The KL6F-RPB1 pathway – setting the transcriptional stage for failed tubular recovery

Episodes of acute kidney injury (AKI) that are followed by incomplete or maladaptive repair contribute to the development of chronic kidney disease (CKD) or AKI to CKD transition. When proximal tubule cells are injured they dedifferentiate, reenter the cell cycle to proliferate and replace and repair lost tubular structures, followed by re-differentiation to restore the original phenotype. In contrast, a maladaptive repair is associated with G2/M cell cycle arrest and sustained dedifferentiation, leading to cell senescence and a prosecretory and profibrotic phenotype^1–3. The zinc-finger transcription factor Krüppel-like factor 6 (KLF6) is an early injury response gene, and its link to branched chain amino acid (BCAA) catabolism has recently been introduced as a relevant process in maladaptive proximal tubule repair. Piret and colleagues had shown that proximal tubule specific KLF6–mediated suppression of branched chain amino acid (BCAA) catabolism, possibly through consequences of excessive BCAA accumulation, contributes to kidney injury and the transition to CKD in mouse models, including in the aristolochic acid (AAI)-induced nephrotoxic model, a murine model of Balkan nephropathy⁴ (Fig. 1). Suggesting potential relevance in the human setting, individuals with CKD showed inverse correlations between KLF6 expression and both BCAA catabolic gene expression and kidney function⁴. Moreover, the group provided evidence that activation of BCAA catabolism can protect against acute kidney injury induced by AAI⁵, and others showed that gene knockout of Slc6a19, the amino acid transporter primarily responsible for the tubular uptake of filtered BCAA, did not affect the initial extent of AAI-induced tubular injury but improved the subsequent kidney recovery⁶.

Fig. 1: The DNA damage-induced KLF6-RPB1 pathway causes a transcriptional switch to sustain tubular dedifferentiation and promote failed recovery and fibrosis.

In the murine model of Balkan nephropathy, the naturally occurring toxic aristolochic acid I (AAI) is systemically injected and primarily taken up into proximal tubule cells by basolateral organic anion transporters OAT1 and OAT3. This is followed by DNA adduct formation, DNA injury response, and direct cytotoxic effects of AAI to cause cell injury and reactive upregulation of KLF6 expression. KLF6 facilitates failed injury recovery by inhibiting catabolism of branched chain amino acids (BCAA)⁴ or, as shown now⁷, by inducing the expression of RNA polymerase II (RNAPII) subunit 2 (Polr2a/RPB1), causing a possible persistent stalling of RNAPII. This seems to reprogram transcription and thereby trap proximal tubular cells with DNA damage in a dedifferentiated state that enhances fibrosis. Suggesting potential for translation, individuals with diabetic kidney disease also show an association between kidney DNA damage and RPB1 expression. See text for more details.

In the current issue, Piret and colleagues introduce a second pathway by which KLF6 has the potential to promote failed tubular recovery⁷. During an effort to better understand the early transcriptional mechanisms by which KLF6 induction exacerbates proximal tubular injury and eventual fibrosis, the group identified Polr2a, encoding RPB1, the largest subunit of RNA polymerase II (RNAPII), as a potential downstream target responsible for a DNA damage-induced transcriptional switch that promotes sustained tubular dedifferentiation and failed repair (Fig. 1). One day after a single injection of AAI, the authors used snRNA-sequencing to identify an injured proximal tubule cluster in which pathways and GO terms found in normal proximal tubule were mostly absent whereas Klf6 expression was increased. The cluster was characterized by enhanced mRNA processing and splicing and translation pathways, and the highest upregulated gene in this cluster was Polr2a. A proximal tubule cluster expressing high Klf6 and Polr2a was also identified when DNA damage and AKI were induced by cisplatin. Additional studies showed that KLF6 can drive POLR2A expression (Fig. 1), and that with repeated AAI injections, proximal tubule cells with RPB1-positive nuclei were preferentially retained relative to overall proximal tubule loss.

RNAPII is involved in the transcription of all protein-encoding genes and non-coding RNA genes, and POLR2A is essential for cell survival. Its role in transcription regulation is the recruitment of transcription factors to the promoter region of target genes thereby promoting transcription. Moreover, regulation of the RNAPII pool is integral to the DNA damage response, which can be triggered by both internal cellular processes, like spontaneous DNA base modifications or oxidative damage from metabolism, and external factors, like radiation or chemical agents. Persistent upregulation of RPB1 expression and RNAPII activity at sites of DNA damage (RNAPII stalling) had previously been linked to a transcriptional switch from medium (>30kb – <100kb) and long (>100kb) genes to short genes (<30kb genomic span)⁸. In this regard, Piret and colleagues found that the injured cluster, showing tubular RPB1 upregulation, was associated with a transcriptional shift to the expression of genes with reduced length, with the downregulated medium/long genes including pathways related to normal proximal tubule cell function, including transport, metabolism, and cellular adhesion.

Supporting a link between POLR2A/RPB1 and maladaptive repair, tubular positivity for RPB1 was associated with dedifferentiated proximal tubules lacking brush border. Moreover, knockdown of POLR2A in AAI-injured HK-2 cells reduced the level of dedifferentiation, DNA damage, and G2/M cell cycle arrest, and attenuated the transcriptional switch from long genes to short genes, and this was associated with reduced inflammatory and fibrotic gene expression. Remarkably, knockdown of POLR2A increased cell death in injured cells; this may relate to the beneficial effects of apoptosis and removal of severely injured cells versus the deleterious consequences of keeping senescent and dedifferentiated cells around. The authors point to potential similarities with effects of senolytics, i.e., the use of drugs that selectively eliminate senescent cells that donť replicate, but can still release harmful substances and that have been linked to aging and injury-related inflammation and tissue dysfunction.

In an effort to probe for potential clinical translation, the authors made use of human kidney data deposited in the Nephroseq database or analyzed data from KPMP. The analyses revealed high POLR2A/RPB1 expression in kidneys rejected after transplantation, DNA-damage induced CKD, DNA-damage associated diabetic kidney disease as well as in dedifferentiated proximal tubule cells. Moreover, the authors’ analyses provided evidence that both adaptive proximal tubule and degenerative proximal tubule clusters increased the expression of KLF6 compared to normal proximal tubule clusters, but only the degenerative cluster showed the transcriptional switch from medium or long genes to short genes, which the authors had observed in the injured cluster having POLR2A/RPB1 upregulation. Finally, some evidence was provided for a KLF6-POLR2A axis by showing that in acute kidney rejection samples, but not in the absence of rejection, the expression of KLF6 and POLR2A was positively correlated.

The work invites thoughts on the therapeutic targeting of a dysregulated KLF6-POLR2A axis. The physiological role of KLF6 in the response to toxic or oxidative/ischemic insults may be the induction of cellular senescence, which then allows repair of the resulting DNA damage⁴. A potential therapeutic targeting of KLF6 may also be complicated by the observation that the outcome can be strongly dependent on the cellular context, e.g., knockdown of KLF6 in podocytes by ~50% on protein level increased the susceptibility in a murine model of focal segmental glomerulosclerosis⁹. It will be important to gain a more comprehensive understanding of KLF6 as a potential therapeutic target and of the regulators and determinants of RPB1 activity, the nature of RNAPII stalling, and their implications for the transcriptional shift in the setting of kidney injury, CKD, and AKI to CKD transition. Also, do other kidney protective strategies interfere with this pathway, e.g., inhibitors of the sodium-glucose cotransporter SGLT2, which reduce the risk of AKI, have protective effects in non-diabetic and diabetic CKD as well as in AAI-induced nephropathy, and also share mechanistic similarities with SLC6A19 and its inhibition¹⁰.

Overall Piret and colleagues provide first and convincing evidence that proximal tubular injury-induced KLF6 can positively regulate RPB1. Moreover, the exciting concept is introduced that activation of the KLF6-RPB1 pathway causes a global transcriptional switch that upregulates short genes but downregulates medium/long genes typically involved in physiological functions, and that this switch is a potential key mechanism driving tubular dedifferentiation and degeneration, at least under conditions of DNA damage-induced tubular injury. Notably and related to transcriptional resetting, recent studies showed that weeks to months after mild-to-moderate ischemia-reperfusion injury-induced AKI, even proximal tubules with adaptive repair showed reduced expression of genes encoding essential functions like transmembrane transport proteins¹. Clearly, the seminal work by Piret and colleagues set the stage to probe the relevance of the proposed KLF6-RPB1 pathway and the transcriptional switch in other forms of AKI and CKD, including those with less prominent DNA damage.