|Home | About | Journals | Submit | Contact Us | Français|
To determine whether urine ubiquitin levels are elevated after burns and to assess whether urine ubiquitin could be useful as a non-invasive biomarker for burn patients.
Forty burn patients (%TBSA: 20±22; modified Baux scores: 73±26) were included (control: 11 volunteers). Urine was collected in 2h-intervals for 72h, followed by 12h-intervals until discharge from the ICU. Ubiquitin concentrations were analyzed by ELISA and Western blot. Total protein was determined with a Bradford assay. Patient characteristics and clinical parameters were documented. Urine ubiquitin concentrations, renal ubiquitin excretion and excretion rates were correlated with patient characteristics and outcomes.
Initial urine ubiquitin concentrations were 362±575 ng/mL in patients and 14±18 ng/mL in volunteers (p<0.01). Renal ubiquitin excretion on day-1 was 292.6±510.8 μg/24h and 21±27 μg/24h in volunteers (p<0.01). Initial ubiquitin concentrations correlated with modified Baux scores (r=0.46; p=0.02). Ubiquitin levels peaked at day 6 post-burn, whereas total protein concentrations and serum creatinine levels remained within the normal range. Total renal ubiquitin excretion and excretion rates were higher in patients with %TBSA ≥20 than with %TBSA <20, in patients who developed sepsis/MOF than in patients without these complications and in non-survivors vs. survivors.
Ubiquitin urine levels are significantly increased after burns. Renal ubiquitin excretion and/or excretion rates are associated with %TBSA, sepsis/MOF and mortality. Although these findings may explain previous correlations between systemic ubiquitin levels and outcomes after burns, the large variability of ubiquitin urine levels suggests that urine ubiquitin will not be useful as a non-invasive disease biomarker.
Despite major advances in the understanding of the pathophysiology of burn injury and improvements in surgical and intensive care, risk stratification of burn patients has remained unchanged throughout decades (1, 2). While potential biomarkers of burn injury have been proposed, none of these molecules has been validated to improve the traditional risk stratification, which is based on age, percent of burned total body surface area (%TBSA) and presence of inhalation injury (1, 3–9). As new diagnostic or prognostic biomarkers for burn patients, however, could facilitate further reduction of burn associated mortality and morbidity, identification of such molecules could be of significant clinical value (10).
Whereas invasive biomarkers require careful consideration of their risk-benefit ratios, identification of non-invasive and cost effective biomarkers would have the advantage that even incremental advances in predicting outcomes after burns and in diagnosing post-burn complications could benefit the patient. Accordingly, several previous studies have evaluated urine as a source for non-invasive biomarkers and some correlations between urine concentrations of various molecules, such as albumin, lipid peroxidation products or neutrophil gelatinase-associated lipocalin, and outcomes after burns have been established (11–16).
Ubiquitin is a small and highly conserved protein in all eukaryotes (17, 18). Besides its intracellular function as a post-translational protein modifier (17–19), ubiquitin is also a natural plasma protein which functions as an immune modulator and non-cognate agonist of the chemokine (C-X-C motif) receptor (CXCR) 4 (20–31). Systemic ubiquitin levels are elevated in various diseases, such as sepsis, trauma, burns or myocardial infarction (22). We have shown previously that serum ubiquitin concentrations correlate with the burn size in patients, which is in agreement with the passive release of damage-associated molecular patterns (DAMP) from damaged cells and tissues after burns (22, 32–36). In contrast to systemic levels of the prototypical DAMPs high mobility group box 1 (HMGB1) and S100A8/9 after burns, however, we observed that patients who develop multiple organ failure or die have a relative deficiency of serum ubiquitin during the first week after injury (32, 33, 36). In combination with the anti-inflammatory and beneficial effects of ubiquitin treatment that have been observed in multiple animal models and species (37–43), these data suggest a protective function of endogenous ubiquitin in the systemic circulation.
Ubiquitin is also detectable in small concentrations in other extracellular body fluids, such as cerebrospinal fluid, lung alveolar lining fluid and urine (22). While increased ubiquitin concentrations have been described in bronchoalveolar lavage fluid after trauma, burn and inhalation injury in animals and patients, increased ubiquitin urine concentrations have been observed in sepsis patients (21, 44, 45). Ubiquitin urine concentrations in patients after burn injuries, however, have not been studied. As the molecular weight of ubiquitin (8.5 kDa) is well below the molecular weight cutoff (30–50 kDa) for glomerular filtration (46, 47), we hypothesized that ubiquitin urine concentrations are also elevated after burn injury and may correlate with the size of the burn injury and patient outcomes. Thus, we performed a prospective observational study to evaluate whether urine ubiquitin levels are elevated after burn injury in patients and may be useful as a new non-invasive disease biomarker.
This study was approved by the Institutional Review Board at Loyola University Chicago. Forty consecutive burn patients (age: 45.5 ± 18.3 years (mean ± SD), 62.5% male) who were admitted to the burn intensive care unit (ICU) between September 2011 and August 2012 were enrolled in the study. Informed consent was obtained from all participants. Burn patients admitted to the burn ICU younger than 18 years of age were excluded from the study. Injury severity was determined based on the percentage of total body surface area (% TBSA) partial and/or full thickness burns, as documented in the patients’ medical records. Patients with suspected inhalation injury underwent diagnostic bronchoscopy and were assigned a grade of 0–4 based on the appearance of the endobronchial lumen during the procedure (grade 0: no injury, absence of carbonaceous deposits, erythema, edema, bronchorrhea, or obstruction; grade 1: mild injury, minor or patchy areas of erythema, carbonaceous deposits in proximal or distal bronchi; grade 2: moderate injury, moderate degree of erythema, carbonaceous deposits, bronchorrhea, with or without compromise of the bronchi; grade 3: severe injury, severe inflammation with friability, copious carbonaceous deposits, bronchorrhea, bronchial obstruction; grade 4: massive injury, evidence of mucosal sloughing, necrosis, endoluminal obliteration), as described (48). All patients received the standard of care based on the Loyola ICU protocols. Fluid resuscitation was performed according to the Parkland formula (4 mL/kg/% TBSA with half given during the first 8 hours of admission and the remaining half given over the next 16 hours).
The following clinical characteristics and outcomes were obtained from electronic medical records and entered into a database: age, gender, length of stay, pre-existing kidney diseases, % TBSA, Baux score (sum of age and % TBSA), modified Baux score (sum of age and % TBSA plus 17 if positive for inhalation injury), grade of inhalation injury, sepsis and/or multiple organ failure (MOF), use of vasoactive agents, urine output and mortality. Sepsis and MOF were diagnosed using the ACCP/SCCM Consensus Conference criteria (49).
Urine was collected at the time of admission to the ICU and then every two hours for 72 hours. Thereafter, urine was collected in 12 h intervals until discharge from the burn ICU. Because urine samples could not be obtained from each patient at each individual time point, the exact number of analyzed urine samples is provided in the Results section. Urine from eleven healthy volunteers without pre-existing kidney disease served as a normal reference. Urine samples were centrifuged at 250 x g for five minutes and the supernatants were stored at −80°C until further analysis.
Ubiquitin urine concentrations were quantified with a competitive ELISA, in which biotinylated ubiquitin and ubiquitin in the test sample compete for a limited number of binding sites in the anti-ubiquitin antibody, as described (32, 44, 50). In brief, microtiter plates were coated with anti-ubiquitin (Sigma-Aldrich, St. Louis, MO) and incubated for 18 h at 4°C. The plates were washed three times with 0.05 % tween 20 in phosphate buffered saline and were incubated with blocking buffer (1 % bovine serum albumin (Sigma-Aldrich), 0.05 % tween 20, in phosphate buffered saline) for 1 h. After washing three times, 50 μl of the standards or samples were mixed with 50 μl of biotinylated ubiquitin (Boston Biochem, Boston, MA) and placed in the plates. Each sample was tested in eight dilutions. Dilutions for the standard curve and the test samples were prepared in blocking buffer. After incubation for 2 h the plates were washed again and a peroxidase-labeled anti-biotin antibody (Amersham Biosciences, Buckinghamshire, UK) was added. After incubation for 2 h the plates were washed again and 100 μL TMB ELISA solution (Sigma-Aldrich) was added. After incubation for 10 – 20 min, the reaction was stopped by addition of 100 μL 2N HCl and optical densities were measured using a micro-ELISA autoreader (Synergy 2, Bio-Tek Instruments Inc., Winooski, VT; test filter: 450 nm; reference filter: 540 nm). The ubiquitin concentration in the test sample was calculated with the KC4 for windows program, Version 3.02 (Bio-Tek Instruments Inc.), from a four parameter logistic fit employing ubiquitin as standard (0 – 1700 ng/mL, Sigma-Aldrich).
Western blotting with mouse anti-ubiquitin (LifeSensors, Malvern, PA; 1:1000) in combination with horseradish peroxidase conjugated anti-mouse-IgG (GE Healthcare, Burr Ridge, IL; 1:5000) was performed as described (51, 52). Chemiluminescence signals were detected with a Chemidoc imaging system (BioRad).
Total protein concentrations in urine samples were measured with the Quick Start Bradford protein assay kit (Bio-Rad, Hercules, CA) using bovine serum albumin as a standard.
Data are described as mean with standard deviation. Data were analyzed with locally weighted scatterplot smoothing (LOWESS) analyses, Spearman correlation analyses (rSpearman), linear regression analyses, Fisher’s exact test, Mann-Whitney U test and two-way analyses of variance (ANOVA) with Bonferroni post-tests, respectively. A two-tailed p < 0.05 was considered significant. All statistical analyses were calculated with the GraphPad Prism 6 program (GraphPad Software, Inc., La Jolla, CA).
The clinical characteristics of the patients are summarized in Table 1. Twenty six patients sustained flame injury and five patients scald burn. Two of the patients sustained a combination of scald and chemical or scald and flame burn. Seven of the patients had inhalation injury without cutaneous burn injury. One patient (inhalation injury grade 3, 0 % TBSA) had preexisting kidney disease; only three urine samples could be obtained from this patient. There were no statistically significant differences in age and gender distribution between healthy volunteers (age: 34.1 ± 8.2 years, 36.4% male, p>0.05 vs. patients) and burn patients.
An initial urine sample on admission to the ICU could be obtained from 25 patients. In these specimens, the average total protein concentration was 200 ± 270 μg/mL. As shown in Fig. 1A, average ubiquitin urine concentrations in patients on admission to the ICU were 26-fold increased, as compared with ubiquitin urine concentrations in healthy volunteers (patients − 362 ± 575 ng/mL; volunteers − 14 ± 18 ng/mL; p<0.01). The average ubiquitin excretion on day 1 in burn patients was 292.6 ± 510.8 μg/24 h, which is 15-fold higher than the estimated daily urine ubiquitin excretion in healthy volunteers (21 ± 27 μg/24 h, p<0.01 vs. burn patients) when an average daily urine volume of 1500 mL is assumed (Fig. 1B).
To confirm the high urine ubiquitin concentrations in patients with a second method, we then performed Western blot analyses of urine samples with anti-ubiquitin. Consistent with the ELISA measurements, we detected a strong band corresponding to free ubiquitin in patient’s urine, whereas no bands or bands with only minimal intensities could be observed in urine from healthy volunteers (Fig. 1C).
To assess whether initial ubiquitin urine levels could have diagnostic or prognostic value after burns, we correlated and compared ubiquitin urine concentrations on admission and the total amount of ubiquitin excreted on day 1 among various clinical parameters. There were no significant correlations between urine ubiquitin concentrations on admission and patient’s age (rSpearman = 0.12), %TBSA (rSpearman= 0.31), grade of inhalation injury (rSpearman = 0.02), number of days with mechanical ventilation (rSpearman = −0.03) or length (days) of ICU treatment (rSpearman = −0.013) (p > 0.05 for all). Ubiquitin urine concentrations on admission, however, correlated significantly with the modified Baux score (Fig. 2A; rSpearman = 0.46; p = 0.02). Although ubiquitin urine concentrations on admission were higher in patients with %TBSA ≥ 20 (151 ± 233 ng/mL) than in patients with %TBSA < 20 (629 ± 763 ng/mL), this difference did not reach statistical significance (Fig. 2B; p = 0.08). However, initial ubiquitin urine concentrations in patients with a modified Baux score ≥ 60 (503 ± 651 ng/mL) were significantly higher than ubiquitin urine concentrations in patients with a modified Baux score < 60 (61 ± 110 ng/mL; p = 0.016) (Fig. 2C). There were no statistically significant differences in ubiquitin urine concentrations on admission when patients were grouped according to presence or absence of inhalation injury, vasopressor treatment, development of sepsis/MOF and survival or non-survival (Fig. 2D–G).
The total average amounts of ubiquitin excreted in urine on day 1 (n = 33) did not correlate significantly with age (rSpearman= −0.019), %TBSA (rSpearman = =0.03), modified Baux scores (rSpearman = 0.112), grade of inhalation injury (rSpearman = 0.26), length of ICU treatment (rSpearman = 0.015) or ventilator days (rSpearman = 0.33) (p>0.05 for all). Furthermore, there were no statistically significant differences when the urine ubiquitin excretion on day 1 was compared between patients grouped according to the various clinical parameters (Fig. 3).
Fig. 4A shows the time course of the average urine total protein and ubiquitin concentrations in patients throughout the observation period. Within the first ten days after burn injury, the average total protein concentrations remained within the normal range. Elevated urine total protein concentrations, however, were detectable at later time points (> day 10 post burn).
As compared with average ubiquitin urine concentrations in healthy volunteers, average ubiquitin urine concentrations in burn patients remained elevated throughout the observation period. Average ubiquitin urine concentrations in burn patients transiently decreased during initial fluid resuscitation to 127 ± 209 ng/mL at 6 h after admission to the ICU. Thereafter, urine ubiquitin concentrations increased 9–10-fold with peak concentrations at post-burn days 5–6. and subsequently decreased towards post-burn day 1 concentrations by post-burn day 10. After day 10 post burn, however, total protein and ubiquitin urine concentrations showed a very similar time course profile. Accordingly, LOWESS curve analyses of the ubiquitin/total protein ratio (ng/mg) in urine showed a monophasic increase of the ubiquitin/total protein ratio with peak values at day 6 post burn (Fig. 4B). Interestingly, LOWESS curve analyses of the average plasma creatinine levels showed a time course profile within 10 days post burn inverse to the urine ubiquitin/total protein ratio. After day 10 post-burn, however, plasma creatinine concentrations and urine ubiquitin/total protein ratios followed a similar time course pattern (Fig. 4B).
The average urine output and the cumulative ubiquitin excretion after burn injury throughout the observation period are shown in Fig. 4C. After initial fluid resuscitation, urine output remained constant between 50–100 mL/h. Within post-burn days 1–10, patients excreted on average 0.98 ± 0.04 mg of ubiquitin/24 h in the urine. Thereafter, the average daily urine ubiquitin excretion was 0.32 ± 0.16 mg/24 h.
The cumulative urine ubiquitin excretion in patients that were grouped based on %TBSA, sepsis/MOF development and survival are shown in Fig. 5A–C. Patients with %TBSA < 20 excreted significantly less ubiquitin than patients with %TBSA ≥ 20. The average ubiquitin excretion rates, which were calculated from linear regression analyses, were 18 ± 0.7 μg/h and 46 ± 2 μg/h in patients with %TBSA <20 and ≥20, respectively (p<0.01). Similarly, patients who developed sepsis/MOF excreted significantly more ubiquitin in urine, as compared with patients who did not develop these complications (Fig. 5B). The average ubiquitin excretion rate was 50 ± 2.7 μg/h in patients with development of sepsis/MOF and 22 ± 0.7 μg/h in patients with uncomplicated clinical course (p<0.01). Because the %TBSA was significantly higher in patients who developed sepsis/MOF (age – 46 ± 15 years; %TBSA – 39 ± 25%, p = 0.0092 vs. patients with uncomplicated recovery) than in patients with uncomplicated recovery (age – 45 ± 19 years; %TBSA – 16 ± 19%), we also compared ubiquitin excretion rates when patients with sepsis/MOF development were matched with patients with uncomplicated recovery based on age and %TBSA (n = 9; age – 48 ± 15 years, %TBSA – 42 ± 26%; p>0.05 vs. patients with sepsis/MOF). In these matched burn patients without development of sepsis/MOF, the average ubiquitin excretion rate was 29 ± 1.6 μg/h (p < 0.01 vs. patients with sepsis/MOF development). Similarly, ubiquitin excretion rates were 34 ± 2 μg/h and 25 ± 1 μg/h in burn patients with and without vasopressor treatment, respectively (p<0.01, not shown). Furthermore, burn patients who died excreted significantly more ubiquitin in urine than burn patients who survived (Fig. 5C). In contrast to surviving burn patients, in which ubiquitin excretion followed a (pseudo)linear or logarithmic trend, patients who died showed an exponential increase in urine ubiquitin excretion (Fig. 5C).
In the present study, we provide the initial description of ubiquitin urine levels in burn patients. There are several new findings from the present study. First, ubiquitin urine levels are significantly increased after burn injury. Second, average ubiquitin excretion and excretion rates are associated with the size of the burn injury and patient outcomes. Third, the variation of ubiquitin urine levels after burns is large and individual measurements of ubiquitin urine concentrations, ubiquitin excretion and ubiquitin excretion rates do not discriminate patient characteristics or outcomes.
The urine ubiquitin concentrations that we determined in healthy volunteers in the present study are in agreement with the previously described presence of small amounts of ubiquitin in normal human urine (21, 53–55). Although the total protein concentrations in urine from burn patients on admission to the ICU were within the normal range (56), we measured significantly increased urine ubiquitin concentrations by ELISA and confirmed this finding by Western blot analyses.
We have shown previously that systemic ubiquitin concentrations in burn patients on admission to the hospital are increased several fold and remain elevated for at least 7 days post-burn (32). As the molecular mass of ubiquitin permits glomerular filtration (46, 47), the increased ubiquitin urine concentrations can be explained as a result of the increased ubiquitin concentrations that have been detected in the systemic circulation after burns. The findings that plasma creatinine levels and total urine protein concentrations on admission were within the normal range in burn patients in the present study suggest that possible changes of the glomerular filter or the proximal tubule scavenging system are less likely to account for the observed increase in urine ubiquitin concentrations (46).
Although the exact cellular origin of ubiquitin in the systemic circulation remains to be determined, passive release of ubiquitin from damaged tissues is currently the most likely source of increased levels that have been observed in various diseases, such as trauma, sepsis or burns (22). Thus, the high ubiquitin/total protein ratio in urine and the high ubiquitin excretion rates within the first 6–10 days after burn injury may reflect ongoing tissue and cell damage. This assumption is further supported by the exponential increase in urine ubiquitin excretion in non-surviving burn patients, which likely reflects extensive cellular injury and cell death.
The observation from the present study that burn patients who develop sepsis/MOF or die excrete significantly more ubiquitin in the urine and have significantly increased average renal ubiquitin excretion rates could explain our previous finding that these patients have a relative deficiency of ubiquitin in the systemic circulation, as compared with %TBSA matched patients who recovered uneventfully (32). The mechanisms underlying increased ubiquitin excretion in patients with poor outcomes after burns, however, remain unknown. It is also unknown to which degree glomerular filtered ubiquitin can be reabsorbed in the proximal tubule. Nevertheless, it could be speculated that differences in the transport capacity of multi-ligand receptors in the proximal tubule, such as megalin or cubilin, may account for the observed differences in ubiquitin excretion (46, 57). Furthermore, chemokine receptor CXCR4 is expressed in the proximale tubule and highly upregulated after renal injury (58, 59). As ubiquitin binding to CXCR4 leads to receptor mediated endocytosis of the receptor-ligand complex (23, 27, 29, 60), changes in the expression level of CXCR4 in the kidney could also contribute to the observed differences in ubiquitin excretion between burn patients with uncomplicated recovery and those who develop sepsis/MOF or die.
As the number of patients who developed post-burn complications or died in the present study was small, it is possible that statistically significant correlations or group differences among the analyzed parameters may become evident in a larger patient population. However, we have quantified ubiquitin urine levels in numerous sequential urine specimens obtained in 2–12 hour intervals. As these measurements showed large intra- and inter-individual variations, it is unlikely that measurements of isolated ubiquitin urine concentrations or excretion rates could be useful as a disease biomarker in the clinical arena, even if significant correlations could be established in a larger patient population.
In conclusion, ubiquitin urine concentrations and renal ubiquitin excretion are significantly increased after burn injury. Average renal ubiquitin excretion and ubiquitin excretion rates are associated with the burn size, development of sepsis/MOF and survival. These findings may explain the previously observed correlations between serum ubiquitin levels, %TBSA and outcomes in burn patients. The large variability of ubiquitin urine levels in the present study, however, suggests that measurements of ubiquitin urine levels will not be useful as a non-invasive disease biomarker.
This research was supported in part by National Institutes of Health Grant NIHT32GM008750 and the Dr. Ralph and Marian Falk Medical Research Trust.
Presented, in part, at the 17th Congress of the International Society for Burn Injuries, Sydney, Australia, October 2014.