXYD1, FXYD2, or FXYD4 is co-expressed with rat α1 subunits in HeLa cells, indicating that the effects shown for purified human proteins are biologically relevant in intact cells (25).

That an excess of FXYD proteins stabilizes the Na⁺/K⁺-ATPase structure and function is supported by how recombinant full-length WT FXYD3 prevents a decrease in α1/β1 Co-Ip detected in cardiac myocyte lysate and prevents a decrease of I_p measured in intact voltage-clamped myocytes during oxidative stress (2). However, while WT- and Cys-mutated recombinant proteins in equimolar concentrations similarly displace the FXYD1 native to myocytes, only the WT FXYD3 has these preventative effects. The proteins share amino acids that have bonds to α- and β subunits, and presence of Cys residues rather than displacement of WT proteins per se accounts for their effects on integrity and function of Na⁺/K⁺-ATPase.

Even with the overexpression of WT FXYD3 in BxPC-3 cells, FXYD3-pep CKCK had a protective effect against Dox-induced cytotoxicity while FXYD3-pep SKSK augmented it. The two peptides share with their full-length equivalent proteins all domains with bonds to the α/β Na⁺/K⁺-ATPase complex. However, Cys residues, including the specific residue that mutational studies indicate confers susceptibility of FXYD proteins to glutathionylation, had been replaced by Ser in FXYD3-pep SKSK ( Figure 2 ) and it is a FXYD protein’s susceptibility to glutathionylation that counters glutathionylation of the β1 subunit (2). Effects of the peptides here indicate that augmentation of cytotoxicity with siRNA-induced downregulation of FXYD3 (9) reflects the role of the Cys residue that is critical for countering β1 subunit glutathionylation.

With selective inhibition of plasmalemmal Na⁺/K⁺-ATPase at ouabain concentrations >50 nM, there was no dose-dependent increase in ouabain’s augmentation of Dox’s cytotoxicity; that is, ouabain did not reproduce the large dose-dependent augmentation of Dox’s cytotoxicity that FXYD3-pep SKSK induced. Therefore, while glutathionylation of the β1 subunit inhibits Na⁺/K⁺-ATPase (26), selective inhibition of plasmalemmal Na⁺/K⁺-ATPase ion transport is unlikely to account for the pattern of sensitization to Dox seen with FXYD3-pep SKSK.

In addition to mediating ion transport, Na⁺/K⁺-ATPase facilitates cell adhesion via intercellular β1–β1 subunit bridges, and exposure to ouabain in nanomolar concentrations augments such cell adhesion in CHO fibroblasts overexpressing β1 subunits (27). Ouabain also prevents β1 subunit glutathionylation induced by oxidative stress, shown for intact cardiac rabbit myocytes and isolated pig kidney Na⁺/K⁺-ATPase-enriched membrane fragments (28). β1 subunit glutathionylation and α1/β1 subunit Co-Ip are inversely related (28), and molecular dynamics simulations suggest that this might reflect the structural disruption that a GSH adduct causes between transmembrane domains of the glutathionylated heterodimer, combined with a disruption of the cytosolic membrane leaflet necessary for access of the cytosolic GSH (29).

FXYD3 from intracellular α/β/FXYD3 Na⁺/K⁺–ATPase complexes in cells overexpressing FXYD3 might be a source that counters glutathionylation of β1 subunits in plasmalemmal Na⁺/K⁺-ATPases. The effect such a source of FXYD proteins can have is indicated by the reversal of decreased myocyte I_p induced by oxidative stress when recombinant FXYD1 or FXYD3 proteins are included in patch pipette solutions perfusing the intracellular compartment. A decrease in the β1 subunit’s glutathionylation in parallel with an increase in its Co-Ip with α1 subunits in lysates of myocytes supported FXYD protein-dependent protection of α1/β1 heterodimer integrity accounting for effects on ion transport (2).

Since intercellular β1–β1 subunit bridges extend to the cytoskeleton through the binding of β1 to α1 subunits, that in turn have bonds to the cytoskeleton (30), disruption of the α1/β1 heterodimer with glutathionylation is expected to disrupt cell adhesion. Conversely, overexpressed FXYD3 might counter such disruption. Consistent with this, FXYD3 knockdown impairs growth and disperses cells in MCF-7 tumors expressed in mice (23), and in human colon adenocarcinoma Caco-2 cells, FXYD3 knockdown decreases the transepithelial electrical resistance of confluent cell layers consistent with decreased intercellular adhesion (31).

In contrast to FXYD3 promoting cell adhesion, transfection of a mouse kidney cell line with FXYD5 reduces adhesion, as indicated by a decrease in transepithelial electrical resistance (32). However, this is expected if disruption of the α1/β1 heterodimer with glutathionylation disrupts cell adhesion because FXYD5 has a Gln residue (33) at the site where a Cys residue bracketed by basic amino acids is obligatory for a FXYD protein to counter glutathionylation of the β1 subunit. A single, selective mutation of the residue (Cys→Ala) eliminates an FXYD protein’s effect to counter β1 subunit glutathionylation (2).

Disruption of cell adhesion with FXYD5 transfection was attributed to a decrease in glycosylation of the β1 subunit’s extracellular domain (32), and inhibiting glycosylation can augment the cytotoxicity of Dox (34). FXYD5 is functionally a WT protein equivalent of FXYD3-pep SKSK in terms of their expected effects on glutathionylation of β1 subunits, and decreased glycosylation or increased glutathionylation of the subunits is not mutually exclusive in potential effects on subunit-dependent intercellular adhesion. The complexity of β1 subunits Na⁺/K⁺-ATPase N-glycans is also important for effects of the subunit on intercellular adhesion (35). However, an effect of FXYD3-pep SKSK on this cannot be invoked at present.

The FXYD peptide derivatives here augment key functions of the WT proteins they are derived from or, with a Cys-mutated modification, eliminate the function by displacing the WT protein from its only known binding partner, Na⁺/K⁺-ATPase. Exposure over minutes–hours avoids confounding cellular adjustments with transfection to overexpress FXYD proteins or silencing to reduce their abundance over days–weeks, and, in contrast to the experimental use of recombinant FXYD proteins (2), purified peptides are easy and inexpensive to make.

The FXYD3 peptide derivatives we report here are useful tools for exploring the role of overexpressed FXYD3 in cancer and its potential as a treatment target. With further modifications, FXYD3-pep SKSK might be developed into a treatment sensitizer. However, one caveat must be kept in mind. The peptide is, at least in part, expected to be a functional equivalent of FXYD5, and FXYD5 expression increases the metastatic potential in a mouse breast cancer model. This is associated with a decrease in plasmalemmal Na⁺/K⁺-ATPase β1 subunit expression and a decrease in intercellular adhesion (36). FXYD5 expression also appears to increase the metastatic potential of human cancers (37).

While the focus here is on FXYD3-pep SKSK, the protective effect of FXYD3-pep CKCK against Dox-induced cytotoxicity reflects properties that also may be useful. The peptide retains the Cys residue and, in turn, an ability to protect integrity and function of the α1/β1 Na⁺/K⁺-ATPase heterodimer (2). A potential use is exemplified by treatment of acute respiratory distress syndromes, including SARS-CoV-2 infection. Reduced Na⁺ export from alveolar cells mediated by Na⁺/K⁺-ATPase and compromised β1 subunit-dependent integrity of the alveolar epithelial barrier are implicated in the pathophysiology of pulmonary alveolar edema (38, 39). A peptide drug based on FXYD3-pep CKCK and modified to resist in vivo breakdown might offer an approach to treatment.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding authors.

C-CL contributed to the design and implementation of the research, to the analysis of the results, and to the writing of the manuscript. YK, RT, AG, AW, and EH performed the experiments. FC, RB, and HR aided in interpreting the results and worked on the manuscript. All authors contributed to the article and approved the submitted version.

The work is supported by a grant from the Avner Pancreatic Cancer Foundation, Heart Research Australia, and Ramsay Research and Teaching Fund (Australia). YK was supported by Research Training Program Stipend Scholarship (Australia) and a Ramsay research top up scholarship (Australia). AG was supported by a UTS Chancellors Postdoctoral Research Fellowship (Australia). EH was supported by Heart Research Australia.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.