Kidney-Heart Crosstalk After Myocardial Infarction

Evidence

Heart and kidney dysfunction often coexist after myocardial infarction, but coexistence alone does not reveal which organ is influencing the other or when. Two independent mouse studies approach that problem from opposite directions. A study first published July 24 asks how pre-existing chronic kidney disease, or CKD, changes the heart's response to infarction. A separate full-text study begins with myocardial infarction and tracks how kidney tissue changes over five and 28 days. Together, they provide complementary evidence about timing and direction, not proof of one continuous molecular circuit.

The fresh study used adenine-induced CKD in mice and compared cardiac remodeling after myocardial infarction with non-CKD controls. CKD mice developed stronger cardiac dysfunction even though the reported analyses found no differences in infarct size, cardiomyocyte apoptosis, or myofibroblast content. They instead had more circulating myeloid cells and greater neutrophil infiltration in the heart. A combination of RNA sequencing, kinome profiling, western blotting, and mass spectrometry linked the CKD condition to greater cardiac oxidative stress, increased S100A8/A9 in blood and heart, p38 MAP-kinase activation, and NR4A1 phosphorylation after infarction.

The same study tested part of that inflammatory signal outside the intact animal. S100A8/A9 acutely impaired calcium flux and sarcomere shortening in cardiomyocytes ex vivo. Its multi-omics analysis also found impaired glycolysis, a reduced glycerol-3-phosphate shuttle, and lower coenzyme A availability in CKD mice after infarction. Those metabolic changes were associated with poorer cardiac performance, while the investigators did not observe intrinsic mitochondrial defects. Human data provided limited corroboration: single-nucleus RNA sequencing detected increased myeloid-cell-derived S100A8/A9 expression in infarcted heart tissue, and patients with CKD had higher post-infarction S100A8/A9 levels than patients without kidney dysfunction. The ingested abstract does not establish that this signal caused human outcomes.

The independent full-text study used permanent coronary ligation in male C57BL/6J mice. At day five, RNA sequencing included infarct-zone hearts and kidneys from nine infarcted mice and corresponding tissues from eight sham-operated mice. The cardiac infarct zone had more than 11,000 differentially expressed genes, compared with only 18 in the kidney. A transcript-based CellChat analysis inferred 131 possible heart-to-kidney interactions after restricting the analysis to upregulated cardiac ligands. Periostin, encoded by Postn, and osteopontin, encoded by Spp1, were the highest-probability candidates. These are computationally inferred interactions; the study did not trace the proteins from heart to kidney.

The kidney response was more extensive later. At 28 days, RNA sequencing of kidneys from seven infarcted mice identified 86 differentially expressed genes: 41 upregulated and 45 downregulated. Histology in four mice per group showed more renal cortical fibrosis than in sham controls. Six of the ten most significantly enriched gene-ontology terms involved lipid metabolism, with reduced expression across genes assigned to intracellular lipolysis and lipid oxidation. Four kidney genes classified as secreted ligands were upregulated: Cck, Slitrk6, Gdf15, and Psap. The investigators prioritized proposed Slitrk6-Ptprs and Gdf15-Tgfbr2 kidney-to-heart interactions because the corresponding receptors were expressed in heart tissue.

One candidate received a functional follow-up. Gdf15 was increased in kidney and Tgfbr2 in the cardiac infarct zone at day 28. Human cardiac fibroblasts exposed to 100 nanograms per milliliter of GDF-15 for 24 hours increased expression of TGF-beta, COL1A1, and CTGF, three fibrosis-associated genes. This experiment shows that GDF-15 can change gene expression in isolated fibroblasts under the tested conditions. It does not establish the kidney as the source of circulating GDF-15, show that the inferred receptor mediates the response, or demonstrate that blocking this candidate alters disease in vivo.

Analysis — Direction Changes With Time

The cross-study pattern is a possible time-dependent feedback system, and that interpretation is analysis rather than a joint result. When CKD exists before infarction, the fresh study finds an early cardiac environment marked by myeloid activation, oxidative stress, altered fuel handling, and poorer function without a larger infarct. When infarction is the starting event, the full-text study finds a highly asymmetric day-five response—extensive cardiac transcriptional change but little renal transcriptional change—followed by renal fibrosis, altered lipid-metabolism programs, and candidate kidney-to-heart signals at day 28. The convergence lies at the process level: inflammation, metabolic adaptation, and fibrosis may connect the organs across different phases. The named molecules do not yet form a verified chain. S100A8/A9, periostin, osteopontin, and GDF-15 are also active in tissues beyond heart and kidney. A decisive experiment would combine CKD and infarction in a factorial, time-resolved design; measure proteins and organ function in both directions; trace the tissue source of candidate signals; and test whether selective interruption changes cardiac and renal outcomes. Until then, a bidirectional loop is a useful hypothesis, not an established mechanism.

Limitations

Both studies rely heavily on mice, and the fresh paper was available to this briefing only as a PubMed abstract. Its sample sizes, exact timing, complete statistical estimates, adenine-model details, and potential exclusions cannot be evaluated from the ingested record. The human S100A8/A9 analyses add relevance but remain observational or transcript-based; they do not reproduce the mouse intervention or demonstrate causality in patients.

The full-text study used permanent coronary ligation, a non-reperfused model that differs from many treated human infarctions. It included only young male mice. The day-five and day-28 analyses used modest, partly different sample sets, and the sham kidneys used for the 28-day comparison came from the earlier time point. CellChat infers communication from ligand and receptor transcripts; it does not prove protein secretion, circulation, receptor binding, tissue of origin, or functional organ-to-organ transfer. Cardiac RNA sequencing was unavailable at day 28, preventing a complete late bidirectional analysis. The GDF-15 experiment used isolated fibroblasts at one reported concentration and duration, and fibrosis-associated gene expression is not the same as organ fibrosis or clinical benefit. Finally, the studies did not use the same CKD model, animals, time points, or candidate pathways. Their synthesis defines testable connections but cannot determine whether any proposed signal is a therapeutic target.