Localization and characterization of proenkephalin-A as a potential biomarker for kidney disease in murine and human kidneys

Abstract Introduction Exact measurement of renal function is essential for the treatment of patients. Elevated serum-creatinine levels, while established, are influenced by other parameters and show a significant time-lag. This drives the search for novel biomarkers of renal function and injury. Beside Lipocalin-2 and kidney-injury-molecule-1 (KIM-1), the endogenous opioid precursor proenkephalin-A (Penk) has recently emerged as a promising marker for renal function. But the cellular origin and regulation of Penk outside the brain has not yet been investigated in depth. Materials and methods This study characterizes the cellular origin of Penk expression with high-resolution in situ hybridization in two models of renal fibrosis in mice and human tissue. Results Interstitial cells are the main expression site for renal Penk. This classifies Penk as biomarker for interstitial damage as opposed to tubular damage markers like Lipocalin-2 and KIM-1. Furthermore, our data indicate that renal Penk expression is not regulated by classical profibrotic pathways. Discussion This study characterizes changing Penk expression in the kidneys. The similarity of Penk expression across species gives rise to further investigations into the function of Penk in healthy and injured kidneys. Conclusion Penk is a promising biomarker for interstitial renal damage that warrants further studies to utilize its predictive potential. Clinical significance Knowledge of real-time renal function is essential for proper treatment of critically ill patients and in early diagnosis of acute kidney injury (AKI). Proenkephalin-A has been measured in a number of patient cohorts as a highly accurate and predictive biomarker of renal damage. The present study identifies Penk as a biomarker for interstitial damage in contrast to the tubular biomarkers such as Lipocalin-2 or KIM-1. Our data show that Penk is regulated independently of classical profibrotic or proinflammatory pathways, indicating it might be more robust against extra-renal influences. Data presented in this study provide fundamental information about cell type-specific localization and regulation of the potential new biomarker Penk across species as foundation for further research.


Introduction
Accurate estimation of renal function is essential for treating critically ill patients (Beunders et al. 2017, Caironi et al. 2018, Hartman et al. 2020). This is not only the case for patients diagnosed with renal failure, but also for cases of sepsis, severe cardiovascular events and diagnosing possible cases of acute kidney injury (AKI) after major surgery or organ transplantation (Kim et al. 2017, Ng et al. 2017, Caironi et al. 2018, Beunders et al. 2020, Luft 2021.
The classical method of monitoring serum creatinine (sCr) levels has several complications. Creatinine is not only filtered, but also actively secreted by the kidneys, thus it is not an ideal marker for renal function. Furthermore, sCr shows a significant delay between renal injury and rising plasma levels. sCr levels are also influenced by a variety of extra-renal factors leading to increasingly complex formulae to calculate glomerular filtration rate (GFR). Due to the kidneys great adaptability, only a strong decline in GFR is reflected by elevated sCr levels (Herget-Rosenthal et al. 2007, Beunders et al. 2017, De Oliveira et al. 2019, Albert et al. 2021, Luft 2021. This has driven the search for new biomarkers in recent years that show rapid changes in GFR and are independent of extra-renal factors (Paragas et al. 2011, Kim et al. 2017, Ng et al. 2017, Caironi et al. 2018, Beunders et al. 2020, Albert et al. 2021.
Among the candidates of newly identified makers for renal injury and predictors of GFR are proenkephalin-A (Penk), lipocalin-2 (Lcn2) and kidney-injury-molecule-1 (KIM-1). These markers have already found some implementation into clinical practice within the last five years and were studied in a number of patient cohorts (Viau et al. 2010, Ng et al. 2017, De Oliveira et al. 2019, Beunders et al. 2020, Hartman et al. 2020. They show considerable potential for early diagnosis of kidney injury and rapid changes in GFR. In addition, these new markers might have predictive advantages by correlating high biomarker measurements with increased risk for severe adverse events in the follow-up periods of these studies (Kim et al. 2017, Caironi et al. 2018, Hartman et al. 2020, Khorashadi et al. 2020, Albert et al. 2021. Of the proposed candidates, Penk has so far been reported as a good marker for predicting accurate renal function and long-term patient survival (Ng et al. 2017, Caironi et al. 2018, Emmens et al. 2019, Khorashadi et al. 2020.
Although a number of studies in humans and different animal models could show Lcn2 and KIM-1 as markers for renal tubular damage (Viau et al. 2010, Paragas et al. 2011, De Oliveira et al. 2019, Albert et al. 2021, little is known about the cellular origin, function and regulation of Penk outside the central nervous system (CNS) (Denning et al. 2008, Caironi et al. 2018. Penk is the approximately 4.5 kDa precursor protein for several different biologically active endogenous opioids such as Met-and Leu-enkephalin. Until recently, Penk has mostly been studied in the CNS, where it plays an important role in nociception and stress handling (Henry et al. 2017, Zeidler et al. 2022. The enkephalins cleaved from Penk activate opioid receptors to mediate their biological effects (Kapusta and Obih 1995, König et al. 1996, Salzet and Tasiemski 2001, Gavériaux-Ruff and Kieffer 2002, Beunders et al. 2017, Luo et al. 2019. The very short half-life-time of about 15 min for enkephalins derived from Penk makes them difficult to measure (Peinado et al. 2003, Beunders et al. 2017. In contrast to the active enkephalins, Penk is more stable and seems to be freely filtered at the glomerulus without known binding to plasma proteins or secretion by tubular cells (Ng et al. 2017, Khorashadi et al. 2020. Unlike Lcn2, Penk seems to be less influenced by extra-renal factors such as systemic inflammation (Beunders et al. 2017, Hartman et al. 2020, Khorashadi et al. 2020. Studies in a number of patient cohorts could show a correlation between Penk plasma levels, GFR and increased risk for worsening renal function and other adverse effects (Kim et al. 2017, Ng et al. 2017, Caironi et al. 2018). These properties make Penk a good candidate as a biomarker for renal function.
However, it is yet unknown whether Penk is a protective or detrimental factor during kidney injury (Beunders et al. 2017, Ng et al. 2017, Emmens et al. 2019, Khorashadi et al. 2020. It is also not yet clear if the elevated plasma Penk levels measured in patients are caused by an increased production of Penk, the reduction in renal filtration or a combination of both (Caironi et al. 2018, Khorashadi et al. 2020. Penk expression outside the central nervous system has previously been reported for the human heart, kidneys and some immune cells (Barron 1999, Sezen 2003, Denning et al. 2008, Ng et al. 2017, Caironi et al. 2018. However, it is not yet clear which cells exactly produce Penk in the kidneys.
This study focuses on identifying which renal cells produce Penk in mice and humans. Furthermore, we investigate how Penk expression changes in two models of renal damage. We combine high-resolution in situ hybridization, quantitative real-time PCR, cell-specific gene deletion and measurement of plasma Penk concentrations to identify the cells producing Penk and their possible regulation in mouse kidneys to provide a better insight into this novel biomarker.

Material and methods
All animal experiments were conducted according to the 'National Institutes of Health guidelines for the care and use of animals in research' and were approved by the local ethics committee (Number 55.2-2532.2-935-28). All mice were kept under optimal conditions, consisting of constant 23 °C room temperature, relative humidity of 55 ± 5% and a constant 12-h dark/light cycle. Standard rodent chow (0.6% NaCl, Ssniff, Deutschland) and tab water were provided ad libitum. All mice were bred in a C57/Bl6 background.
To study the relation of Cox-2 and Penk expression, mice with Cre recombinase under control of the PDGFR-β promotor were crossed with mice bearing the floxed alleles for Cox-2 (Cox-2 fl/fl Stock No: 030785, Jackson Laboratories, RRID:IMSR_JAX:030785) (Ishikawa andHerschman 2006, Yang andLiu 2017). Successful excision of exon 4 and 5 of the Cox-2 gene was verified by genotype-PCR as described previously (Ishikawa and Herschman 2006). Littermates without Cre expression were used as control animals and underwent the same treatment. Adult animals were euthanized by dislocation of the neck after sedation (ketamine, 80 mg/kg; xylazine 10 mg/kg, intraperitoneal). Neonatal mice of both genders were euthanized by decapitation after anaesthesia overdose. Male mice at an average age of 16 weeks were used for the experiments unless stated otherwise.
Genotyping was performed with primers listed in Table 1.

Adenine-induced nephropathy
Adenine-induced fibrosis was initiated in adult mice as described previously (Jia et al. 2013, Buchtler et al. 2018, Rahman et al. 2018, Klinkhammer et al. 2020 After tamoxifen treatment and the following rest period, male mice were fed an adenine containing diet (0.2%; 1324, Altromin, Germany) continually for 3 weeks. Experiments were performed after exactly 3 weeks (3wks adenine) or mice were given an additional 3 weeks on standard diet to recover (3 weeks adenine + 3 weeks recovery).

Unilateral ureteral obstruction
Under inhalation anaesthesia, an ureteral ligation was placed close to the right kidney through a small abdominal incision (Chevalier et al. 2009). Mice were kept under close observation after the operation for 72 h. Ten days after the procedure, mice were sacrificed and perfused for RNAscope or kidneys were removed for mRNA quantification.

Fixation of renal tissue
Adult mice were anaesthetized and euthanized as described earlier.
To fixate the tissue for RNAscope, the abdominal cavity was opened. To reduce the number of mice needed for this study, the right kidney was ligated to preserve it for mRNA measurements and the left kidney was fixed by retrograde perfusion through the abdominal artery. All mice were first perfused with 40 mL of sterile PBS to clear erythrocytes and then with 40 mL of 10% neutral buffered formalin solution. After the perfusion, the ligated (non-perfused) right kidneys were removed, frozen in liquid nitrogen and stored at -80 °C for mRNA measurements. For embryonic and neonatal tissues, one kidney was frozen for mRNA measurements and the other kidney was fixed by immersion in 15 mL of 10% neutral buffered formalin solution under slow agitation.

Determination of mRNA expression by real-time PCR
Total RNA was isolated from kidneys as described by Chomczynski and Sacchi (1987) and quantified by photometer. Of the resulting RNA, 1 µg was used for reverse transcription. cDNA was synthesized by Moloney murine leukaemia virus RT (Thermo Fisher Scientific, Waltham, MA). For quantification of mRNA expression, real-time PCR was performed using a Light Cycler Instrument and the LightCycler 480 SYBR Green I Master Kit (Roche Diagnostics, Mannheim, Germany). mRNA expression data were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Table 2 lists primer sequences.

Determination of Penk plasma concentrations
Blood samples were collected from tail vein into EDTA-coated capillary tubes (KABE LABORTECHNIK GmbH, Nümbrecht, Germany). Plasma was collected after centrifugation (4 min, 12,000 rpm, RT) and stored at -80 °C until used. Penk plasma values were determined with the Penk ELISA Kit for mouse (Cat. OKCD02795, Aviva Systems biology, San Diego, USA) according to the manufacture's guidelines.

Human tissue
For investigation of human kidney tissue, formalin-fixed paraffin-embedded (FFPE) specimens provided by the Department of Nephropathology [Friedrich-Alexander-University (FAU) Erlangen-Nuremberg] were used. The STARD-Diagramm for selection of human tissue samples is represented in Supplemental figure 1. In total, six different human samples were analyzed. Representative details shown in this study were taken from archival kidney tissue sample of a 68-year-old female patient. It was obtained from distant portions of a kidney surgically excised because of the presence of a localized neoplasm. Another kidney sample used for the presented figure was from a transplant nephrectomy from a 45-year-old male patient with chronic rejection. The analysis of archived renal biopsies was approved by the local Ethics committee (Reference Number 4415).

Microscopy
All micrographs were captured with an Axio Observer.Z1 Microscope (Zeiss, Jena, Germany) and a Plan-Apochromat ×20/0.8 objective, a ×1 tube lens and an Apotome.2 system. The Colibri7 (Zeiss, Jena, Germany) was used as light source. Fluorescent images were captured with an Axiocam 506mono camera. Excitation wavelength used were 650 nm, 548 nm, 747 nm and 385 nm and emissions were detected at 673 nm, 561 nm, 773 nm and 465 nm, respectively. For detail images, 5 to 15 z-stacked images were combined for maximum projection. Overviews of whole kidney sections were generated by combination of tiles taken at ×20 magnification. Images represented in the same figure were captured with the same light intensities and exposure times.

Statistical analyses
All data are presented as mean ± SEM. To test for normal distribution of values, Shapiro-Wilk test was performed and the corresponding Q-Q plots were created for graphical verification. Statistical significance between groups was determined by ANOVA and Tukey's multiple comparisons test. p < 0.05 was considered statistically significant. The data were analyzed using Graph Pad Prism9 (Graphpad Software, San Diego, USA, RRID:SCR_002798).

Expression of Penk mRNA in human patients
To investigate the cellular origin of Penk, high-resolution in situ hybridization was performed on kidney tissue of human patients.
Penk mRNA was detected in stromal cells coexpressing the fibroblast marker platelet-derived growth factor receptor-β (PDGFR-β). These cells were partially associated with renal vessels ( Figure 1A). Additional Penk expressing cells could be detected in large cell conglomerates, most likely infiltrates ( Figure 1B) in deeper kidney zones. In these cell infiltrates, Penk was expressed by PDGFR-β + and CD45 + cells.

Penk expression in developing kidneys and healthy adult kidneys of mice
After establishing that Penk was expressed in human kidneys, its expression in mice was investigated to evaluate this species as a potential model organism to study Penk. To this end, Penk mRNA expression was first investigated in healthy murine kidneys during development.
Relative mRNA abundance of Penk was measured in developing mouse kidneys from embryonic day 16 up to day 12 after birth and at the age of 5 weeks. Penk mRNA showed a peak expression on day 3-5 after birth and was rapidly downregulated with completion of post-natal renal development around day 7 after birth ( Figure 2B). Using RNAscope, Penk mRNA was localized exclusively in PDGFR-β expressing interstitial cells and pericytes of developing mouse kidneys (Figure 2A,C). Expression of Penk was restricted to the inner and outer medulla. Figure  2 shows Penk expression on tissue of nine-day old mice as the different structures of the kidney can be more easily distinguished at this point than in three-or five-day old mice. In the renal cortex, Penk mRNA was only expressed by few pericytes around vessels. Tubules did not express Penk mRNA ( Figure 2C). The majority of PDGFR-β + interstitial cells in the cortex and glomerular mesangial cells did not express Penk mRNA.
In adult mice, the expression pattern of Penk mRNA was comparable to neonatal kidneys, but mRNA signals were detected with lower density than in nine-day-old kidneys ( Figure 3A). Penk expressing PDGFR-β + cells of the inner medulla and the inner stripe of the outer medulla also  coexpressed tenascin-C (TNC). TNC + interstitial cells constitute a subpopulation of PDGFR-β + interstitial cells (Broeker et al. 2020). Penk + cells could also be detected along vascular bundles in the outer medulla. In the cortex, only few vessel associated pericytes ( Figure 3D) showed Penk mRNA signals. The majority of PDGFR-β + interstitial cells of the cortex did not express Penk mRNA. Throughout the kidney cortex Penk mRNA was also detected in some CD45 + immune cells ( Figure 3C). Tubular structures of adult mouse kidneys did not show any signals for Penk. Further colocalization studies showed that basal Penk mRNA expression did not co-localize with markers for endothelial cells like CD31. In lung and heart tissue of healthy mice, signals for Penk mRNA could be detected, although these were fewer than in the kidneys. In these organs the majority of Penk + cells also expressed PDGFR-β mRNA (Supplemental figure 2).

Penk is highly upregulated in renal interstitial cells in two models of experimental fibrosis
To investigate how expression of Penk mRNA and Penk plasma concentrations are affected by renal damage in mice, two different models of experimental kidney fibrosis were analyzed. Unilateral ureteral obstruction (UUO) for ten days, served as acute damage model and adenine-induced nephropathy (AN) for three weeks was used as chronic model. Both models led to a strong upregulation of Penk mRNA expression ( Figure 4E,F). Additional Penk mRNA in damaged kidneys could be detected in PDGFR-β + myofibroblasts of the outer and inner medulla ( Figure 4A-C). Results of co-localization studies are shown here exemplary on tissue of mice with AN ( Figure 5). The strong increase in mRNA abundance of Penk in AN was also reflected by significantly increased plasma Penk levels. When the mice were given a three-week recovery period from the fibrotic stimulus, mRNA and plasma levels were no longer significantly different from basal conditions ( Figure 4D,E). In UUO kidneys, Penk was upregulated in the remaining tissue of the inner and outer medulla. The cell types expressing Penk in this model were the same as after three weeks adenine treatment ( Figure 4D,F).

Lcn2 and KIM-1 are expressed exclusively in tubules of mice during experimental kidney fibrosis
To gain more information about a possible differential expression pattern of the novel biomarkers, the spatial expression pattern of Penk in combination with Lcn2 and KIM-1 mRNAs was analyzed.
Penk was expressed by interstitial fibroblast like cells and myofibroblasts after renal damage ( Figure 5). In contrast, Figure 2. rNascope on tissue of nine-day old wildtype mice for Penk (red signals) and PDGFr-β (green signals) mrNa and measurement of mrNa abundance during nephrogenesis. in developing kidneys, Penk mrNa (red) can be detected in PDGFr-β (green) expressing interstitial cells of the inner and outer medulla (a). Tubular structures are negative for Penk. in the inner and outer medulla almost every PDGFr-β + cell coexpresses Penk (C) while PDGFr-β + cells in the cortex and glomeruli do not express Penk. measurement of mrNa abundance shows that during the late embryonic stage and in the period of postnatal renal development Penk expression is about 200-fold higher than in adult kidney tissue (B), reaching its peak between day 3 and 5 after birth. rectangle in a indicates area of high magnification for C. Nuclei were counterstained with DaPi (grey). scale bar in a is 200 µm; scale bar in C is 50 µm.
Lcn2 and KIM-1 mRNA were detected exclusively in tubular structures of damaged kidneys after UUO and adenine-induced nephropathy ( Figure 6). Lcn2 was detected in distal tubules and aquaporin 2 (AQP2) ( Figure 6B) expressing collecting ducts in the outer and inner medulla. Damaged tubules with strong Lcn2 expression were surrounded by Penk + interstitial cells. There was, however, no co-localization of these two makers in the same cells ( Figure  6A). KIM-1 expression on the other hand was restricted to proximal tubules co-expressing megalin in the renal cortex ( Figure 6C). No expression of Penk mRNA could be detected in or around KIM-1 + tubules. Expression of tubule specific markers like megalin or AQP2 was weaker in damaged tubules with increased expression of Lcn2 and KIM-1. Exemplary details of the co-localization studies are shown in tissues of mice after 3 weeks AN. mRNA expression patterns and cell-types expressing Penk were the same in UUO damaged kidneys.

Penk mRNA expression is not regulated through classical profibrotic pathways
The influence of proinflammatory prostaglandin signalling and profibrotic TGFβ-1 signalling in PDGFR-β + cells on Penk expression was investigated in damaged kidneys. Both pathways were interrupted by the cell-specific deletion of a central signalling component. For prostaglandin signalling, Cox-2 the prostaglandin producing enzyme was deleted in PDGFR-β + cells (PDGFR-β ERT2Cre/+ Cox-2 ff mice). For TGFβ-1 signalling, the essential receptor TGFβ-R2 was deleted (PDGFR-β ERT2Cre/+ TGFβ-R2 ff mice). Progression and degree of tissue damage was unchanged in these animals for both Figure 3. Characterization of Penk (red) mrNa expressing cells in the kidneys of healthy adult wildtype mice. in the inner and outer medulla Penk + (red) cells coexpress PDGFr-β (green) mrNa (a, arrows). interstitial Penk and PDGFr-β + cells of the inner medulla and inner stripe of the outer medulla additionally express mrNa for the matrix protein tenascin-C (green, B, arrows). Penk mrNa expressing cells in the renal cortex are CD45 + leucocytes (green, arrow C) and vessel associated PDGFr-β + (green) pericytes (asterisks) surrounding medium sized vessels, marked by myh11 (blue, D) expression. Nuclei were counterstained with DaPi (grey). scale bars 20 µm. Figure 4. overview of mouse kidney sections with rNasccope for Penk mrNa (red) under basal and different pathological conditions and measurement of Penk mrNa abundance and plasma concentrations. Under basal conditions Penk expression is low and restricted mainly to the inner medulla and inner stripe of the outer medulla (a). after 3 weeks on an adenine containing diet, Penk mrNa expression is strongly upregulated (e) and robust signals are detected up to the cortico-medullary border (B). after 10 days UUo Penk is detected in the remaining tissue of the outer medulla and in the cortex (C). Nuclei were counterstained with DaPi (grey). scale bar 200 µm. measurement of Penk in plasma of aN-treated mice shows a significant increase in Penk levels after three weeks aN, which returns to levels no longer significant from basal conditions after the recovery period (D). mrNa abundance in line with plasma values is significantly higher in damaged kidneys after 3wks adenine (e) and UUo (F), but returns to basal levels in mice that are given a recovery period after adenine feeding on a normal diet (e). Data are mean ± sem of at least five animals per condition. significant p values are stated above the respective pairs. Figure 5. Details of rNascope for Penk mrNa (red) and different profibrotic and cellular markers (green) on tissue of mice after 3 weeks adenine diet. in damaged kidneys, the majority of Penk expressing cells (red) also expresses PDGFr-β (green, a). These interstitial PDGFr-β/Penk + cells are identified as eCm producing cells by their expression of Col1a1 (green, B) and as myofibroblasts by co-hybridization with α-sma (green, arrows, C). interstitial Penk expressing cells also show mrNa signals for TGFβ-r2 (green, arrows D), but tubules expressing TGFβ-r2 were negative for Penk mrNa (D, asterisks). Nuclei were counterstained with DaPi (grey). scale bars 50 µm. tested models of kidney fibrosis. The cell-specific deletion of Cox-2 or TGFβ-R2 in PDGFR-β cells had no effect on Penk mRNA expression after 3 weeks AN (Figure 7) or after 10 days UUO.

Discussion
Knowledge of renal function is essential for the proper treatment of patients in order to preserve kidney function during renal and non-renal afflictions (Caironi et al. 2018, De Oliveira et al. 2019, Khorashadi et al. 2020. Serum creatinine values are progressively modified by more complex calculation methods to compensate the deficiencies inherent to the marker (Herget-Rosenthal et al. 2007). So far the search for more reliable biomarkers of renal function has turned up a number of candidates, among them the enkephalin precursor Penk (De Oliveira et al. 2019, Hartman et al. 2020, Khorashadi et al. 2020). This study aimed to identify the renal cell types expressing Penk mRNA in humans and mice. Furthermore, we investigated changes in expression and possible regulatory pathways of Penk in more detail.
Penk expression could be detected in PDGFR-β + interstitial fibroblast-like cells of the inner medulla, along renal vessels in the outer medulla and in CD45 + immune cells on human and murine kidney tissues (Figures 1-3). The distinct expression pattern of Penk is in line with a previously proposed hypothesis that subpopulations exist within the renal interstitium (Stefańska et al. 2013, Kurtz 2017, Broeker et al. 2020. Therefore, we would hypothesize that Penk/PDGFR-β + cells represent a new endocrine subpopulation of renal interstitial cells. Penk mRNA was also strongly expressed in PDGFR-β + cells of embryonic and neonatal kidneys during the postnatal renal development of mice. However, Penk mRNA was strongly downregulated after completion of renal development in mice (Figure 2). These findings are in good accordance with a recent study, which found that Penk values in children were higher than in adults (Hartman et al. 2020). We hypothesize that Penk expression could be linked to the Figure 6. Details for Penk, Lcn2 and kim-1 rNascopes with respective markers on tissue of wildtype mice after 3 weeks adenine treatment. Penk mrNa (red) expressing interstitial cells can be detected around damaged Lcn2 + (green, a) tubules. There is no co-localization of Penk and Lcn2 in the same cells. expression of Lcn2 is strong in collecting ducts co-expressing aQP-2 mrNa (blue, B), highly damaged collecting ducts lost expression of aQP-2 and only express Lcn2. kim-1 mrNa (red) expression is restricted to megalin positive proximal tubules in the renal cortex (green, C). Nuclei were counterstained with DaPi (grey). Circle indicates glomerulus. scale bars 50 µm. prevalent tissue hypoxia reported for developing kidneys of rats and humans (Bernhardt et al. 2006). The localization of basal or stimulated Penk in adult mice in the inner zone of the outer medulla, a zone of low tissue oxygenation would be in line with this hypothesis (Figures 2 and 3) (Brezis andRosen 1995, Fry et al. 2014). A link between increased levels of protective endogenous enkephalins, derived from Penk, and hypoxia signalling has previously been reported in the central nervous system and the heart (Gao et al. 2012, Kleinbongard et al. 2017). However, a definitive connection of hypoxia signalling and Penk expression in the kidneys needs further experimental evidence.
Closer characterization of Penk + cells in two mouse models of renal fibrosis showed an about eightfold upregulation of Penk mRNA in PDGFR-β + , matrix producing cells of the inner and outer medulla (Figures 4 and 5). In line with the elevated mRNA abundance, plasma Penk levels were increased. Renal Penk expression was dependent on the persistence of the damaging stimuli. Plasma levels and mRNA expression were significantly reduced if the animals were given a recovery period, down to levels no longer significantly different from basal levels ( Figure 4E). Using in-situ hybridization Penk mRNA was detectable in some PDGFR-β + cells of murine heart and lung tissue in accordance with a previous study (Denning et al. 2008). However, unlike the strong increase in the kidneys, only a very slight increase in Penk mRNA signals could be detected in heart tissue after 3 weeks adenine treatment. No apparent change in the expression of Penk could be observed in lung tissue (Supplemental figure 2).
To learn more about possible pathways involved in the regulation of Penk during experimental fibrosis, we analyzed Penk in different genetic mouse models (Fuchs et al. 2021). These data indicate that Penk expression is not regulated by proinflammatory prostaglandin or profibrotic TGFβ-1 signalling in PDGFR-β + cells of mouse kidneys (Figure 7).
The strong renal interstitial expression of Penk in the pathological models investigated for this study is in clear contrast to other newly proposed markers for kidney injury. Both Lcn2 and Kim-1 are exclusively expressed in different parts of damaged tubules ( Figure 6). This is in good accordance with previous studies of these two markers (Paragas et al. 2011, Humphreys et al. 2013, De Oliveira et al. 2019. The differential expression sites of these new biomarkers indicate that caution should be used when evaluating their predictive performance for different diseases (De Oliveira et al. 2019). But it might also provide an opportunity to establish a method, where a profile of different biomarkers is measured to give a more detailed diagnosis of renal diseases.
Taken together, we show a characterization of Penk expressing cells in human kidneys as well as in murine kidneys during post-natal renal development and in different models of renal damage. In contrast to the tubular injury markers Lcn2 and Kim-1, this study identifies Penk as an . mrNa measurement for Penk in different genetic mouse models with and without adenine induced damage. in damaged kidneys neither interruption of prostaglandin signalling by deletion of Cox-2 nor disruption of TGFβ-1 signalling had any effect on renal Penk mrNa levels. Data are mean ± sem of at least 6 animals per condition. significant p values are stated above the respective pairs. interstitial cell marker responding to acute fibrotic stimuli. In addition, we could show that Penk does not seem to be regulated directly through classical profibrotic pathways like TGF-β1 or prostaglandin signalling. Until we learn more about the function of these novel markers in renal damage, a clear distinction of the clinical causes for renal damage is necessary when evaluating the predictive performance of these markers.