Early Urinary Markers of Kidney Injury in Adult People with Cystic Fibrosis Correlate with Neutrophil Activation and Lung Infection

Grace M. Rosner, Himanshu B. Goswami, Katherine Sessions, Lindsay K. Mendyka, Brenna Kerin, Irma Vlasac, Diane Mellinger, Lorraine Gwilt, Thomas H. Hampton, Martha Graber, Alix Ashare, William T. Harris, Brock Christensen, Agnieszka Swiatecka-Urban, Bruce A. Stanton, Sladjana Skopelja-Gardner
doi: https://doi.org/10.1101/2023.11.10.23298378

Abstract

Background Adult people with cystic fibrosis (PwCF) have a higher risk of end-stage kidney disease than the general population. The nature and mechanism of kidney disease in CF are unknown. This study quantifies urinary kidney injury markers and examines the hypothesis that neutrophil activation and lung infection are associated with early kidney injury in CF.

Methods Urinary total protein, albumin, and markers of kidney injury and neutrophil activation, normalized to creatinine, as well as urinary immune cells, were quantified in CF (n = 48) and healthy (n = 33) cohorts. Infection burden and chronicity were defined by sputum culture and serum titers of anti-bacterial antibodies.

Results PwCF had increased urinary protein levels, consisting of low-molecular-weight tubular injury markers, independent of glomerular filtration rate (eGFR). This finding suggests subclinical renal injury processes. Urinary analysis of the CF cohort identified two renal tubular injury signatures: one associated with aminoglycoside exposure and one associated with increased urinary levels of neutrophil activation. High urinary KIM-1 levels and increased prevalence of neutrophils among urine immune cells correlated with decreased lung function in PwCF. The relationship between tubular injury and neutrophil activation was most prominent in patients harboring chronic Pseudomonas aeruginosa infection.

Conclusions Increased urinary tubular injury markers in PwCF suggest early subclinical renal injury not readily detected by eGFR. The strong association of high urinary KIM-1 and neutrophils with diminished lung function and high Pseudomonas aeruginosa burden suggests that pulmonary disease may contribute to renal injury in CF.

Introduction

Besides the lungs, mutations in the Cystic Fibrosis Transmembrane Conductance Regulator (CFTR) can cause pathology in other organs, including the pancreas, upper airways, liver, intestine, kidneys, and reproductive organs 1,2. Even though lung disease is the primary cause of morbidity and mortality in CF, extrapulmonary manifestations resulting from CFTR dysfunction in the pancreas, intestine, or kidney may contribute to disease outcomes. Advancements in early diagnosis, intensive antibiotic treatment, and most notably, high-efficiency CFTR modulators (HEMT) have remarkably extended the lifespan of people with CF (PwCF), now exceeding 50 years of age. This increased longevity has brought to light a concerning rise in long-term complications, such as chronic kidney disease (CKD), which affects more than 1 in 7 adults in the United States. The literature estimates that 14–29% of PwCF suffer from CKD 2–4. Moreover, PwCF with CKD have up to 20-fold higher risk of developing end-stage renal disease (ESRD) compared to non-CF patients with CKD 4–6. Despite these alarming figures, the nature of early renal injury and the mechanisms by which inflammatory events in the lung may contribute to kidney pathogenesis in CF remain unknown.

Potential contributors to acute kidney injury (AKI) in CF are aminoglycosides 3,7,8, used to treat frequent pulmonary exacerbations triggered by Pseudomonas aeruginosa (P. aeruginosa) infections. Aminoglycosides are retained in the lysosomes of proximal tubular cells, where they can disrupt tubular reabsorption and lead to necrosis 9. However, several studies showed that kidney disease in PwCF may be independent of prolonged aminoglycoside use or cumulative lifetime aminoglycoside exposure 3,10,11, suggesting that other factors relevant to CF disease, such as CFTR dysfunction in the kidney, repeated episodes of dehydration, or systemic inflammation, may propagate kidney damage. Moreover, studies showing an association between aminoglycoside use and AKI at the time of a pulmonary exacerbation did not consider the inflammatory events triggered by lung infection12. That inflammatory events in the lung can cause kidney injury was first recognized in acute respiratory disease syndrome (ARDS)-mediated AKI 13, but likely extends to other lung illnesses 14. In chronic obstructive pulmonary disease (COPD), ∼24% of patients had persistent albuminuria, compared to 4% of control subjects, and lower estimated glomerular filtration rate (eGFR) in COPD patients correlated with decreased lung function 15. In in vivo murine studies, intratracheal instillation of lipopolysaccharide (LPS) caused acute renal injury accompanied by both pulmonary and renal inflammation 16. Whether infection-triggered inflammation or diminished lung function in CF contributes to early kidney injury is unknown. Given the high neutrophil infiltration in the CF lung before and during infection 17–20, and the reported ability of neutrophils to mediate both local and distal tissue damage 21,22, we hypothesized that activated neutrophils contribute to kidney injury in PwCF.

Several studies have reported increased levels of urinary kidney injury markers such as kidney injury molecule 1 (KIM-1) and β2-microglobulin (β2MG) in both adult and pediatric CF cohorts 4,7,8,23. While these studies were limited to a small number of tubular injury markers, they suggested that kidney damage in PwCF may develop before it is detected by standard clinical tests such as albuminuria or changes in eGFR. Neither a decrease in eGFR, which may occur as late as after >50% of kidney function loss, nor an increase in albuminuria, a marker of glomerular injury, provide an insight into disease etiology or capture early signs of renal injury. This study aims to define early urinary markers of kidney damage and asks if lung infection and neutrophil-mediated inflammation may play a role in subclinical renal injury in CF.

Methods

Study Cohorts

Urine and blood were collected by written informed consent from a cohort of 48 adult (> 18 years old) CF patients, recruited between the years 2017 and 2022, at Dartmouth-Hitchcock Medical Center (DHMC), Lebanon, New Hampshire, through the Clinical and Translational Research Core (CTRC) of the Dartmouth CF P30. Frozen urine and serum samples and demographic and clinical data were obtained from the CTRC. Sputum cultures, patient lung function, and other laboratory measurements were assayed by clinical laboratories at DHMC. The eGFR was computed using the CKD-EPI Creatinine Equation. Urine from 33 healthy controls (HC) was collected by written informed consent. Cohort characteristics are described in Table 1, and Supplemental Table 1. The correlation between urinary kidney injury markers and clinical parameters in the CF Cohort are summarized in Table 3. An additional CF cohort of 11 patients (2022-2023, characteristics in Supplemental Table 2) was recruited to obtain urinary cell pellets for methylation studies.

View this table:
Table 1: Characteristics of the Healthy and CF cohort.

Urine collection and processing

Urine samples were processed immediately after collection and tested for urinary tract infections using Chemstrip® Test Strips (Roche). Samples used for urinary analysis of kidney injury and neutrophil activation were aliquoted and immediately frozen at −80°C. For urine DNA methylation studies, urine samples were centrifuged (200g) to separate cell-free urine from the cell pellet. Cell pellets were preserved at −80°C until DNA extraction.

Quantification of urinary renal injury markers and neutrophil activation

Total urine protein levels were quantified with Bradford assay (Pierce™). Urinary albumin and creatinine levels were determined using the Albuwell Hu and Creatinine Companion ELISA kit (Ethos Biosciences). The levels of urinary kidney injury proteins were quantified by ELISA: VCAM-1 (R&D); podocalyxin (Invitrogen); beta-NAG (Abcam); prostasin (ThermoFisher); uromodulin (R&D) or Human Kidney Biomarker Panel Luminex® (R&D): β2MG, EGF, cystatin C, osteopontin, KIM-1, NGAL, and TFF-3. Urinary S100A8/A9 was measured by ELISA (R&D). Urinary NETs were quantified by ELISA measuring MPO: DNA complexes as previously described 24. Isolated NETs (PMA-treated neutrophils) were used to establish a standard curve. All protein concentrations were normalized to urinary creatinine levels.

Anti-P. aeruginosa and anti-S. aureus immunoglobulin G (IgG) quantification

Serum concentrations of anti-P. aeruginosa and anti-Staphylococcus aureus (S. aureus) IgG were quantified by ELISA using bacterial lysates generated from overnight cultures of P. aeruginosa PA14 and S. aureus USA100 strains as previously established in our lab 25.

DNAm-based deconvolution of cell types in urine

DNA from urine cell pellets was extracted using Zymogen Mini/Micro Prep Kit and 250 ng of DNA was bisulfide converted and run on the EPIC Illumina DNAm array at the Dartmouth Genomics Core. Infinium Methylation EPIC BeadChip v1.0 (EPIC) raw intensity data (IDAT) files were preprocessed using minfi and Enmix. Probes with out-of-band hybridization >0.05 were excluded from the analyses, followed by normalization and background correction using FunNorm, (minfi). Probe type correction for EPIC was completed using beta-mixture quantile normalization (BMIQ). EpiDISH 2.12/HEpiDISH was used to calculate cell type proportions in urine samples 26. We used an immune-specific reference data set using twelve immune cell types based on the iterative algorithm Identifying Optimal Libraries (IDOL) 26. Robust partial correlation methods were used.

Statistical analysis

All analyses were performed using GraphPad Prism software and R version 2023.06.0+421. Differences between two means were analyzed using Student’s t-test with a statistical significance of p<0.05. Correlations were assessed using Pearson’s or Spearman’s analysis (p<0.05) for ordinal or nominal data, respectively. Principal component analysis was performed using GraphPad Prism using urine proteomic values as well as clinical factors.

Results

Proteinuria and urinary tubular injury markers are elevated in PwCF

Adult PwCF (n=48) and healthy controls (HC; n=33) were recruited for this study. Patient characteristics are summarized in Table 1. Total urinary protein, normalized to urine creatinine, was significantly higher in 29.2% (N = 14) of PwCF, compared to HC (Fig. 1A). Urinary protein did not correlate with eGFR (Fig. 1A). Increased total urinary protein in HC but not PwCF correlated with age (Fig. S1). The increased urinary protein in PwCF was not due to albuminuria, as the urine albumin/creatinine ratio was comparable between the CF and HC cohorts (Fig. 1B).

Figure 1.
Figure 1. Urinary levels of tubular proteins indicate subclinical renal injury in adult PwCF.

(A) Total urine protein levels were measured by Bradford assay and normalized to urine creatinine in healthy control (HC, n =33) and CF (CF, n= 48) cohorts and stratified by eGFR, computed using the CKD-EPI Creatinine Equation: blue: <60, green: 60-90, grey: 90-120, pink: >120. Significance was defined by Student’s t-test, **p < 0.01. The dotted line represents a positive cutoff (0.87) based on mean HC (0.45) + 2 S.D. (0.21). (B) Urinary levels of albumin (uAlbumin) normalized to creatinine in HC and CF cohorts. Student’s t-test defined significance, ns = not significant (p > 0.05). (C) Urinary levels of kidney injury markers in HC (pink) and CF (black) cohorts, normalized to urine creatinine. Significance was determined by multiple unpaired t-tests with *p<0.05. (D) Pearson correlation analysis between urinary levels of epidermal growth factor (uEGF) normalized to urine creatinine and eGFR, computed as in A. (E) Principal component (PC) analysis (left panel) of urinary kidney injury markers measured in panel C in HC and CF cohorts. Circles indicate two subpopulations of the CF cohort with differential urinary levels of injury markers. The Biplot (right panel) showing vectors of differentiation (blue) indicates which urinary proteins contribute to the segregation of the two subpopulations in the PCA. (F) PCA (left panel) of urinary markers measured in panel B in the CF cohort, as a function of positive (red) or negative (blue) total urine protein (defined as in A). Biplot (right panel) showing vectors of differentiation (blue) indicates which urinary proteins contribute to the segregation of two subpopulations in the PCA. (G) total urine protein normalized to creatinine, in PwCF with (pink) or without (grey) a diagnosis of diabetes, were compared by Student’s t-test; ns = not significant (p > 0.05). (H) Principal component (PC) analysis of urinary markers measured in C in the CF cohort stratified by clinical presentation of diabetes mellitus (blue) or no clinical presentation of diabetes (CFRD and type II, red). (F) Spearman correlation analysis between urinary kidney injury markers measured in C, % FEV1, eGFR, and status of antibiotic, azithromycin (AZ), tobramycin (Tobi), or aztreonam use at the time of sample collection. The color of the squares indicates a significant positive (blue) or negative (red) correlation, and the r coefficient is shown within each square. β2m = β2-microglobulin, Kim-1 = kidney injury marker 1, EGF = epidermal growth factor, OPN = osteopontin, NGAL = Lipocalin 2, TFF3 = Trefoil factor 3, PDX = Podocalyxin, βNAG = N-acetyl-β-d-glucosaminidase.

To test if elevated urinary protein levels indicate subclinical tubular injury and to identify early biomarkers of such processes, we used Luminex renal injury panels and ELISA assays to quantify urinary levels of low-molecular-weight kidney injury markers. As summarized in Table 2, urinary levels of tubular injury markers, normalized to urine creatinine, including β2MG, Cystatin C, KIM-1, TFF3, β-NAG, and VCAM-1, were significantly higher in PwCF compared to HC (Fig. 1C). Using the HC to define the baseline urinary levels of each biomarker (mean of HC + 2 S.D.), we found that 17–48.9% of CF samples had high urinary levels of one or more kidney injury markers (Table 2). Even though some urinary injury markers were not significantly increased in the overall CF cohort (e.g., NGAL), a subpopulation (16.7%) of PwCF had elevated urinary NGAL. Contrary to the increased urinary levels of tubular injury markers, urinary podocalyxin, which shedding has been associated with glomerular disease 27, was significantly decreased in PwCF, compared to HC (Fig. 1C, Table 2). Consistent with signs of tubular injury, there was a trend in reduced urinary epidermal growth factor (uEGF) in PwCF (Table 2), which positively correlated with eGFR (Fig. 1D). Levels of other tubular injury markers did not correlate with eGFR, further supporting that eGFR fails to reflect early signs of tubular damage.

View this table:
Table 2: Urinary kidney injury markers in the CF vs. HC cohorts
View this table:
Table 3: Correlation between urinary kidney injury markers and clinical parameters and treatment in the CF Cohort.

Principal component analysis (PCA) based on urinary measurements of the 10 kidney injury markers demonstrated 33.3% (n=16/48) of PwCF grouped separately from HC (Fig. 1E left panel). Differential urinary levels of kidney injury markers segregated PwCF into two subpopulations (Fig. 1E left panel, circled) as shown in the biplot: i) KIM-1, EGF, b-NAG, NGAL, OPN, and podocalyxin in the upper part of the plot and ii) Cystatin-C, VCAM-1, TFF3, and β2MG in the bottom part of the plot (Fig. 1E, right panel). Urinary levels of most tubular injury markers correlated with total urinary protein levels (Table 2). However, stratifying CF samples by proteinuria (positive vs. negative, established as the mean of HC + 2 S.D.) maintained the separation of the CF cohort into two populations (Fig. 1F, left panel), driven by the same two groups of urinary injury markers as in Fig. 1C (Fig. 1F, right panel biplot). These findings suggest two potentially distinct mechanisms of tubular damage may contribute to proteinuria in PwCF.

We first considered the potential role of diabetes mellitus in kidney injury as it is one of the leading causes of CKD in the general population. Despite its high prevalence in PwCF (37.5% in our cohort, Table 1), the presence of diabetes was not associated with increased total urine protein/creatinine ratio (Fig. 1G) or with a unique urinary injury signature in PwCF (Fig. 1H). Given the nephrotoxic nature of aminoglycosides, which are indicated for the treatment of acute disease exacerbations and chronic lung infection in PwCF, we next asked if increased urinary levels of tubular injury markers in PwCF are a consequence of treatment with tobramycin (nephrotoxic aminoglycoside, 25% of PwCF), compared to antibiotics with low nephrotoxic risk: azithromycin (41.6% of PwCF) and aztreonam (4.1% of PwCF). Urinary levels of β2MG, Cystatin C, and TFF3 were significantly increased in PwCF receiving tobramycin therapy at the time of urine collection, while only urinary TFF3 was elevated in PwCF receiving azithromycin (Fig. 1I and Table 3). Interestingly, the three renal injury markers elevated in tobramycin-treated patients segregated together in the initial PC analysis (Fig. 1E, right panel). Together, these findings demonstrate subclinical injury processes in PwCF that are independent of diabetes mellitus and are only partially associated with aminoglycoside therapy.

Early tubular injury in CF patients is associated with worse lung function and chronic P. aeruginosa infection

While lung-kidney crosstalk has been recognized in several conditions of lung inflammation (e.g., severe illness, COPD, asthma, and pulmonary fibrosis 28), how these organs may communicate in CF patients is unknown. To establish a link between the lung and the kidney in CF we examined if lung function (% forced expiratory volume in 1 second, %FEV1) correlates with urinary levels of kidney injury markers. We found that urinary levels of tubular injury marker KIM-1, normalized to creatinine, negatively correlated with %FEV1 (Fig. 2A), suggesting an inverse relationship between lung function and kidney injury in PwCF. While neither %FEV1 nor urinary KIM-1 levels were associated with aminoglycoside use (Figure 1I, Table 3), low %FEV1 was most notable in PwCF with a high burden of mucoid P. aeruginosa in sputum culture (Fig. 2A, red and blue circles). Additionally, PwCF with increased sputum burden of mucoid P. aeruginosa had significantly higher urinary levels of KIM-1 than those with lower infection burden (Fig. 2B). No association was found between non-mucoid P. aeruginosa in sputum and %FEV1 or uKIM-1 levels (not shown). In support of a relationship between chronic P. aeruginosa infection and renal injury, we found that anti-P. aeruginosa IgG levels in the serum of PwCF positively correlated with urinary KIM-1 levels (Fig. 2C). Contrary to the associations with chronic P. aeruginosa infection, PwCF who had higher sputum S. aureus burden had significantly lower urinary KIM-1 compared to PwCF with no or low S. aureus burden (Fig. 2D), suggesting that chronic S. aureus infection may not be associated with kidney injury in CF. Indeed, anti-S. aureus IgG levels in the serum were negatively correlated with urinary KIM-1 levels (Fig. 2E). These findings suggest that chronic lung infections with mucoid P. aeruginosa strains may be an important mediator of the lung-kidney axis in CF. In support of this model, patients with mucoid P. aeruginosa, but not S. aureus, positive sputum culture also had significantly lower urinary levels of EGF (Fig. 2F).

Figure 2.
Figure 2. High urinary KIM-1 levels are associated with worse lung function and increased mucoid P. aeruginosa burden in PwCF.

(A) Pearson correlation analysis between urinary KIM-1 (uKIM-1) normalized to urine creatinine and percent predicted forced expiratory volume in 1 second (%FEV1, n =48) in the CF cohort. Distribution of the mucoid P. aeruginosa burden (none = black, few = green, moderate = blue, many = red) in sputum culture in relation to %FEV1 and uKIM-1. (B) Levels of uKIM-1 in PwCF with no (none), few, and moderate/many mucoid P. aeruginosa colonies in the sputum culture were compared by one-way ANOVA with Bonferroni post-hoc (ns = not significant, p > 0.05; *p < 0.05; **p < 0.01). (C) Pearson correlation analysis between uKIM-1 levels and anti-P. aeruginosa IgG serum titres (n = 27 PwCF). (D) Levels of uKIM-1 in PwCF with no (none), few, and moderate/many S. aureus colonies in the sputum culture were compared by one-way ANOVA with Bonferroni post-hoc (ns = not significant, p > 0.05; *p < 0.05). (E) Pearson correlation analysis between uKIM-1 levels and anti-S. aureus IgG serum titres (n = 27 PwCF). (F) Urinary levels of EGF (uEGF), normalized to urine creatinine, in PwCF colonized with P. aeruginosa (left) or S. aureus (right) by sputum culture were compared to PwCF not colonized with each pathogen by Student’s t-test (n=48, ns = not significant, p > 0.05; *p < 0.05).

High levels of renal injury markers positively correlate with neutrophil activation, in the context of chronic P. aeruginosa infection

Neutrophils are highly abundant innate immune cells in the infected CF lung 18,29. In other models of local tissue injury neutrophils have been shown to disseminate to distal organs, including our own findings that following skin injury neutrophils migrate to the kidney to cause renal damage 21,30,31. To ask if neutrophil activation associates with kidney injury in PwCF, we quantified urine levels of calprotectin (S100A8/9) and neutrophil extracellular traps (NETs, MPO: DNA complexes), known markers of neutrophil activation 24. Urinary calprotectin and/or NETs levels, normalized to creatinine, positively correlated (blue squares) with several urinary markers of tubular (NGAL, KIM-1, TFF3, β-NAG) and glomerular (PDX) injury (Fig. 3A, Table 4). The positive correlation between uNETs and uKIM-1, which was elevated in PwCF with low %FEV1 and high serum anti-P. aeruginosa IgG levels (Fig. 2A–C), implicates neutrophil activation as the mechanistic link between chronic P. aeruginosa lung infection and kidney injury. Indeed, urinary levels of NETs were positively correlated with serum titers of anti-P. aeruginosa IgG, but not with anti-S. aureus IgG (Fig. 3B–C), suggesting pathogen specificity in neutrophil-mediated responses and their role in early renal injury. Similarly, urinary NGAL strongly positively correlated with the presence of uNETs as well as with serum anti-P. aeruginosa IgG levels (Fig. 3A, D) but not with anti-S. aureus IgG (Fig. 3E). These observations support the model by which kidney injury in PwCF is related to neutrophil activation in a pathogen-dependent manner.

Figure 3.
Figure 3. High urinary levels of tubular injury markers correlate with urinary calprotectin and NETs, in the context of chronic P. aeruginosa infection.

(A) Pearson correlation matrix of urinary kidney injury markers measured in Figure 1C, normalized to urine creatinine, with urinary markers of neutrophil activation: neutrophil extracellular traps (NETs, MPO: DNA complexes) and calprotectin (S100A8/A9), normalized to creatinine. The color of the squares indicates a significant positive (blue) or negative (red) correlation, with the r coefficient shown within each square (n = 48). (B-C) Pearson correlation analysis between (B) anti-P. aeruginosa IgG serum titers or (C) anti-S. aureus IgG serum titers in PwCF and urinary NETs (uNETs, n = 27, AU = Arbitrary Units defined based on standard curve of PMA-induced NETs extract). (D-E) Pearson correlation analysis between (D) anti-P. aeruginosa IgG serum titers or (E) anti-S. aureus IgG serum titers of PwCF and urinary neutrophil gelatinase-associated lipocalin (uNGAL, n = 27).

View this table:
Table 4: Correlation between urinary kidney injury and neutrophil activation markers.

Urinary neutrophil levels are elevated in CF patients with decreased lung function

While increased urinary calprotectin and NETs implicate neutrophil activation in renal injury in PwCF (Fig. 3), these are indirect measurements of neutrophil presence in urine. Since recent studies have demonstrated that urine cell profiles closely mimic cellular diversity in the kidney tissue 32,33, we quantified neutrophil levels in the urine of PwCF by leveraging cell-specific DNA methylation (DNAm) signatures 34,35. As cells commit to different lineages, they inherit characteristic DNAm signatures in their epigenome. Utilizing reference-based cell-type deconvolution of differentially methylated regions (DMRs), we quantified immune cells in the urine cell pellets from 11 PwCF (Fig. 4A). For these studies, we recruited a 2nd cohort of PwCF as the samples collected in the initial CF cohort did not include urinary cell pellets (patient characteristics in Supplemental Table 2). Deconvolution of urinary DNAm data identified myeloid cells (Myel: monocytes and macrophages), neutrophils (Neu), memory B cells (Bmem), and regulatory T cells (Treg) as the most prevalent immune populations in the cell pellets of urine samples from PwCF (Fig. 4B). Among innate immune cells, the percentage of neutrophils within all immune cells was significantly higher than that of myeloid cells (Fig. 4C).

Figure 4.
Figure 4. Urinary neutrophil levels correlate with worse lung function in PwCF.

(A) Urine cell pellets and cell-free supernatants were separated by centrifugation. DNA was extracted from frozen cell pellets, and 250ng was run on an Illumina EPIC DNA methylation (DNAm) array for deconvolution of different immune cell populations. (B) Deconvolution of DNAm data demonstrated the presence of neutrophils (Neu), myeloid cells (Myel: monocytes and/or macrophages), regulatory T cells (Treg), memory B cells (Bmem), and basophils (Bas) in the urine of PwCF, within all immune cells. (C) The percentages of neutrophils and myeloid cells within all immune cells in urine were compared by Student’s t-test, n = 11, *p < 0.05. (D) Pearson correlation analyses between the percentage of Neu and Myel cells within urine immune cells (left) or the percentage of Treg and Bmem (right) and percent predicted forced expiratory volume in 1 second (%FEV1, n=11).

Interestingly, percent neutrophils in urine immune cells inversely correlated with %FEV1 (Fig. 4D). No relationship was observed between myeloid cells, Treg, or memory B cells with %FEV1 (Fig. 4D). These findings suggest that increased neutrophil infiltration in the kidney is associated with worse lung function and support the role for neutrophils in renal injury in CF.

Discussion

In this study, we report that PwCF display signs of subclinical renal injury, including increased levels of total urine protein in ∼29% of individuals and several urinary tubular injury markers (KIM-1, TFF3, β2MG, cystatin C, VCAM-1, and b-NAG) in 17–48.9% of PwCF, in the absence of albuminuria. The CF cohort was segregated into two subpopulations according to the differential urinary levels of kidney injury markers, potentially indicating two distinct mechanisms of injury. These were independent of total proteinuria or the diagnosis of diabetes. That only a subgroup of urinary injury markers (β2MG, cystatin C, and TFF3) associated with antibiotic course suggests urinary tubular injury in PwCF occurs more as a response to disease than treatment. Our studies reveal several novel associations that propose a model of lung-kidney pathogenic axis mediated by neutrophils in the setting of chronic P. aeruginosa lung infection: i) increased urinary levels of the tubular injury marker KIM-1 and increased prevalence of neutrophil in urine immune cells inversely correlate with lung function; ii) urinary levels of kidney injury markers positively correlate with urinary markers of neutrophil activation, and iii) both urinary levels of renal injury and neutrophil activation markers are preferentially elevated in PwCF with chronic P. aeruginosa, but not S. aureus, infection.

Though renal injury in a subpopulation of our CF cohort may have reflected aminoglycoside-triggered AKI at the time of sample collection, a distinct urinary protein signature of increased KIM-1, NGAL, VCAM-1, b-NAG, and TFF3 was independent of treatment. While urinary increases in these tubular proteins have been reported in both AKI and CKD, longitudinal studies are needed to define their role in predicting CKD in PwCF. Interestingly, increased urinary levels of these injury markers in PwCF correlated with higher levels of urinary calprotectin and NETs, markers of neutrophil activation, suggesting a relationship between inflammation and renal injury in a subset of PwCF. Neutrophils are known inflammatory responders and are abundant in the CF airway. In other settings of local injury, neutrophils have been shown to disseminate to other organs, including the kidney, where they mediate inflammatory and injury processes (e.g., neutrophils mediated increased Kim-1 and Ngal expression following skin sterile injury 21). In our studies, urinary NETs and KIM-1 were both strongly associated with chronic P. aeruginosa infection in the CF cohort, suggesting that infection-driven inflammation originating in the CF lung may stimulate systemic, including renal, neutrophil activation. This model of a lung-kidney pathogenic axis in CF is further supported by a strong inverse relationship between % FEV1 and urinary KIM-1 as well as %FEV1 and the level of neutrophils in urine cell pellets. Studies in ARDS and COPD demonstrate that inflammation in the lung can have both acute and chronic consequences on the kidney 15,36. Whether and by which mechanism neutrophils mediate renal injury in these diseases is unknown. The lack of an association between urinary NETs or Kim-1 and chronic S. aureus infection points to pathogen-specificity that may be driven by the differential effects of these two bacteria on neutrophils. S. aureus has been shown to inhibit neutrophil function and suppress the production of IL-8, a major neutrophil chemoattractant, which may dampen the overall neutrophil influx into the lung and consequently the systemic effects of neutrophils 37,38.

Another important contributor to the renal susceptibility to injury in PwCF, independently of and in the context of lung infection, may be aberrant CFTR function in the kidney. CFTR is highly expressed by proximal and distal tubular cells in the apical endosomes, where it plays a role in the endocytic uptake of low-molecular-weight (LMW) proteins 39–41. The defect in endocytosis of LMW proteins by the proximal tubular cells in the CFTR-deficient mice is attributed to decreased levels of cubilin 42. Therefore, the accumulation of proteins such as β2MG, TFF3, or NGAL, in the urine of PwCF may in part be a consequence of intrinsic renal CFTR defects. Studies of lung epithelial and endothelial cells have demonstrated increased expression of adhesion molecules, neutrophil chemoattractants (IL-8, IL-1b), and inflammatory cytokinesin the absence of a functional CFTR 43,44. Moreover, high peribronchial neutrophil infiltration in CF infants prior to any infection 45,46, suggests that renal endothelial and epithelial intrinsic defects in the CFTR may also contribute to neutrophil recruitment and activation in the kidney. We were not able to capture any CFTR mutation-specific effects due to the small number of non-Class II PwCF and the small number of patients treated with Trikafta in this retrospective analysis. However, even in a CF cohort predominantly treated with Trikafta (n = 9/11), we detected a wide range in neutrophil prevalence among urinary immune cells, which was strongly linked with lung function (Figure 4). These data suggest that renal damage might still occur in PwCF through neutrophil-mediated inflammatory mechanisms even after restoring normal CFTR function in the kidney.

The limitations of this study include the relatively small sample size and the cross-sectional design, which limit our ability to establish causality. Urine samples in CF cohort 1 and matched healthy controls were directly frozen and thus some of the proteins could be derived from cells in urine. Future studies with larger cohorts and longitudinal design are necessary to better elucidate the effect of treatment type and duration, lung disease flares, and co-morbidities on the underlying kidney injury in PwCF. Due to the size of the cohort, sex as a variable was not carefully considered and will be integrated into statistical analyses in future studies.

Nonetheless, our report highlights the presence of subclinical kidney injury in PwCF that may be secondary to Pseudomonal lung infection or the primary defect in the renal CFTR. Moreover, our studies identify a lung-kidney axis of injury in CF that may be driven by neutrophil activation in the context of chronic lung infection. These results contribute to a better understanding of kidney injury mechanisms that are not solely driven by aminoglycoside treatment and underscore the importance of identifying early urinary renal injury signatures in PwCF. Further research in this area is warranted to explore multiple pathogenic pathways, particularly in the era of HEMT.

Data Availability

All data produced in the present study are available upon reasonable request to the authors

Supplemental Material

View this table:
Supplemental Table 1: Clinical parameters of the CF cohort.
View this table:
Supplemental Table 2: CF Cohort 2 patient characteristics.
Supplemental Figure 1:
Supplemental Figure 1: Correlation between total urine protein and age

Supplemental Figure 1. Relationship between total urine protein/creatinine ratio and age in HC and CF cohorts. (A) Total urine protein levels were measured by Bradford assay and normalized to urine creatinine in healthy control (HC, n =33) and CF (CF, n= 48) cohorts. Pearson correlation analysis between total urinary protein normalized to urine creatinine and age.

Footnotes

  • ↵# co-first authors

References

  1. 1.↵
    Rowe, S. M., Miller, S. & Sorscher, E. J. mechanisms of disease Cystic Fibrosis. N Engl J Med vol. 352 www.nejm.org (2005).
  2. 2.↵
    Chen, Q., Shen, Y. & Zheng, J. A review of cystic fibrosis: Basic and clinical aspects. Animal Models and Experimental Medicine vol. 4 220–232 Preprint at doi:10.1002/ame2.12180 (2021).
    OpenUrlCrossRef
  3. 3.↵
    Quon, B. S., Mayer-Hamblett, N., Aitken, M. L., Smyth, A. R. & Goss, C. H. Risk factors for chronic kidney disease in adults with cystic fibrosis. Am J Respir Crit Care Med 184, 1147–1152 (2011).
    OpenUrlCrossRefPubMedWeb of Science
  4. 4.↵
    Lai, S. et al. Renal involvement and metabolic alterations in adults patients affected by cystic fibrosis. J Transl Med 17, (2019).
  5. 5.
    Santoro, D. et al. Cystic Fibrosis: A Risk Condition for Renal Disease. Journal of Renal Nutrition 27, 470–473 (2017).
    OpenUrl
  6. 6.↵
    Stevanovic, M. & Graber, M. Primary Causes of ESRD in the US Cystic Fibrosis Population. Journal of the American Society of Nephrology 33, (2022).
  7. 7.↵
    Steinkamp, G. et al. Renal function in Cystic Fibrosis: proteinuria and enzymuria before and after tobramycin therapy.
  8. 8.↵
    Al-Aloul, M. et al. Renal impairment in cystic fibrosis patients due to repeated intravenous aminoglycoside use. Pediatr Pulmonol 39, 15–20 (2005).
    OpenUrlCrossRefPubMedWeb of Science
  9. 9.↵
    Mingeot-Leclercq, M.-P. & Tulkens, P. M. Aminoglycosides: Nephrotoxicity. ANTIMICROBIAL AGENTS AND CHEMOTHERAPY vol. 43 (1999).
  10. 10.↵
    Steinkamp, G. et al. Renal function in Cystic Fibrosis: proteinuria and enzymuria before and after tobramycin therapy.
  11. 11.↵
    McWilliam, S. J., Antoine, D. J., Jorgensen, A. L., Smyth, R. L. & Pirmohamed, M. Urinary Biomarkers of Aminoglycoside-Induced Nephrotoxicity in Cystic Fibrosis: Kidney Injury Molecule-1 and Neutrophil Gelatinase-Associated Lipocalin. Sci Rep 8, (2018).
  12. 12.↵
    McWilliam, S. J., Antoine, D. J., Jorgensen, A. L., Smyth, R. L. & Pirmohamed, M. Urinary Biomarkers of Aminoglycoside-Induced Nephrotoxicity in Cystic Fibrosis: Kidney Injury Molecule-1 and Neutrophil Gelatinase-Associated Lipocalin. Sci Rep 8, (2018).
  13. 13.↵
    Park, B. D. & Faubel, S. Acute Kidney Injury and Acute Respiratory Distress Syndrome. Critical Care Clinics vol. 37 835–849 Preprint at doi:10.1016/j.ccc.2021.05.007 (2021).
    OpenUrlCrossRef
  14. 14.↵
    Komaru, Y., Bai, Y. Z., Kreisel, D. & Herrlich, A. Interorgan communication networks in the kidney– lung axis. Nat Rev Nephrol (2023) doi:10.1038/s41581-023-00760-7.
    OpenUrlCrossRef
  15. 15.↵
    Shayo, F. K. & Lutale, J. Albuminuria in patients with chronic obstructive pulmonary disease: A cross-sectional study in an African patient cohort. BMC Pulm Med 18, (2018).
  16. 16.↵
    Peng, X. et al. Protective effects of sphingosine 1-phosphate in murine endotoxin-induced inflammatory lung injury. Am J Respir Crit Care Med 169, 1245–1251 (2004).
    OpenUrlCrossRefPubMedWeb of Science
  17. 17.↵
    Laucirica, D. R., Garratt, L. W. & Kicic, A. Progress in Model Systems of Cystic Fibrosis Mucosal Inflammation to Understand Aberrant Neutrophil Activity. Frontiers in Immunology vol. 11 Preprint at doi:10.3389/fimmu.2020.00595 (2020).
    OpenUrlCrossRef
  18. 18.↵
    Downey, D. G., Bell, S. C. & Elborn, J. S. Neutrophils in cystic fibrosis. Thorax vol. 64 81–88 Preprint at doi:10.1136/thx.2007.082388 (2009).
    OpenUrlAbstract/FREE Full Text
  19. 19.
    Lyczak, J. B., Cannon, C. L. & Pier, G. B. Lung infections associated with cystic fibrosis. Clinical Microbiology Reviews vol. 15 194–222 Preprint at doi:10.1128/CMR.15.2.194-222.2002 (2002).
    OpenUrlAbstract/FREE Full Text
  20. 20.↵
    Gifford, A. M. & Chalmers, J. D. The role of neutrophils in cystic fibrosis. Current Opinion in Hematology vol. 21 16–22 Preprint at doi:10.1097/MOH.0000000000000009 (2014).
    OpenUrlCrossRefPubMed
  21. 21.↵
    Skopelja-Gardner, S. et al. Acute skin exposure to ultraviolet light triggers neutrophil-mediated kidney inflammation. doi:10.1073/pnas.2019097118/-/DCSupplemental.
    OpenUrlCrossRef
  22. 22.↵
    Wang, J. Neutrophils in tissue injury and repair. Cell and Tissue Research vol. 371 531–539 Preprint at doi:10.1007/s00441-017-2785-7 (2018).
    OpenUrlCrossRef
  23. 23.↵
    Shahin, W. et al. Early renal involvement in children with cystic fibrosis. European Respiratory Journal 52, PA4627 (2018).
  24. 24.↵
    Bach, M. et al. A Neutrophil Activation Biomarker Panel in Prognosis and Monitoring of Patients With Rheumatoid Arthritis. Arthritis and Rheumatology 72, 47–56 (2020).
    OpenUrl
  25. 25.↵
    Skopelja-Gardner, S. et al. Autoimmunity to bactericidal/permeability-increasing protein in bronchiectasis exhibits a requirement for Pseudomonas aeruginosa IgG response. European Respiratory Journal vol. 53 Preprint at doi:10.1183/13993003.01891-2018 (2019).
    OpenUrlAbstract/FREE Full Text
  26. 26.↵
    Zheng, S. C., Breeze, C. E., Beck, S. & Teschendorff, A. E. Identification of differentially methylated cell types in epigenome-wide association studies. Nat Methods 15, 1059–1066 (2018).
    OpenUrlCrossRef
  27. 27.↵
    Zhang, L.-H. et al. Early podocyte injury and elevated levels of urinary podocyte-derived extracellular vesicles in swine with metabolic syndrome: role of podocyte mitochondria. Am J Physiol Renal Physiol 317, 12–22 (2019).
    OpenUrl
  28. 28.↵
    Bollenbecker, S., Czaya, B., Gutierrez, O. M. & Krick, S. Lung-kidney interactions and their role in chronic kidney disease-associated pulmonary diseases. American Journal of Physiology - Lung Cellular and Molecular Physiology vol. 322 L625–L640 Preprint at doi:10.1152/AJPLUNG.00152.2021 (2022).
    OpenUrlCrossRef
  29. 29.↵
    Tirouvanziam, R., et al. Profound functional and signaling changes in viable inflammatory neutrophils homing to cystic fibrosis airways. www.pnas.org/cgi/content/full/ (2008).
  30. 30.↵
    Owen-Woods, C. et al. Local microvascular leakage promotes trafficking of activated neutrophils to remote organs. Journal of Clinical Investigation 130, 2301–2318 (2020).
    OpenUrl
  31. 31.↵
    Wang, J. et al. Visualizing the function and fate of neutrophils in sterile injury and repair. https://www.science.org.
  32. 32.↵
    Abedini, A. et al. Urinary single-cell profiling captures the cellular diversity of the kidney. Journal of the American Society of Nephrology 32, 614–627 (2021).
    OpenUrlAbstract/FREE Full Text
  33. 33.↵
    Fava, A. et al. Urine Proteomics and Renal Single-Cell Transcriptomics Implicate Interleukin-16 in Lupus Nephritis. Arthritis and Rheumatology 74, 829–839 (2022).
    OpenUrl
  34. 34.↵
    Zhang, Z. et al. HiTIMED: hierarchical tumor immune microenvironment epigenetic deconvolution for accurate cell type resolution in the tumor microenvironment using tumor-type-specific DNA methylation data. J Transl Med 20, (2022).
  35. 35.↵
    Salas, L. A. et al. Enhanced cell deconvolution of peripheral blood using DNA methylation for high-resolution immune profiling. Nat Commun 13, (2022).
  36. 36.↵
    Park, B. D. & Faubel, S. Acute Kidney Injury and Acute Respiratory Distress Syndrome. Critical Care Clinics vol. 37 835–849 Preprint at doi:10.1016/j.ccc.2021.05.007 (2021).
    OpenUrlCrossRef
  37. 37.↵
    Tomlinson, K. L. et al. Staphylococcus aureus stimulates neutrophil itaconate production that suppresses the oxidative burst. Cell Rep 42, (2023).
  38. 38.↵
    Zurek, O. W., Pallister, K. B. & Voyich, J. M. Staphylococcus aureus Inhibits Neutrophil-derived IL-8 to Promote Cell Death. Journal of Infectious Diseases 212, 934–938 (2015).
    OpenUrlCrossRefPubMed
  39. 39.↵
    Jouret, F. et al. Cystic fibrosis is associated with a defect in apical receptor-mediated endocytosis in mouse and human kidney. Journal of the American Society of Nephrology 18, 707–718 (2007).
    OpenUrlAbstract/FREE Full Text
  40. 40.
    Morales, M. M., Falkenstein, D. & Lopes, A. G. The Cystic Fibrosis Transmembrane Regulator (CFTR) in the Kidney *.
  41. 41.↵
    Souza-Menezes, J. & Morales, M. M. CFTR structure and function: Is there a role in the kidney? Biophysical Reviews vol. 1 3–12 Preprint at doi:10.1007/s12551-008-0002-3 (2009).
    OpenUrlCrossRef
  42. 42.↵
    Jouret, F. et al. Cystic fibrosis is associated with a defect in apical receptor-mediated endocytosis in mouse and human kidney. Journal of the American Society of Nephrology 18, 707–718 (2007).
    OpenUrlAbstract/FREE Full Text
  43. 43.↵
    Cabrini, G. et al. Role of Cystic Fibrosis Bronchial Epithelium in Neutrophil Chemotaxis. Frontiers in Immunology vol. 11 Preprint at doi:10.3389/fimmu.2020.01438 (2020).
    OpenUrlCrossRef
  44. 44.↵
    Hunter, M. J. et al. Expression of wild-type CFTR suppresses NF-κB-driven inflammatory signalling. PLoS One 5, (2010).
  45. 45.↵
    Margaroli, C. et al. Elastase exocytosis by airway neutrophils is associated with early lung damage in children with cystic fibrosis. Am J Respir Crit Care Med 199, 873–881 (2019).
    OpenUrlPubMed
  46. 46.↵
    Chandler, J. D. et al. Myeloperoxidase oxidation of methionine associates with early cystic fibrosis lung disease. European Respiratory Journal 52, (2018).
Back to top
Posted November 11, 2023.
Download PDF
Data/Code
Email
Early Urinary Markers of Kidney Injury in Adult People with Cystic Fibrosis Correlate with Neutrophil Activation and Lung Infection
Grace M. Rosner, Himanshu B. Goswami, Katherine Sessions, Lindsay K. Mendyka, Brenna Kerin, Irma Vlasac, Diane Mellinger, Lorraine Gwilt, Thomas H. Hampton, Martha Graber, Alix Ashare, William T. Harris, Brock Christensen, Agnieszka Swiatecka-Urban, Bruce A. Stanton, Sladjana Skopelja-Gardner
medRxiv 2023.11.10.23298378; doi: https://doi.org/10.1101/2023.11.10.23298378
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area