Long-read single-nucleus sequencing exposes a layer of kidney injury biology missed by short reads
Acute kidney injury is usually discussed in terms of cell loss, tubular stress, and altered gene expression. This preprint argues that the story is broader than gene-level counts alone. By pairing single-nucleus long-read RNA sequencing with conventional single-cell and single-nucleus approaches, investigators mapped transcript diversity in healthy mouse kidney and in kidneys subjected to bilateral renal ischemia-reperfusion injury.
The central point is straightforward: short-read methods are strong at quantifying genes, but they struggle to identify which transcript isoforms are actually being used. In the kidney, where cell state changes can be rapid and compartment-specific, that limitation leaves out an important layer of regulation. Long reads preserve splice structure and transcript architecture, allowing investigators to distinguish protein-coding and non-protein-coding isoforms rather than collapsing them into a single gene-level signal.
Three healthy adult male C57BL/6J mice provided single-cell and single-nucleus long-read datasets, and another three adult male C57BL/6J mice underwent ischemia-reperfusion injury before kidney harvesting for single-nucleus long-read analysis. Across samples, the team generated 1.3 billion reads, with a median read length of 757 to 884 bases. Those are large datasets by any standard, and they matter because isoform detection depends heavily on sequencing depth and read structure. A shorter read can tell you a gene is active; it cannot always tell you which transcript form is carrying that activity.
Single-nucleus libraries behaved differently from single-cell libraries. They assigned a smaller proportion of reads to isoforms, which fits their larger fraction of intronic reads. Yet the nucleus-based datasets detected more distinct isoforms and greater isoform expression diversity. The explanation is biologically plausible: nuclei capture nascent and incompletely processed RNA, and they avoid much of the mitochondrial transcript background that can crowd out other signals in whole-cell preparations. In practical terms, that makes the nucleus an efficient compartment for isoform discovery, even when raw assignment rates look less favorable at first glance.
A single sentence captures the methodological surprise: the nuclear approach found more transcript forms even while assigning fewer reads per isoform.
Ischemia-reperfusion injury shifted isoform usage in proximal tubules and beyond
The injured-kidney analysis focused on how acute kidney injury reshaped transcript usage in specific cell populations. Proximal tubule cells showed downregulation of oxidative phosphorylation-associated genes at the gene-expression level, and the same pathway also shifted at the isoform level. This pairing matters biologically. Proximal tubules are among the most energy-demanding cells in the kidney, relying heavily on oxidative phosphorylation to maintain transport function and ionic homeostasis. When ischemia deprives tissue of oxygen and reperfusion adds oxidative stress, the cell’s response is not just reduced expression of energy genes; it can also involve changing which versions of those genes are transcribed.
Isoform switching can alter protein domains, regulatory motifs, localization signals, or transcript stability. Even without a direct protein assay in hand, the possibility is clinically and mechanistically important. Two transcripts from the same gene can have different consequences for mitochondrial maintenance, stress adaptation, or recovery after ischemic insult. A gene-level readout can flatten those differences into a single number and miss the underlying reprogramming.
Beyond the oxidative phosphorylation pathway, two genes stood out as showing prominent shifts in isoform usage after injury: Fut9 and Prrc2b. The abstract does not provide additional functional annotation for those shifts, so the data should be read as transcriptomic signals rather than mechanistic proof. Even so, their appearance alongside energy-metabolism genes suggests that ischemia-reperfusion injury drives a broader remodeling of RNA processing than standard differential expression workflows usually reveal.
The finding is not that one pathway turned on and another turned off. The more interesting observation is that the kidney appears to rewrite the transcript structure of stress-responsive genes at the same time it changes their abundance. In a metabolically vulnerable tissue like the proximal tubule, that kind of RNA-level remodeling may help determine whether a cell adapts, stalls, or dies after ischemia.
What the nuclear long-read approach adds to kidney research
This preprint positions single-nucleus long-read RNA sequencing as a way to extract information that conventional short-read analysis leaves inaccessible. The advantage is especially clear in kidney injury, where the same cell type may express multiple isoforms of a gene depending on oxygen tension, mitochondrial stress, and injury phase. By preserving transcript structure, long-read sequencing can identify isoform diversity that would otherwise be averaged away.
The approach also brings a set of technical tradeoffs. Nuclear libraries are enriched for intronic material, which reduces the fraction of reads that can be confidently assigned to mature isoforms. At the same time, that same nuclear content appears to widen the window onto transcript diversity. For studies of injured tissue, where cell dissociation can distort vulnerable populations, the nucleus may also offer a more stable substrate than whole-cell preparation.
Because this is a preprint, the findings should be treated as early evidence rather than established practice-changing data. The sample size was small, limited to six male mice in total, and the work was performed in a controlled experimental model rather than in human kidney disease. The abstract also does not report validation at the protein level, functional perturbation experiments, or outcome data after recovery from injury. Those gaps leave open the question of which isoform shifts are causal, compensatory, or simply epiphenomena of ischemic stress.
Even with those limits, the study makes a clear case for transcript isoform analysis as more than a technical curiosity. In kidney injury, transcript choice itself appears to change. Gene counts alone no longer tell the full story.