Transcriptomic signals separate CKD from ESRD at the molecular level
Progression from chronic kidney disease to end-stage renal disease is not only a clinical trajectory defined by falling estimated glomerular filtration rate and rising treatment burden. It is also a biological transition in which injured kidney tissue and systemic physiology appear to reorganize around hypoxia, mitochondrial stress, inflammatory activation, and altered oxygen handling. This preprint used integrated transcriptomic analysis to look for stage-associated markers that might help distinguish CKD from ESRD more precisely than routine laboratory measures alone.
The study combined two discovery datasets comparing CKD or ESRD with healthy controls, then layered differential expression analysis with weighted gene co-expression network analysis. Candidate genes were narrowed further by intersecting differentially expressed genes with key co-expression modules and then ranking proteins in a protein-protein interaction network using Degree, Closeness, and Maximal Clique Centrality algorithms. Independent GEO validation datasets, quantitative real-time PCR, and western blotting in clinical blood samples were used to check whether the same signals held up outside the discovery phase.
One pattern stood out immediately. EPB42 moved in opposite directions across disease stages.
EPB42 changed direction between CKD and ESRD
Among the ten hub genes identified, EPB42, KLF1, and H3F3A were prioritized for additional evaluation. EPB42 showed decreased expression in CKD but increased expression in ESRD, a bidirectional pattern that may carry more staging value than a single abnormality that simply rises or falls with renal dysfunction. In practical terms, that kind of reversal can be more informative than a monotonic biomarker because it suggests a molecular switch may occur as kidney injury deepens and systemic compensation fails.
EPB42 encodes erythrocyte membrane protein band 4.2, a protein tied to red blood cell structural integrity. Its involvement fits with the study’s enrichment findings, which pointed toward oxygen transport and related biology. CKD and ESRD are both states in which chronic anemia, disturbed red cell biology, and impaired oxygen delivery can amplify tissue hypoxia. A marker linked to erythrocyte architecture therefore makes biologic sense in a disease process where diminished oxygen transport and persistent ischemic stress help drive fibrosis and organ decline.
KLF1 behaved differently. It did not differ significantly between healthy controls and CKD, but it increased in ESRD. As a transcription factor central to erythroid maturation, KLF1 again ties the transcriptomic findings to red cell biology rather than to a purely kidney-local process. H3F3A, by contrast, was increased in CKD and rose further in ESRD, suggesting a cumulative disease-associated signal rather than a stage reversal.
Oxygen transport, mitochondria, and inflammation converged in the pathway analysis
Functional enrichment analysis linked the candidate genes to oxygen transport, mitochondrial function, muscle-associated processes, and inflammation. Those pathways are not incidental. Progressive kidney disease reduces oxygen delivery to renal tissue while also exposing cells to oxidative injury and metabolic failure, conditions that can damage mitochondria and push injured tissue toward maladaptive repair. In that setting, mitochondrial dysfunction and hypoxia are not separate stories; they reinforce one another.
The authors also pointed to possible involvement of hypoxia-associated oxidative injury and ferroptosis-related processes. Ferroptosis, an iron-dependent form of regulated cell death driven by lipid peroxidation, has attracted attention in kidney injury because renal tubular cells are especially vulnerable to oxidative stress and disturbed iron handling. If the observed gene signatures reflect that biology, they may help explain why some patients continue to progress despite standard measures aimed at blood pressure control, volume management, and proteinuria reduction. The molecular damage may already be embedded in the cellular stress response.
Why these markers may matter for staging and prognosis
The principal clinical appeal of this work lies in staging. CKD and ESRD are often discussed as points on a continuous decline, but bedside decisions frequently depend on whether a patient has crossed a biologic threshold that signals a fundamentally different risk state. A marker such as EPB42, if validated in larger and more diverse cohorts, could help capture that threshold by reflecting a change in the underlying molecular program rather than simply the severity of reduced kidney function.
The blood-based validation approach is also important. Tissue-level transcriptional changes can be informative but hard to translate clinically if they require invasive sampling. By checking qRT-PCR and western blotting in clinical blood samples, the study moves toward a form of biomarker development that could eventually be tested in routine practice. Even so, the preprint remains an exploratory analysis. It identifies candidates, not clinical assays.
Validation remains necessary before these genes can enter care pathways
The study design strengthens internal consistency by pairing bioinformatics with experimental confirmation, but several limits remain inherent to the approach described here. Discovery datasets drawn from public repositories can vary in platform, sampling conditions, and patient composition. Validation in GEO datasets helps, yet independent prospective cohorts will be needed to determine whether the gene signatures retain performance across different populations, disease causes, and treatment backgrounds.
The bedside question is not whether CKD progression has molecular correlates. It clearly does. The question is whether these transcripts can outperform existing staging tools, add prognostic value beyond routine nephrology markers, or help separate patients whose disease is accelerating from those whose trajectory is more stable. Until that is tested in larger cohorts with clear clinical endpoints, EPB42, KLF1, and H3F3A should be viewed as candidate biomarkers with biologic plausibility rather than ready-to-use diagnostic tools.
The most distinctive finding is still the direction change in EPB42. In a disease continuum usually described by gradual decline, a gene that flips from suppression in CKD to increased expression in ESRD suggests a sharper molecular boundary may exist than clinical staging alone can show.