Skip to main content
PMC home page
PLOS ONE logoLink to PLOS ONE
. 2020 Apr 21;15(4):e0231555. doi: 10.1371/journal.pone.0231555

Dysregulated activities of proline-specific enzymes in septic shock patients (sepsis-2)

Gwendolyn Vliegen 1, Kaat Kehoe 1, An Bracke 1, Emilie De Hert 1, Robert Verkerk 1, Erik Fransen 2, Bart ‘s Jongers 3, Esther Peters 4, Anne-Marie Lambeir 1, Samir Kumar-Singh 3, Peter Pickkers 4, Philippe G Jorens 5, Ingrid De Meester 1,*
Editor: Paul Proost6
PMCID: PMC7173796  PMID: 32315321

Abstract

The proline-specific enzymes dipeptidyl peptidase 4 (DPP4), prolylcarboxypeptidase (PRCP), fibroblast activation protein α (FAP) and prolyl oligopeptidase (PREP) are known for their involvement in the immune system and blood pressure regulation. Only very limited information is currently available on their enzymatic activity and possible involvement in patients with sepsis and septic-shock. The activity of the enzymes was measured in EDTA-plasma of patients admitted to the intensive care unit (ICU): 40 septic shock patients (sepsis-2) and 22 ICU control patients after major intracranial surgery. These data were used to generate receiver operating characteristic (ROC) curves. A survival analysis (at 90 days) and an association study with other parameters was performed. PRCP (day 1) and PREP (all days) enzymatic activities were higher in septic shock patients compared to controls. In contrast, FAP and DPP4 were lower in these patients on all studied time points. Since large differences were found, ROC curves were generated and these yielded area under the curve (AUC) values for PREP, FAP and DPP4 of 0.88 (CI: 0.80–0.96), 0.94 (CI: 0.89–0.99) and 0.86 (CI: 0.77–0.95), respectively. PRCP had a lower predicting value with an AUC of 0.71 (CI: 0.58–0.83). A nominally significant association was observed between survival and the DPP4 enzymatic activity at day 1 (p<0.05), with a higher DPP4 activity being associated with an increase in survival. All four enzymes were dysregulated in septic shock patients. DPP4, FAP and PREP are good in discriminating between septic shock patients and ICU controls and should be further explored to see whether they are already dysregulated in earlier stages, opening perspectives for their further investigation as biomarkers in sepsis. DPP4 also shows potential as a prognostic biomarker. Additionally, the associations found warrant further research.

1. Introduction

Sepsis remains a major problem in the intensive care unit (ICU), where approximately one third of admitted patients is diagnosed with sepsis [1]. Sepsis is currently defined as a life-threatening organ dysfunction caused by a dysregulated host response to infection and septic shock is defined as a subset of sepsis in which particularly profound circulatory, cellular, and metabolic abnormalities are associated with a greater risk of mortality than with sepsis alone [2]. It is now acknowledged that sepsis involves both pro- and anti-inflammatory responses in combination with alterations in other immunologic and non-immunologic pathways [2–5]. To date, there is no standard diagnostic test that can identify patients with sepsis accurately, although early identification is necessary to improve their outcome [6].

The proline-specific peptidases studied herein are dipeptidyl peptidase 4 (DPP4; EC 3.4.14.5), prolylcarboxypeptidase (PRCP; EC 3.4.16.2), fibroblast activation protein α (FAP; 3.4.21.B28) and prolyl oligopeptidase (PREP; EC 3.4.21.26). All four enzymes are proline-specific preferentially cleaving peptides after proline residues. However, the exact position of the proline in the peptide and the preferred substrates differ between the enzymes. In addition, they share structural properties and therefore belong to related peptidase families, namely S9 (PREP, FAP and DPP4) and S28 (PRCP) [7,8]. Here, we studied the enzymatic activities of DPP4, PRCP, FAP and PREP in the plasma of patients with septic shock (sepsis-2), since they are known for their involvement in the immune system and blood pressure regulation, important factors playing a key role in sepsis, as will be discussed below. Additionally, they are easily measured in plasma or serum with specific assays. Apart from one study on DPP4 activity in severe sepsis patients, the activity of these enzymes in sepsis patients have not been reported in the past. However, it is conceivable that they could be involved in pathogenesis of sepsis and septic shock, as will be outlined below.

DPP4 activity has been shown to be decreased in the serum of patients with sepsis compared to healthy controls [9]. DPP4, also known as CD26, is a co-stimulatory molecule for T-cells and has several substrates that can be involved in sepsis. Examples are neuropeptide Y [10], glucagon-like peptide-1 [11–13], procalcitonin [14,15] and several chemokines [16]. A nested case-control study in type 2 diabetes patients admitted for sepsis did not find a significant association between the use of a DPP4-inhibitor and the development of sepsis [17]. This is reassuring from the therapeutic point of view but tells us nothing about a possible value of DPP4 activity in plasma as a biomarker for sepsis.

FAP, a structurally related ectopeptidase, is, under normal conditions, mostly absent from adult tissues. However, a soluble form can be found in the blood [18]. In a baboon model of sepsis, gene expression of FAP in the lung was maximal at 24 hours post E. coli challenge, probably reflecting the active tissue remodeling [19]. FAP also cleaves several substrates, such as α2-antiplasmin and fibroblast growth factor (FGF) 21. Interfering with the cleavage of α2-antiplasmin by FAP might enhance its thrombolytic activity [20] and it has been reported to be increased in patients with septic melioidosis compared to healthy controls [21]. FGF21 has been shown to be increased in patients with sepsis compared to healthy controls and decreased with clinical improvement [22].

The third enzyme studied here is PRCP, Erdös and co-workers reported an increased PRCP activity in the plasma of dogs 20 minutes after endotoxin injection [23]. PRCP has a dual position in the kallikrein-kinin and renin-angiotensin system, by activating prekallikrein and hydrolyzing angiotensin II to form angiotensin (1–7) [24–26]. PRCP is involved in the regulation of blood pressure and hypotension is common in sepsis and septic shock, moreover, angiotensin II is recently approved as therapy for distributive shock. Therefore, a possible role for PRCP in the pathogenesis of sepsis and septic shock is conceivable [4,27,28].

To the best of our knowledge, there are no reports on PREP in sepsis or septic shock, but since it shares several substrates with PRCP, it could be implicated in sepsis and septic shock [29–32]. In addition, PREP shares structural properties and substrate specificities with the other proline-specific peptidases and is therefore included in the study.

The goals of this study were first to evaluate the activities of these enzymes in patients with septic shock (sepsis-2). Since large differences between the septic shock and the ICU control patients were found, ROC curves were generated to test whether these enzymes should be further explored as diagnostic biomarkers. Secondly, the potential of these enzymes as prognostic biomarkers of survival at day 90 in septic shock was evaluated. By exploring associations between the enzymes and a variety of inflammatory, hemodynamic, metabolic parameters, measured on the same days, we additionally aimed to deepen our insights in their possible involvement in septic shock.

2. Materials and methods

2.1. Patient samples

This prospective cohort study was conducted in patients with septic shock (≥ 18 years) admitted to the ICU of the Radboud University Medical Center (Radboudumc, Nijmegen, the Netherlands). Septic shock was defined according to the definitions stated by ACCP and SCCM consensus conference: a suspected infection, two or more systemic inflammatory response syndrome (SIRS) criteria and the need for vasopressor therapy (Sepsis-2) [33]. All patients received standard of care according to the surviving sepsis campaign guidelines [6,34]. A group of 40 phenotypically well-characterized patients in whom samples were available at all 4 time points (days 1, 3, 5 and 7 after diagnosis) was used. The non-septic shock ICU control group consisted of 22 consecutively admitted patients (≥ 18 years) undergoing major intracranial surgery (resection of a cerebral tumor or clipping of an aneurysm) who were admitted to the ICU of the Antwerp University Hospital for postoperative monitoring. Collection of samples from septic shock patients was done in accordance with the applicable rules concerning the review of the Ethics Committee of UMC Radboud (CMO-nr: 2016/2923) and informed consent was given by the patient or his/her closest relative. All ICU control patients gave written informed consent and the study was approved by the Ethics Committee of the Antwerp University Hospital/ University of Antwerp (Amendment 17/10/119 ref. B300201732219). The study was performed in accordance with the Declaration of Helsinki. Additionally, 30 healthy controls were included, more information on this study group can be found in S1 Appendix and S1 Fig.

2.2. Blood sampling

Ethylenediaminetetraacetic acid (EDTA) and lithium heparin-anticoagulated blood from septic shock patients was collected from the arterial catheter within 24 h after diagnosis of septic shock (day 1), and on days 3, 5 and 7. Blood was centrifuged (1600 x g, 4 °C, 10 min) and the plasma was stored at -80 °C until further analysis. Demographic, clinical and laboratory data were collected on the days of blood sampling. EDTA-plasma from the ICU control group was collected the morning after surgery. All samples from the different groups were handled and stored identically as in the septic shock group.

2.3. Enzyme activity measurements

Enzyme activities were measured using in-house validated assays in EDTA-plasma samples. The enzyme activities are expressed in units per liter (U/L) where one unit defines the amount of enzyme that hydrolyses 1 μmol of substrate per minute under the given assay conditions. Buffers were pH adjusted at room temperature.

DPP4 activity was measured colorimetrically using the substrate glycyl-prolyl-paranitroanilide (Gly-Pro-pNA; Bachem, Bubendorf, Switzerland) as described earlier [35]. The release of pNA from the substrate was measured kinetically at 405 nm during 10 minutes at 37 °C, pH 8.3, using the Infinite™ 200 (Tecan Trading AG, Switzerland).

FAP activities were measured using N-benzyloxycarbonyl-Gly-Pro-7-amido-4-methylcoumarine (Z-Gly-Pro-AMC; Bachem, Bubendorf, Switzerland) as a substrate. Samples were diluted 26 times in a 0.1 M Tris buffer pH 8.0 containing 300 mM NaF, 1 mM NaN3, 1 mM EDTA and 50 mM salicylic acid. 6 μL of 4.6 mM substrate in methanol was added to 150 μL of diluted sample and incubated for 2 h at 37 °C. Reactions were stopped with 500 μL of 1.5 M acetic acid and fluorescence was measured (λex = 370 nm, λem = 440 nm) on the Shimadzu Fluorimeter RF-5000 (Den Bosch, The Netherlands) [36,37]. Concentration of the generated AMC was determined by means of a standard curve.

A reversed-phase high-performance liquid chromatography technique was used to measure the PRCP activity in the samples as described earlier [38]. PRCP activity was determined by measuring the hydrolysis of N-benzyloxycarbonyl-Pro-Phe (Z-Pro-Phe; Bachem, Bubendorf, Switzerland). Plasma samples were incubated for 2 h with Z-Pro-Phe at 37 °C, subsequently stop solution (10% perchloric acid and 20% acetonitrile solution in purified water (v/v)) was added to stop the enzymatic reaction. The enzymatically formed Z-Pro was tracked by its UV absorbance at 210 nm after separation on a Shimadzu HPLC apparatus equipped with a Chromolith C18e 100x3 mm column, a LC-20AT pump, a SIL-20AC HT autosampler and SPD-20 UV-VIS detector (Shimadzu, Den Bosch, The Netherlands). Quantification was performed by peak height measurements.

PREP in plasma was measured after activation with dithiothreitol (DTT) using the fluorogenic substrate Z-Gly-Pro-AMC as described before with some modifications [37,39]. To 20 μL of plasma, 100 μL of activation buffer containing 100 mM potassium phosphate buffer pH 8.0, 1 mM NaN3, 1 mM EDTA and 10 mM DTT was added. As a blank, 100 μL of the same buffer without DTT was added to 20 μL of sample. 6 μL of 4.6 mM substrate in dimethyl sulfoxide was added to the blank and activation sample tubes followed by incubation for 2 h at 37 °C. The reaction was stopped with 500 μL of 1.5 M acetic acid and fluorescence was measured as described above for FAP activity.

2.4. Statistical analysis

Data are represented as median with the interquartile range (IQR), unless otherwise stated.

Comparison of enzyme activities

The comparison of enzyme activities between the control group and the different time points in the cases group was carried out using a linear mixed model with ‘day’ as fixed effect having five levels (day 1, 3, 5, 7 and controls). A random intercept term for individual was added, to account for the non-independence between observations from the same individual. To compare the control level with the four measurements in the patients, a post-hoc analysis was carried out with Dunnett’s correction for multiple testing. In addition, changes in mean enzymatic activity between the different days were tested for significance in the patients, using a post-hoc analysis with Tukey’s correction for multiple comparisons. For PREP, the test was carried out on the log-transformed enzyme activity due to the non-normality of the residuals.

Receiver operating characteristic (ROC) curves

ROC curves were computed to assess the four enzymes as possible biomarkers for septic shock. For the septic shock patients (n = 40) only day 1 was used. The patient’s condition on day 1 is clinically the most relevant, since it is of utmost importance to identify patients as soon as possible to improve their outcome [6]. ROC curves were generated using the ICU control group. Cutoff values were determined using two methods, one in which equal weight is given to false positives and false negatives, also called the Youden index, and another one in which all patients with sepsis are identified, meaning that the sensitivity reaches 100%. To assess combinations of enzymes as predictors of septic shock, the two enzymes were entered as independent variables into a logistic regression model with disease status as dependent variable. Subsequently, the predicted probability of this model was used as an input for the ROC-curve upon which the AUC calculation was based.

Survival analysis

Survival analysis was performed using a Cox proportional hazard model with the survival time (censored at 90 days) as outcome and the four longitudinal enzyme activities as independent variables. In addition, a logistic regression model with survival up to day 90 as binary outcome was fitted.

All statistical analyses have been performed using R (Version 3.1.2, R Core Development Team (2008)). A p-value below 0.05 was considered significant, unless otherwise stated.

3. Results

3.1. Study populations

The characteristics of the septic shock patients are summarized in Table 1, including all the inflammatory, hemodynamic and metabolic parameters selected for this study. The septic shock patients were of older age, several of them suffered from multiple conditions, such as various malignancies, hypertension, myocardial infarction, COPD and others. None of the patients, however, used DPP4-inhibitors. The most common sources of sepsis were pneumonia (n = 17), abdominal sepsis (n = 10), mediastinitis (n = 4) and soft tissue/muscular infection (n = 3). In the other patients, the sites of infection were central line sepsis, wound infection, myocarditis and leptospirosis. In 2 patients multiple sites were identified (pneumonia/central line sepsis). The median (IQR) APACHE II score of 25 (20–29), a duration of vasopressor treatment of 3.0 (1.5–8.5) days and the necessity to restart this treatment in 6 out of the 40 patients illustrate the severity of septic shock encountered in these patients. Additionally, 13 patients had died at day 90. More information on the ICU control patients can be found in Table 1 and S1 Table.

Table 1. Study populations: Characteristics and the selected septic shock related parameters.

Longitudinally measured parameters were determined on days 1, 3, 5 and 7 after diagnosis. Data are expressed as median (IQR), unless stated otherwise.

A. Study populations characteristics Patients (n = 40) ICU controls (n = 22)
Age, years (mean ± SD) 63 ± 15 51 ± 12
Gender Female: n = 19 Female: n = 13
Male: n = 21 Male: n = 9
B. Septic shock parameters in patients [median (IQR)]
Longitudinally measured Day 1 Day 3 Day 5 Day 7
Mean arterial pressure, mmHg 75.00 (68.00–80.00) 76.00 (70.25–83.50) 76.00 (70.25–87.25) 82.50 (69.75–98.25)
Heart rate, beats/minute 106.00 (90.00–117.75) 89.50 (80.25–110.25) 96.50 (77.25–109.75) 89.00 (75.25–103.50)
PaO2/FiO2 ratio 30.44 (23.40–35.60) 30.38 (19.77–35. 94) 30.43 (26.89–38.65) 31.14 (25.46–36.81)
Thrombocytes, x 109/L 195.00 (141.25–321.50) 152.00 (90.50–320.25) 140.00 (65.00–306.00) 163.50 (97.00–336.00)
Total bilirubin, μmol/L 13.00 (9.00–26.00) 15.00 (8.50–66.75) 21.50 (12.00–28.25) 25.50 (8.50–33.75)
Creatinine, μmol/L 139.00 (73.75–279.00) 119.00 (60.75–234.25) 105.00 (52.00–146.00) 94.00 (52.00–146.00)
Leucocytes, x 109/L 14.20 (7.20–19.40) 14.40 (6.35–19.30) 11.30 (6.50–16.30) 11.85 (7.35–17. 53)
Lactate, mmol/L 1.80 (1.20–2.65) 1.40 (1.20–1.80) 1.35 (1.20–2.23) 1.60 (1.20–1.85)
Noradrenalin infusion rate, μg/kg/min 0.20 (0.11–0.50) 0.10 (0.00–0.30) 0.00 (0.00–0.10) 0.00 (0.00–0.08)
TNFα, pg/mL 33.42 (19.90–72.00) 23.63 (15.53–41.46) 22.90 (14.83–35.42) 20.93 (11.92–32.90)
IFNγ, pg/mL 9.50 (3.20–21.25) 8.00 (3.20–15.20) 5.32 (3.20–16.15) 5.36 (3.20–16.32)
IL-1β, pg/mL 3.20 (0.70–3.44) 3.20 (0.70–3.20) 3.20 (0.84–3.20) 3.20 (0.84–3.80)
IL-1RA, pg/mL 186.54 (53.10–694.25) 66.34 (36.93–198.91) 69.25 (41.65–132.36) 62.38 (38.54–120.76)
IL-6, pg/mL 216.42 (48.18–1489.25) 45.90 (12.44–137.43) 25.32 (11.50–76.98) 31.42 (8.92–61.63)
IL-8, pg/mL 135.60 (59.72–327.23) 54.77 (31.82–141.15) 45.26 (28.03–110.96) 50.54 (28.54–99.23)
IL-10, pg/mL 78.43 (26.35–262.50) 38.78 (8.75–96.05) 27.69 (9.88–54.15) 26.60 (7.71–44.05)
I-FABP, pg/mL 374.00 (192.00–912.90) 522.13 (187.00–1422.23) 455.16 (241.09–1324.25) 789.50 (361.25–1347.32)
Dialysis (yes/total) 5/40 5/40 9/40 11/40
Ventilation (yes/total) 36/40 32/40 32/40 29/40
Measured once
Length of hospital stay, days 41.00 (28.25–53.75)
Length of Intensive Care stay, days 15.50 (9.50–32.50)
APACHE II score 25.00 (20.00–29.00)
Duration of noradrenalin treatment, days 3.00 (1.50–8.50)
Restart of noradrenalin treatment Yes: 6
No: 34
Mortality at day 90 13/40
SOFA score on day 1 (min-max) 8 (4–15)
Open in a new tab

Abbreviations used: I-FABP: intestinal fatty-acid binding protein; IFNγ: interferon γ; IL: interleukin; IL-1RA: interleukin-1 receptor antagonist; IQR: interquartile range; PaO2/FiO2 ratio: ratio of arterial oxygen partial pressure to fractional inspired oxygen; TNFα: tumor necrosis factor α.

3.2. Enzyme activities in septic shock patients and ICU controls

The median (IQR) values of the enzymatic activities can be found in Fig 1 and S2 Table. FAP and DPP4 activity were both significantly lower in patients with septic shock compared to ICU controls, across all four days (p ≤ 0.001). Within the septic shock patients, FAP showed a significant decrease from day 1 to day 3 (p ≤ 0.001), whereas DPP4 differed between day 1 and days 3 and 5 (p = 0.003 and p = 0.028). PRCP activity was higher in the septic shock patient group on day 1 (p = 0.001) compared to the control patients. Within the septic shock patients, a decrease in PRCP activity was observed between day 1 and day 5 (p = 0.017) and day 7 (p = 0.007). On all four days of measurement (day 1, 3, 5 and 7), the septic shock patients had a significantly higher PREP activity (p ≤ 0.001) than the ICU control patients. Some septic shock patients had a very high PREP activity on day 1 but normalized to control levels afterwards. For DPP4, FAP and PREP, the activity levels did not normalize to control levels during the 7-day study period. We also measured the enzyme activities in 30 healthy controls (see S1 Appendix) and compared the values with the ICU controls. Between these two controls groups, no statistically significant differences were identified, and a multivariate analysis via principal components (see S1 and S2 Figs), gave no indication that the two groups were different.

Fig 1. The plasma activity of DPP4, FAP, PRCP and PREP in septic shock patients and ICU controls.

Fig 1

Open in a new tab

The plasma activity of DPP4 (A), FAP (B), PRCP (C) and PREP (D), expressed in U/L (median with interquartile range) on days 1, 3, 5 and 7 of the septic shock patients (n = 40) and in the ICU control group (n = 22). Differences between the control group and the different days in the patients were tested using linear mixed models, followed by a post-hoc analysis with Dunnett’s correction. Within the septic shock patients, pairwise differences between the days were modeled using linear mixed models, followed by a post-hoc Tukey’s correction. For PREP, the test was carried out on the log-transformed enzyme activity due to the non-normality of the residuals. * p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001. See also S2 Table. Abbreviations used: DPP4: dipeptidyl peptidase 4; FAP: fibroblast activation protein α; PRCP: prolylcarboxypeptidase; PREP: prolyl oligopeptidase; U/L: units per liter.

3.3. Receiver operating characteristic curves

Since large differences were found in the activities of all four enzymes, ROC curves were computed to evaluate the ability of each of the enzymes individually to distinguish septic shock patients from ICU controls. DPP4, FAP and PREP turned out to be very good in discriminating septic shock from non-septic shock ICU control patients with areas under the curve (AUC) of 0.86 (95% confidence interval [CI]: 0.77–0.95), 0.94 (95% CI: 0.89–0.99) and 0.88 (95% CI: 0.80–0.96), respectively. PRCP had a lower diagnostic value with an AUC of 0.71 (95% CI: 0.58–0.83). The four ROC curves are depicted in Fig 2. Cutoff values, either determined using the Youden index (giving equal weight to false positives and false negatives) or to obtain 100% sensitivity (to identify all patients with sepsis), are listed in Table 2 together with the sensitivity, specificity, negative and positive predictive value and the negative and positive likelihood ratios. Two maxima were obtained for PRCP, in this case the cutoff that would yield the highest sensitivity was selected, as that would identify more patients with septic shock. The complete dataset can be found in S3 Table (DPP4), S4 Table (FAP), S5 Table (PRCP) and S6 Table (PREP). In some cases, but not always, combining enzymes lead to better predictions. Therefore, we chose to evaluate whether the combination of FAP and DPP4 and FAP with PREP would make a better model. We chose these two combinations since FAP had the highest AUC value and DPP4 and PREP had very similar AUC values. Combining FAP and DPP4 showed an AUC value of 0.94 (95% CI: 0.89–0.99), which is not different from FAP alone. On the other hand, when FAP and PREP were combined as predictors using logistic regression, the AUC value increased to 1 (95% CI: 0.95–1), which would mean that this model correctly predicts at a wide range of cutoffs. However, this latter estimate may be overly optimistic, as the prediction model was never validated in an independent dataset, and the current dataset is too small to carry out cross validation. The ROC curves can be found in S3 Fig.

Fig 2. Receiver operating characteristic (ROC) curves for DPP4, FAP, PRCP and PREP.

Fig 2

Open in a new tab

ROC curves for DPP4 (A), FAP (B), PRCP (C), PREP (D) measured in EDTA-plasma. Cutoff values are given in Table 2. Also see S3, S4, S5 and S6 Tables. Abbreviations used: AUC: area under the curve; DPP4: dipeptidyl peptidase 4; FAP: fibroblast activation protein α; PRCP: prolylcarboxypeptidase; PREP: prolyl oligopeptidase.

Table 2. Cutoff values based upon the ROC curves of DPP4, FAP, PRCP and PREP.

For the septic shock patients day 1 was used. Cutoff values were determined using the Youden index or when all patients with septic shock would be identified (sensitivity = 100%). For DPP4 and FAP, enzymatic activities lower than the indicated cutoff value would suggest a diagnosis with septic shock. For PRCP and PREP this would be the case if the activity lies above the calculated cutoff value. The complete datasets can be found in S3 Table (DPP4), S4 Table (FAP), S5 Table (PRCP) and S6 Table (PREP).

DPP4 (Controls: n = 22; Cases: n = 40) FAP (Controls: n = 22; Cases: n = 40) PRCP (Controls: n = 22; Cases: n = 40) PREP (Controls: n = 22; Cases: n = 40))
Youden index (J = 0.63) All septic shock patients Youden index (J = 0.80) All septic shock patients Youden index (J = 0.38) All septic shock patients Youden index (J = 0.65) All septic shock patients
Cutoff value, U/L 11.41 22.06 1.31 2.73 1.03 0.17 0.69 0.29
Sensitivity 0.68 1.00 0.80 1.00 0.48 1.00 0.65 1.00
Specificity 0.95 0.18 1.00 0.45 0.91 0.00 1.00 0.41
Positive predictive value 0.96 0.69 1.00 0.77 0.90 0.65 1.00 0.75
Negative predictive value 0.62 1.00 0.73 1.00 0.49 NA 0.61 1.00
Positive likelihood ratio 14.85 1.22 ∞ 1.83 5.23 1.00 ∞ 1.69
Negative likelihood ratio 0.34 0.00 0.20 0.00 0.58 NA 0.35 0.00
Open in a new tab

Abbreviations used: DPP4: dipeptidyl peptidase 4, FAP: fibroblast activation protein α; NA: not applicable; PRCP: prolylcarboxypeptidase; PREP: prolyl oligopeptidase; U/L: units per liter.

3.4. Survival analysis

A survival analysis was performed to investigate if there is an association between the enzyme activities and the time till death. At day 90, 13 people had died (32.5%). An overall effect on survival time was found for PRCP and PREP (p = 0.043 and p = 0.048, respectively) in the Cox proportional hazard model. Despite this overall effect, none of the enzymes at the individual time points was significantly associated with mortality before day 90. A nominally significant association was observed between mortality before day 90 and the DPP4 enzymatic activity at day 1 (p = 0.04 and p = 0.03, respectively) in both the Cox proportional hazard model and the logistic regression model, with an increased DPP4 enzymatic activity being associated with an increase in survival.

3.5. Association analysis

Because we wanted to see whether the enzymatic activities were associated with other parameters frequently measured in sepsis and septic shock (which are presented in Table 1), we additionally performed an association analysis (see S1 Appendix). Unfortunately, only a few strong associations were identified (PREP with intestinal fatty-acid binding protein (I-FABP) and interleukin-1 receptor antagonist (IL-1RA)). See S4 Fig.

4. Discussion

To the best of our knowledge, this is the first report on the activity of the proline-specific peptidases DPP4, PRCP, FAP and PREP in a well-characterized population of septic shock patients over a time-interval of 7 days. Since they are involved in blood pressure regulation and inflammatory pathways, both are overtly disturbed in sepsis and septic shock, it is conceivable that these enzymes are dysregulated or involved in the pathogenesis of sepsis and septic shock. Indeed, both DPP4 and FAP show a significantly lower enzyme activity in plasma of patients with septic shock for all time points. This is in accordance with previously published data reporting a lower DPP4 activity in patients with severe sepsis [9]. Lower plasma DPP4 activity, however, is not specific for sepsis and has also been reported in other inflammatory disorders such as rheumatoid arthritis [40,41] and in certain cancers as reviewed in [42]. FAP enzyme activity is known to be increased in certain liver diseases and decreased in certain malignancies, but remains unchanged in rheumatoid arthritis or scleroderma [43]. Unfortunately, the exact sources or mechanisms by which these enzymes are released into the bloodstream are not fully understood [44,45]. It is plausible that these secretion or shedding mechanisms are disturbed.

We report that PRCP activity is increased in the septic shock patient group on day 1. As PRCP is an angiotensin II degrading enzyme and angiotensin II is recently approved as therapy for septic shock [27,28], PRCP has the potential to become a promising predictive biomarker for response to angiotensin II. Angiotensin II treatment probably will not have the desired effect in patients with high PRCP activity. A dedicated study, including associations between PRCP activity and angiotensin II levels, angiotensin (1–7) levels and ideally other markers of the renin-angiotensin system is needed to confirm this hypothesis.

At present, there is no good diagnostic biomarker for sepsis (for an overview of biomarkers studied in sepsis we refer to [46,47]). Other biomarkers that have been studied in the past, such as procalcitonin and C-reactive protein, lack specificity [47]. To see whether surgery itself would also influence the enzymatic activities, we chose patients that were admitted to the ICU after major intracranial surgery as a relevant control group. In our study, the enzymatic activity of the ICU controls did not differ from a healthy control group, reassuring that the observed differences between the septic shock patients and ICU control patients are indeed due to the disease state of the septic shock patients and not to the surgery or the underlying disease of the ICU control patients. Additionally, reference values for DPP4 have been published in the past, and further confirm that the observed DPP4 enzymatic activities are lower in the septic shock patients, while the ICU control group falls within the expected range [48].

DPP4, FAP and PREP can discriminate between the septic shock patients (day 1) and the ICU control group, as reflected by high AUC values. When the Youden index is applied, FAP reaches the highest sensitivity (80%) and specificity (100%). Since septic shock has a high mortality, another cutoff value was calculated in which no patients would be missed (sensitivity is 100%). In this case, each of the enzymes gives an insufficient specificity as more than half of the ICU control patients would be falsely diagnosed with septic shock and would receive unnecessary treatment. The ability of the enzymes to discriminate between an ICU control group and septic shock patients, does not necessarily mean that these enzymes are indeed good biomarkers for sepsis itself. Septic shock is a very extreme condition, which does not need a specific biomarker. A biomarker for sepsis is needed in the early stage, when one is not sure whether a patient does indeed have sepsis or not. Therefore, in the future, it would be very interesting to study larger groups of patients and to prospectively include all patients at high risk of developing sepsis and measure, for example, their DPP4- and FAP-activities on a daily basis. This would provide us with the information whether the observed reduction in enzymatic activity in the most extreme form of sepsis develops gradually or at once and whether DPP4 and/or FAP should be used as a target to watch in patients with a high risk of developing sepsis. Thereby, identifying these patients earlier and enabling the caregivers to initiate therapy earlier on, in order to improve the outcome of these patients. Most likely, a combination of these enzymes with other parameters used in sepsis could result in a diagnostic biomarker panel. For now, especially FAP looks very promising and also the combination of FAP and PREP should be further explored. A limitation of this study is that only samples from patients who stayed in the ICU during the entire week, were available. Follow-up studies should include all patients diagnosed with sepsis. This may even increase the biomarker value of the enzymes studied here. Moreover, a good biomarker should be cheap and easily measurable, for DPP4 this is certainly the case, since it only requires a commercially available, chromogenic substrate and results can be obtained within 1 hour after blood sampling. We recently published a new and faster method for the simultaneous measurement of FAP and PREP using a commercially available fluorogenic substrate, making also these measurements readily available [37].

Only a marginally significant association with mortality at day 90 was found for DPP4 activity on day 1. An overall association with PREP and PRCP was seen in the Cox proportional hazard model. However, given the large number of associations tested here, the reported p-values are not very promising and therefore we cannot conclude from this data that PRCP or PREP are good prognostic biomarkers for sepsis. Additionally, it should be taken into consideration that only patients, of whom samples on days 1, 3, 5 and 7 were available, were included in the study. A stronger association might be identified when the DPP4-activity on day 1 is measured in a larger and more diverse sepsis patient population.

In the association analysis, a large panel of parameters was studied. The selection of the parameters was based upon their use as biomarkers in sepsis, their involvement in inflammation and/or importance for hemodynamic patient characterization. Rather surprisingly, only a few strong associations were found. Most remarkable are the associations between PREP and I-FABP and PREP with IL-1RA. I-FABP is quickly released into the circulation after intestinal damage [49]. Since PREP is also highly expressed in the gastro-intestinal tract [50,51] and intestinal damage occurs in sepsis [52], it is possible that PREP could serve as a blood-based biomarker for intestinal damage. The associations between PREP and IL-1RA and PREP and I-FABP are new and warrant further research.

5. Conclusion

We show here that the activities of the proline-specific peptidases DPP4, PRCP, FAP and PREP are dysregulated in septic shock patients and can significantly discriminate between septic shock patients and the selected ICU control group (PREP, FAP and DPP4: AUC values of 0.88 (CI: 0.80–0.96), 0.94 (CI: 0.89–0.99) and 0.86 (CI: 0.77–0.95)).

The present study strongly supports a further exploration of these enzymes as potential diagnostic and/or prognostic biomarkers in sepsis and septic shock.

Supporting information

S1 Fig. Plasma activity of DPP4, FAP, PRCP and PREP in healthy and ICU controls.

The plasma activity of DPP4 (A), FAP (B), PRCP (C) and PREP (D), expressed in U/L (median with interquartile range), measured in a healthy control group (n = 29–30) and the ICU control group (n = 22). Differences between the two control groups were tested using Mann-Whitney U tests. Abbreviations used: DPP4: dipeptidyl peptidase 4; FAP: fibroblast activation protein α; NS: non-significant; PRCP: prolylcarboxypeptidase; PREP: prolyl oligopeptidase; U/L: units per liter.

(DOCX)

Click here for additional data file. (201.5KB, docx)
S2 Fig. Principal component analysis.

Plot of the first two principal components scores for the ICU control group and the healthy control group, based upon PREP, PRCP, FAP and DPP4. Abbreviations used: ICU; intensive care unit; PC: principal component.

(DOCX)

Click here for additional data file. (91.8KB, docx)
S3 Fig. Receiver Operating Characteristic (ROC) curves of the combination FAP and DPP4 and FAP with PREP.

(DOCX)

Click here for additional data file. (97.9KB, docx)
S4 Fig. Visual representation of the statistical analysis of the studied associations.

(A) Dot plot of the log-transformed p-value for the interaction (pInt) for the longitudinally measured parameters. The dashed line indicates a p-value of 0.01. Significant interactions were obtained for PRCP with total bilirubin and TNFα. P-values are based upon linear mixed models. (B) Dot plot of the log-transformed p-value for the main effect (pMain) for the longitudinally measured parameters. In case of a significant interaction between time and the parameter, no main effect p-value is shown in the dot plot. The dashed line indicates a p-value of 0.01. Linear mixed models were fitted to obtain the p-values. (C) R² for the longitudinally measured parameters. R2 is only depicted when pInt ˃ 0.01. P-values are based upon linear mixed models. ○ DPP4; + PRCP; △ FAP; ● PREP. Abbreviations used: DPP4: dipeptidyl peptidase 4; FAP: fibroblast activation protein α; I-FABP: intestinal fatty acid-binding protein; IFNγ: interferon γ; IL: interleukin; IL-1RA: interleukin-1 receptor antagonist; MAP: mean arterial pressure; NIR, noradrenalin infusion rate; PaO2/FiO2 ratio: ratio of arterial oxygen partial pressure to fractional inspired oxygen; PRCP: prolylcarboxypeptidase; PREP: prolyl oligopeptidase; TNFα: tumor necrosis factor α.

(DOCX)

Click here for additional data file. (207KB, docx)
S1 Table. Characteristics of the non-septic shock ICU control group.

A total of 22 patients were included for this study.

(DOCX)

Click here for additional data file. (15.6KB, docx)
S2 Table. The enzyme activity of DPP4, FAP, PRCP and PREP.

Enzymatic activities were determined in EDTA-plasma of septic shock patients on day 1, 3, 5 and 7 and non-septic shock ICU control patients. Data are represented as median (interquartile range). Abbreviations used: DPP4: dipeptidyl peptidase 4; FAP: fibroblast activation protein α; PRCP: prolylcarboxypeptidase; PREP: prolyl oligopeptidase; U/L: units per liter.

(DOCX)

Click here for additional data file. (18.3KB, docx)
S3 Table. Cutoff values receiver operating characteristic curve of dipeptidyl peptidase 4 (DPP4).

The 1-specificity, sensitivity, negative predictive value (NPV), positive predictive value (PPV), positive likelihood ratio (LR+), negative likelihood ratio (LR-) and Youden index for every cutoff value expressed in U/L for DPP4. For the septic shock patients day 1 was used. Blue indicates the maximum for the Youden index. Red indicates where the sensitivity reaches 1. In the case of DPP4, a value lower than the indicated cutoff would suggest a diagnosis with septic shock. ICU controls: n = 22; Septic shock patients: n = 40.

(DOCX)

Click here for additional data file. (24.6KB, docx)
S4 Table. Cutoff values receiver operating characteristic curve of fibroblast activation protein α (FAP).

The 1-specificity, sensitivity, negative predictive value (NPV), positive predictive value (PPV), positive likelihood ratio (LR+), negative likelihood ratio (LR-) and Youden index for every cutoff value expressed in U/L for FAP. For the septic shock patients day 1 was used. Blue indicates the maximum for the Youden index. Red indicates where the sensitivity reaches 1. In the case of FAP, a value lower than the indicated cutoff would suggest a diagnosis with septic shock. ICU controls: n = 22; Septic shock patients: n = 40.

(DOCX)

Click here for additional data file. (24.4KB, docx)
S5 Table. Cutoff values receiver operating characteristic curve of prolylcarboxypeptidase (PRCP).

The 1-specificity, sensitivity, negative predictive value (NPV), positive predictive value (PPV), positive likelihood ratio (LR+), negative likelihood ratio (LR-) and Youden index for every cutoff value expressed in U/L for PRCP. For the septic shock patients day 1 was used. Blue indicates the maximum for the Youden index. Two maxima were obtained for PRCP, in this case the cutoff that would yield the highest sensitivity was selected, as that would identify more patients with septic shock. Red indicates where the sensitivity reaches 1. In the case of PRCP, a value higher than the indicated cutoff would suggest a diagnosis with septic shock. ICU controls: n = 22; Septic shock patients: n = 40.

(DOCX)

Click here for additional data file. (23.7KB, docx)
S6 Table. Cutoff values receiver operating characteristic curve of prolyl oligopeptidase (PREP).

The 1-specificity, sensitivity, negative predictive value (NPV), positive predictive value (PPV), positive likelihood ratio (LR+), negative likelihood ratio (LR-) and Youden index for every cutoff value expressed in U/L for PREP. For the septic shock patients day 1 was used. Blue indicates the maximum for the Youden index. Red indicates where the sensitivity reaches 1. In the case of PREP, a value higher than the indicated cutoff would suggest a diagnosis with septic shock. ICU controls: n = 22; Septic shock patients: n = 40.

(DOCX)

Click here for additional data file. (23.1KB, docx)
S1 Appendix. Dysregulated activities of proline-specific enzymes in septic shock patients(sepsis-2).

(DOCX)

Click here for additional data file. (30.2KB, docx)

Abbreviations

APACHE II

Acute Physiology and Chronic Health Evaluation II

AUC

area under the curve

CI

confidence interval

DPP4

dipeptidyl peptidase 4

DTT

dithiothreitol

EDTA

ethylenediaminetetraacetic acid

FAP

fibroblast activation protein α

FGF21

fibroblast growth factor 21

Gly-Pro-pNA

glycyl-prolyl-paranitroanilide

ICU

intensive care unit

I-FABP

intestinal fatty-acid binding protein

IFNγ

interferon γ

IL

interleukin

IL-1RA

interleukin-1 receptor antagonist

IQR

interquartile range

MAP

mean arterial pressure

NA

not applicable

PaO2/FiO2 ratio

ratio of arterial oxygen partial pressure to fractional inspired oxygen

PRCP

prolylcarboxypeptidase

PREP

prolyl oligopeptidase

ROC

receiver operating characteristic

SIRS

systemic inflammatory response syndrome

TNFα

tumor necrosis factor α

U/L

units per liter

Z-Gly-Pro-AMC

N-benzyloxycarbonyl-Gly-Pro-7-amido-4-methylcoumarine

Z-Pro-Phe

N-benzyloxycarbonyl-Pro-Phe

Data Availability

The SPSS file containing all raw data is available from Figshare (DOI 10.6084/m9.figshare.10252877.). Processed data are within the manuscript and its Supporting Information Files.

Funding Statement

EDH and BSJ are research assistants of the Fund for Scientific Research Flanders (1S22417N and SB151525, respectively; www.fwo.be). This work was supported by a GOA BOF 2015 grant of the University of Antwerp to SKS, AML and IDM (No. 30729; www.uantwerp.be). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

  • 1.Vincent J-L, Sakr Y, Sprung CL, Ranieri VM, Reinhart K, Gerlach H, et al. Sepsis in European intensive care units: results of the SOAP study. Crit Care Med. 2006;34(2):344–53. 10.1097/01.ccm.0000194725.48928.3a [DOI] [PubMed] [Google Scholar]
  • 2.Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, et al. The third international consensus definitions for sepsis and septic shock (Sepsis-3). JAMA. 2016;315(8):801 10.1001/jama.2016.0287 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Merx MW, Weber C. Sepsis and the heart. Circulation. 2007;116(7):793–802. 10.1161/CIRCULATIONAHA.106.678359 [DOI] [PubMed] [Google Scholar]
  • 4.Neviere R. Sepsis syndromes in adults: Epidemiology, definitions, clinical presentation, diagnosis, and prognosis—UpToDate [Internet]. 2019 [cited 2019 Jul 31]. https://www.uptodate.com/contents/sepsis-syndromes-in-adults-epidemiology-definitions-clinical-presentation-diagnosis-and-prognosis
  • 5.Singer M, De Santis V, Vitale D, Jeffcoate W. Multiorgan failure is an adaptive, endocrine-mediated, metabolic response to overwhelming systemic inflammation. Lancet. 2004;364(9433):545–8. 10.1016/S0140-6736(04)16815-3 [DOI] [PubMed] [Google Scholar]
  • 6.Dellinger R, Levy M, Rhodes A. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med. 2013;41(2):580–637. 10.1097/CCM.0b013e31827e83af [DOI] [PubMed] [Google Scholar]
  • 7.Rawlings ND, Barrett AJ. MEROPS: the Peptidase Database [Internet]. 2019 [cited 2019 Sep 5]. http://merops.sanger.ac.uk/index.shtml
  • 8.Rawlings ND, Barrett AJ, Bateman A. MEROPS: the database of proteolytic enzymes, their substrates and inhibitors. Nucleic Acids Res. 2012;40(Database issue):D343–50. 10.1093/nar/gkr987 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Bergmann A, Bohuon C. Decrease of serum dipeptidylpeptidase activity in severe sepsis patients: relationship to procalcitonin. Clin Chim Acta. 2002;321(1–2):123–6. 10.1016/s0009-8981(02)00042-6 [DOI] [PubMed] [Google Scholar]
  • 10.Felies M, von Horsten S, Pabst R, Nave H. Neuropeptide Y stabilizes body temperature and prevents hypotension in endotoxaemic rats. J Physiol. 2004;561(Pt 1):245–52. 10.1113/jphysiol.2004.073635 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Steven S, Hausding M, Kröller-Schön S, Mader M, Mikhed Y, Stamm P, et al. Gliptin and GLP-1 analog treatment improves survival and vascular inflammation/dysfunction in animals with lipopolysaccharide-induced endotoxemia. Basic Res Cardiol. 2015;110(2):6 10.1007/s00395-015-0465-x [DOI] [PubMed] [Google Scholar]
  • 12.Brakenridge SC, Moore FA, Mercier NR, Cox M, Wu Q, Moldawer LL, et al. Persistently elevated glucagon-like peptide-1 levels among critically ill surgical patients after sepsis and development of chronic critical illness and dismal long-term outcomes. J Am Coll Surg. 2019;229(1):58–67.e1. 10.1016/j.jamcollsurg.2019.04.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Lebherz C, Schlieper G, Möllmann J, Kahles F, Schwarz M, Brünsing J, et al. GLP-1 levels predict mortality in patients with critical illness as well as end-stage renal disease. Am J Med. 2017;130(7):833–841.e3. 10.1016/j.amjmed.2017.03.010 [DOI] [PubMed] [Google Scholar]
  • 14.Wrenger S, Kähne T, Bohuon C, Weglöhner W, Ansorge S, Reinhold D. Amino-terminal truncation of procalcitonin, a marker for systemic bacterial infections, by dipeptidyl peptidase IV (DP IV). FEBS Lett. 2000;466(1):155–9. 10.1016/s0014-5793(99)01779-2 [DOI] [PubMed] [Google Scholar]
  • 15.Becker KL, Snider R, Nylen ES. Procalcitonin in sepsis and systemic inflammation: A harmful biomarker and a therapeutic target. Br J Pharmacol. 2010;159(2):253–64. 10.1111/j.1476-5381.2009.00433.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Mortier A, Gouwy M, Van Damme J, Proost P, Struyf S. CD26/dipeptidylpeptidase IV-chemokine interactions: double-edged regulation of inflammation and tumor biology. J Leukoc Biol. 2016;99(6):955–69. 10.1189/jlb.3MR0915-401R [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Shih C-J, Wu Y-L, Chao P-W, Kuo S-C, Yang C-Y, Li S-Y, et al. Association between use of oral anti-diabetic drugs and the risk of sepsis: a nested case-control study. Sci Rep. 2015;5:15260 10.1038/srep15260 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Lee KN, Jackson KW, Christiansen VJ, Lee CS, Chun JG, McKee PA. Antiplasmin-cleaving enzyme is a soluble form of fibroblast activation protein. Blood. 2006;107(4):1397–404. 10.1182/blood-2005-08-3452 [DOI] [PubMed] [Google Scholar]
  • 19.Zhu H, Tang Y, Ivanciu L, Centola M, Lupu C, Taylor FB, et al. Temporal dynamics of gene expression in the lung in a baboon model of E. coli sepsis. BMC Genomics. 2007;8(1):58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Lee KN, Jackson KW, Christiansen VJ, Dolence EK, McKee PA. Enhancement of fibrinolysis by inhibiting enzymatic cleavage of precursor α2-antiplasmin. J Thromb Haemost. 2011;9(5):987–96. 10.1111/j.1538-7836.2011.04195.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Kager LM, Weehuizen TA, Wiersinga WJ, Roelofs JJTH, Meijers JCM, Dondorp AM, et al. Endogenous α2-antiplasmin is protective during severe gram-negative sepsis (Melioidosis). Am J Respir Crit Care Med. 2013;188(8):967–75. 10.1164/rccm.201307-1344OC [DOI] [PubMed] [Google Scholar]
  • 22.Gariani K, Drifte G, Dunn-Siegrist I, Pugin J, Jornayvaz FR. Increased FGF21 plasma levels in humans with sepsis and SIRS. Endocr Connect. 2013;2(3):146–53. 10.1530/EC-13-0040 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Sorrells K, Erdös EG. Prolylcarboxypeptidase in biological fluids. Adv Exp Med Biol. 1971;23:393–7. 10.1007/978-1-4615-9014-9_36 [DOI] [PubMed] [Google Scholar]
  • 24.Shariat-Madar Z, Schmaier AH. The plasma kallikrein/kinin and renin angiotensin systems in blood pressure regulation in sepsis. J Endotoxin Res. 2004;10(1):3–13. 10.1179/096805104225003807 [DOI] [PubMed] [Google Scholar]
  • 25.Shariat-Madar Z, Mahdi F, Schmaier AH. Identification and characterization of prolylcarboxypeptidase as an endothelial cell prekallikrein activator. J Biol Chem. 2002;277(20):17962–9. 10.1074/jbc.M106101200 [DOI] [PubMed] [Google Scholar]
  • 26.Yang HYT, Erdös EG, Chiang TS. New Enzymatic Route for the Inactivation of Angiotensin. Nature. 1968;218(5148):1224–6. 10.1038/2181224a0 [DOI] [PubMed] [Google Scholar]
  • 27.Senatore F, Jagadeesh G, Rose M, Pillai VC, Hariharan S, Liu Q, et al. FDA Approval of Angiotensin II for the Treatment of Hypotension in Adults with Distributive Shock. Am J Cardiovasc Drugs. 2019;19(1):11–20. 10.1007/s40256-018-0297-9 [DOI] [PubMed] [Google Scholar]
  • 28.Khanna A, English SW, Wang XS, Ham K, Tumlin J, Szerlip H, et al. Angiotensin II for the Treatment of Vasodilatory Shock. N Engl J Med. 2017;377(5):419–30. 10.1056/NEJMoa1704154 [DOI] [PubMed] [Google Scholar]
  • 29.Koida M, Walter R. Post-proline cleaving enzyme. Purification of this endopeptidase by affinity chromatography. J Biol Chem. 1976;251(23):7593–9. [PubMed] [Google Scholar]
  • 30.Goossens F, De Meester I, Vanhoof G, Scharpé S. Distribution of prolyl oligopeptidase in human peripheral tissues and body fluids. Eur J Clin Chem Clin Biochem. 1996;34(1):17–22. 10.1515/cclm.1996.34.1.17 [DOI] [PubMed] [Google Scholar]
  • 31.Brandt I, Scharpé S, Lambeir A-M. Suggested functions for prolyl oligopeptidase: A puzzling paradox. Clin Chim Acta. 2007;377(1):50–61. [DOI] [PubMed] [Google Scholar]
  • 32.Van Elzen R, Lambeir A-M. Structure and function relationship in prolyl oligopeptidase. CNS Neurol Disord Drug Targets. 2011;10(3):297–305. 10.2174/187152711794653814 [DOI] [PubMed] [Google Scholar]
  • 33.American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference: definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med. 1992;20(6):864–74. [PubMed] [Google Scholar]
  • 34.Levy MM, Dellinger RP, Townsend SR, Linde-Zwirble WT, Marshall JC, Bion J, et al. The Surviving Sepsis Campaign: results of an international guideline-based performance improvement program targeting severe sepsis. Crit Care Med. 2010;38(2):367–74. 10.1097/CCM.0b013e3181cb0cdc [DOI] [PubMed] [Google Scholar]
  • 35.Matheeussen V, Lambeir A-M, Jungraithmayr W, Gomez N, Mc Entee K, Van der Veken P, et al. Method comparison of dipeptidyl peptidase IV activity assays and their application in biological samples containing reversible inhibitors. Clin Chim Acta. 2012;413(3–4):456–62. 10.1016/j.cca.2011.10.031 [DOI] [PubMed] [Google Scholar]
  • 36.Baerts L, Brouns R, Kehoe K, Verkerk R, Engelborghs S, De Deyn PP, et al. Acute ischemic stroke severity, progression, and outcome relate to changes in dipeptidyl peptidase IV and fibroblast activation protein activity. Transl Stroke Res. 2017;8(2):157–64. 10.1007/s12975-016-0493-3 [DOI] [PubMed] [Google Scholar]
  • 37.Bracke A, Van Elzen R, Van Der Veken P, Augustyns K, De Meester I, Lambeir A-M. The development and validation of a combined kinetic fluorometric activity assay for fibroblast activation protein alpha and prolyl oligopeptidase in plasma. Clin Chim Acta. 2019;495:154–60. 10.1016/j.cca.2019.04.063 [DOI] [PubMed] [Google Scholar]
  • 38.Kehoe K, Verkerk R, Sim Y, Waumans Y, Van der Veken P, Lambeir A-M, et al. Validation of a specific prolylcarboxypeptidase activity assay and its suitability for plasma and serum measurements. Anal Biochem. 2013;443(2):232–9. 10.1016/j.ab.2013.09.002 [DOI] [PubMed] [Google Scholar]
  • 39.Goossens F, De Meester I, Vanhoof G, Scharpé S. A sensitive method for the assay of serum prolyl endopeptidase. Eur J Clin Chem Clin Biochem. 1992;30(4):235–8. [PubMed] [Google Scholar]
  • 40.Ulusoy H, Kamanli A, Ilhan N, Kuru O, Arslan S, Alkan G, et al. Serum levels of soluble CD26 and CD30 and their clinical significance in patients with rheumatoid arthritis. Rheumatol Int. 2012;32(12):3857–62. 10.1007/s00296-011-2302-3 [DOI] [PubMed] [Google Scholar]
  • 41.Buljevic S, Detel D, Pucar LB, Mihelic R, Madarevic T, Sestan B, et al. Levels of dipeptidyl peptidase IV/CD26 substrates neuropeptide Y and vasoactive intestinal peptide in rheumatoid arthritis patients. Rheumatol Int. 2013;33(11):2867–74. 10.1007/s00296-013-2823-z [DOI] [PubMed] [Google Scholar]
  • 42.Enz N, Vliegen G, De Meester I, Jungraithmayr W. CD26/DPP4—a potential biomarker and target for cancer therapy. Pharmacol Ther. 2019;198:135–59. 10.1016/j.pharmthera.2019.02.015 [DOI] [PubMed] [Google Scholar]
  • 43.Lay AJ, Zhang HE, McCaughan GW, Gorrell MD. Fibroblast activation protein in liver fibrosis. Front Biosci (Landmark Ed). 2019;24:1–17. [DOI] [PubMed] [Google Scholar]
  • 44.Nargis T, Chakrabarti P. Significance of circulatory DPP4 activity in metabolic diseases. IUBMB Life. 2018;70(2):112–9. 10.1002/iub.1709 [DOI] [PubMed] [Google Scholar]
  • 45.Brennen WN, Isaacs JT, Denmeade SR. Rationale behind targeting fibroblast activation protein-expressing carcinoma-associated fibroblasts as a novel chemotherapeutic strategy. Mol Cancer Ther. 2012;11(2):257–66. 10.1158/1535-7163.MCT-11-0340 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Faix JD. Biomarkers of sepsis. Crit Rev Clin Lab Sci. 2013;50(1):23–36. 10.3109/10408363.2013.764490 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Prucha M, Bellingan G, Zazula R. Sepsis biomarkers. Clin Chim Acta. 2015;440:97–103. 10.1016/j.cca.2014.11.012 [DOI] [PubMed] [Google Scholar]
  • 48.Durinx C, Neels H, Van der Auwera J-C, Naelaerts K, Scharpé S, De Meester I. Reference values for plasma dipeptidyl peptidase IV activity and their association with other laboratory parameters. Clin Chem Lab Med. 2001;39(2). [DOI] [PubMed] [Google Scholar]
  • 49.Grootjans J, Lenaerts K, Buurman WA, Dejong CHC, Derikx JPM. Life and death at the mucosal-luminal interface: New perspectives on human intestinal ischemia-reperfusion. World J Gastroenterol. 2016;22(9):2760–70. 10.3748/wjg.v22.i9.2760 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Myöhänen TT, Pyykkö E, Männistö PT, Carpen O. Distribution of prolyl oligopeptidase in human peripheral tissues and in ovarian and colorectal tumors. J Histochem Cytochem. 2012;60(9):706–15. 10.1369/0022155412453051 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.The Human Protein Atlas. PREP [Internet]. [cited 2019 Jul 31]. http://www.proteinatlas.org/ENSG00000085377-PREP/tissue
  • 52.Neviere R. Pathophysiology of sepsis—UpToDate [Internet]. 2018/10/10. [cited 2019 Jul 26]. https://www.uptodate.com/contents/pathophysiology-of-sepsis

Decision Letter 0

Paul Proost

12 Dec 2019

PONE-D-19-30816

Dysregulated activities of proline-specific enzymes in septic shock patients (sepsis-2)

PLOS ONE

Dear Full professor De Meester,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Three external reviewers each gave specific comments on your manuscript (see details below). Their main questions are related to additional information on the clinical parameters/history of the patient and control samples and further analysis of potential correlations. Provide an answer to their comments in a point-by-point reply. Also clearly indicate in the manuscript that the patient population is limited if no additional samples are available. The legends of the figures need to be included in the text according to the Journal's recommendations. These will be combined with the correct figures once the manuscript would be accepted for publication.

We would appreciate receiving your revised manuscript by Jan 26 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

  • A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

  • A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

  • An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Paul Proost, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

Reviewer #3: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This paper is very interestiong and well written. But I have several concerns.

Major points.

1)What is the cause of sepsis? Pneumonia? peritonitis? Trauma? Tumor? You should show the cause of sepsis.

2)Please show the SOFA score (Day 1,3,5,7) in the sepsis group.

3)Were there past medical history, especially diabetis? Was there a patient taking a DPP4 inhibitor? If so, did you think that DPP4 inhibitor have an effect on this outcome?

4) Were there case where steroids were used for refractory shock? If so, did you think that steroid administration have an effect on this outcome?

Reviewer #2: PLOS ONE – VLIEGEN ET AL – DECEMBER 2019

SUMMARY

These authors investigated the activity of several proline-specific enzymes (PSE) (dipeptidyl peptidase 4 (DPP4), prolylcarboxypeptidase (PRCP), fibroblast activation protein α (FAP) and prolyl oligopeptidase (PREP)) in 40 septic shock patients, 22 ICU controls (neurosurgical patients) and 30 healthy controls. PRCP and PREP were higher while FAP and DPP4 activity were lower in septic shock than in controls (as per Figure 1). There were AUCs of 0.86 – 0.94 with septic shock. The PSE activity levels of the healthy and ICU controls dd not differ. These PSE cleave proteins near proline residues. The investigators used their own lab-developed activity assays. The study is novel and provides further insights into the roles of PSE in septic shock. It is a small study so only really opens the door to the role of these enzymes as septic shock biomarkers.

MAJOR

1. Results: “Combination of FAP and DPP4 did not increase the AUC value compared to FAP alone.” This is a random statement because combinations were not mentioned in methods and no other combinations are mentioned. Please delete or do a thorough combination analysis. Combinations of biomarkers are often better, and I would recommend taking this on as an additional analysis. Indeed, the authors make this statement themselves in the discussion: “Most likely, a combination of these enzymes with other parameters used in sepsis could result in a diagnostic biomarker panel.”

2. It would be useful to understand further how these enzyme activities are associated with altered clinical outcomes in addition to mortality. I would expect and like to see whether enzyme activity was associated with organ dysfunction as assessed by days alive and free of vasopressors, ventilation and renal replacement therapy (RRT). The enzyme activity is associated with mortality, but we don’t know if that also alters cardiovascular, respiratory or renal function as reflected by the days alive and free of vasopressors, ventilation and renal replacement therapy (RRT).

3. The sample size is relatively small and there no validation cohort. This is a critical limitation of believing the credibility of the results herein.

4. The statistical analysis for the diagnostic aim should examine in addition only the day 1 values, since there is survival/non-survivor risk of bias due to early deaths after day 1. There could be non-random drop out due to death that could bias the results herein. The authors state in the discussion:” Additionally, it should be taken into consideration that only patients, of whom samples on days 1, 3, 5 and 7 were available, were included in the study.”

5. PRCP modifies/ presumably lowers angiotensin II levels (as per Yang HYT, Erdös EG, Chiang TS. New Enzymatic Route for the Inactivation of. Angiotensin. Nature. 1968;218(5148):1224–6) and angiotensin II is a recently approved therapy for septic shock(1, 2) so that has greater relevance for possibly discovering a predictive biomarker for response to angiotensin II. PRCP levels were increased in septic shock and higher levels were associated with a poor prognosis in your study. Furthermore, there are good studies of abnormalities of the renal-angiotensin-aldosterone (RAAS) system in sepsis(3). Accordingly, it would be useful to examine the PRCP association with markers of RAAS actively (plasma rein activity etc.). Of note, PRCP was not a great diagnostic (Low AUC) but was a good biomarker of prognosis – this suggests that the RAAS is a better marker of recovery or not from septic shock. Also, pre-sepsis use of ACE inhibitors could modify the PRCP activity – do you have that information available?

MINOR

1. Results. Study populations paragraph: would be useful to state the mortality rates in the Table 1 and that paragraph.

2. 30 healthy controls aren’t mentioned until the results section. These should be introduced and described in methods (re how and where they were recruited). They should also be described re age, gender etc. in Table 1.

3. The figure legends appear in the text but should be moved to the end.

4. Results, page 15. “For the septic shock patients day 1 was used.” I presume this refers to the ROC for diagnostic biomarker. Please clarify and introduce the rationale for this in methods.

5. Discussion:” Since septic shock has a high mortality, another cutoff value was calculated in which no patients would be missed. In this case, each of the enzymes gives an insufficient specificity as more than half of the ICU control patients would be falsely diagnosed with septic shock.” This is not clear and requires better explanation.

6. The Kaplan-Meier 90-day survival curves for the best PSE activity cutoffs should be shown as a figure in the main manuscript. The reader would like to see visually how PSE levels are associated with prognosis in septic shock.

REFERENCES

1. Khanna A, English SW, Wang XS, Ham K, Tumlin J, Szerlip H, et al. Angiotensin II for the Treatment of Vasodilatory Shock. N Engl J Med. 2017;377(5):419-30.

2. Senatore F, Jagadeesh G, Rose M, Pillai VC, Hariharan S, Liu Q, et al. FDA Approval of Angiotensin II for the Treatment of Hypotension in Adults with Distributive Shock. Am J Cardiovasc Drugs. 2019;19(1):11-20.

3. Emdin M, Fatini C, Mirizzi G, Poletti R, Borrelli C, Prontera C, et al. Biomarkers of activation of renin-angiotensin-aldosterone system in heart failure: how useful, how feasible? Clin Chim Acta. 2015;443:85-93.

Reviewer #3: The manuscript highlights markers that have the potential to become biomarkers for sepsis, for which currently no concrete association exists (as mentioned by authors as well). However, the associations described for most of the proline-specific enzymes were not very strong. As the authors mentioned further research is warranted. This manuscript will greatly benefit from more detailed analysis of age, gender, ethnicity to determine where the enzymes have any stronger correlation in any specific group. It is understandable data from human patients are not easily available however, to correlate the enzymes with progression in sepsis, this study requires control known inflammatory/anti-inflammatory patterns demonstrated side by side. This analysis will reveal the correlation of the inflammatory status, disease progression of sepsis 2 patient along with the enzymes can be potential biomarkers.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2020 Apr 21;15(4):e0231555. doi: 10.1371/journal.pone.0231555.r002

Author response to Decision Letter 0

28 Feb 2020

General:

1.) Also clearly indicate in the manuscript that the patient population is limited if no additional samples are available.

Next to what was already present in the text (lines 428-430) we added the following sentence to the Discussion: “A limitation of this study is that only samples from patients who stayed in the ICU during the entire week, were available.” (lines 414-415)

2.) The legends of the figures need to be included in the text according to the Journal's recommendations. These will be combined with the correct figures once the manuscript would be accepted for publication.

We provide the figure legends in the manuscript where the figure should be included, as is stated in the submission guidelines.

3.) Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming.

We have split the supporting information into S1 Appendix, S1 Fig, S2 Fig, S3 Fig, S4 Fig, S1 Table, S2 Table, S3 Table, S4 Table, S5 Table and S6 Table. We apologize for the mistake and hope that we now meet PLOS ONE’s style requirements.

Reviewer #1:

We would like to thank the reviewer for the careful revision of the manuscript and for the useful suggestions. We will address all remarks in detail below.

Major points:

1.) What is the cause of sepsis? Pneumonia? peritonitis? Trauma? Tumor? You should show the cause of sepsis.

We have added this information into 3.1 Study populations (lines 224-228).

“The most common sources of sepsis were pneumonia (n = 17), abdominal sepsis (n = 10), mediastinitis (n = 4) and soft tissue/muscular infection (n = 3). In the other patients, the sites of infection were central line sepsis, wound infection, myocarditis and leptospirosis. In 2 patients multiple sites were identified (pneumonia/central line sepsis).”

2.) Please show the SOFA score (Day 1,3,5,7) in the sepsis group.

The SOFA score has been assessed on day 1. This information has been included in Table 1 and was analysed statistically (see S1 appendix lines 39-41), to see whether there were associations with the enzymatic activities on day 1. However, none of the enzymatic activities was significantly associated with the SOFA score on day 1 (data not shown).

3.) Were there past medical history, especially diabetis? Was there a patient taking a DPP4 inhibitor? If so, did you think that DPP4 inhibitor have an effect on this outcome?

The patients studied herein, were of older age, several of them suffered from multiple conditions, such as various malignancies, hypertension, myocardial infarction, COPD and others. A sentence describing these patients has been added to 3.1 Study populations (lines 221-223).

“The septic shock patients were of older age, several of them suffered from multiple conditions, such as various malignancies, hypertension, myocardial infarction, COPD and others.”

Of the 40 patients, 6 patients had diabetes. As mentioned in the Introduction “A nested case-control study in type 2 diabetes patients admitted for sepsis did not find a significant association between the use of a DPP4-inhibitor and the development of sepsis [1].” None of the patients used a DPP4-inhibitor. This information has also been added to 3.1 Study populations (line 223-224).

4.) Were there case where steroids were used for refractory shock? If so, did you think that steroid administration have an effect on this outcome?

Multiple patients indeed received corticosteroids during their course. Our study has not been designed to evaluate the effect of corticosteroid use on the enzymatic activities, but limited data on this subject can be found in the literature for DPP4 and we do not expect an effect of (acute) corticosteroid treatment on serum DPP4 enzymatic activity [2,3]. On the other enzymes, we could not find literature data and it would be an interesting idea for the future to look into the effect of corticosteroid treatment on the enzymatic activity.

Reviewer #2:

First of all, we would like to thank the reviewer for the useful suggestions. We hope that our answers and adjustments are within the expectations.

SUMMARY

These authors investigated the activity of several proline-specific enzymes (PSE) (dipeptidyl peptidase 4 (DPP4), prolylcarboxypeptidase (PRCP), fibroblast activation protein α (FAP) and prolyl oligopeptidase (PREP)) in 40 septic shock patients, 22 ICU controls (neurosurgical patients) and 30 healthy controls. PRCP and PREP were higher while FAP and DPP4 activity were lower in septic shock than in controls (as per Figure 1). There were AUCs of 0.86 – 0.94 with septic shock. The PSE activity levels of the healthy and ICU controls dd not differ. These PSE cleave proteins near proline residues. The investigators used their own lab-developed activity assays. The study is novel and provides further insights into the roles of PSE in septic shock. It is a small study so only really opens the door to the role of these enzymes as septic shock biomarkers.

MAJOR

1.) Results: “Combination of FAP and DPP4 did not increase the AUC value compared to FAP alone.” This is a random statement because combinations were not mentioned in methods and no other combinations are mentioned. Please delete or do a thorough combination analysis. Combinations of biomarkers are often better, and I would recommend taking this on as an additional analysis. Indeed, the authors make this statement themselves in the discussion: “Most likely, a combination of these enzymes with other parameters used in sepsis could result in a diagnostic biomarker panel.”

We fully agree with the reviewer that we should clarify how the combination analysis was done. Therefore, we have added the following sentence to 2.4. Statistical analysis (lines 207-210):

“To assess combinations of enzymes as predictors of septic shock, the two enzymes were entered as independent variables into a logistic regression model with disease status as dependent variable. Subsequently, the predicted probability of this model was used as an input for the ROC-curve upon which the AUC calculation was based.”

The sentence “Combination of FAP and DPP4 did not increase the AUC value compared to FAP alone.” in the results section has been replaced with (lines 285-295):

“In some cases, but not always, combining enzymes lead to better predictions. Therefore, we chose to evaluate whether the combination of FAP and DPP4 and FAP with PREP would make a better model. We chose these two combinations since FAP had the highest AUC value and DPP4 and PREP had very similar AUC values. Combining FAP and DPP4 showed an AUC value of 0.94 (95% CI : 0.89 – 0.99) , which is not different from FAP alone. On the other hand, when FAP and PREP were combined as predictors using logistic regression, the AUC value increased to 1 (95% CI : 0.95-1), which would mean that this model correctly predicts at a wide range of cutoffs. However, this latter estimate may be overly optimistic, as the prediction model was never validated in an independent dataset, and the current dataset is too small to carry out cross validation. The ROC curves can be found in S3 Fig.”

In the Discussion the text has been adapted to (lines 411-417):

“Most likely, a combination of these enzymes with other parameters used in sepsis could result in a diagnostic biomarker panel. For now, especially FAP looks very promising and also the combination of FAP and PREP should be further explored. A limitation of this study is that only samples from patients who stayed in the ICU during the entire week, were available. Follow-up studies should include all patients diagnosed with sepsis. This may even increase the biomarker value of the enzymes studied here.”

2.) It would be useful to understand further how these enzyme activities are associated with altered clinical outcomes in addition to mortality. I would expect and like to see whether enzyme activity was associated with organ dysfunction as assessed by days alive and free of vasopressors, ventilation and renal replacement therapy (RRT). The enzyme activity is associated with mortality, but we don’t know if that also alters cardiovascular, respiratory or renal function as reflected by the days alive and free of vasopressors, ventilation and renal replacement therapy (RRT).

We assessed the mortality using a Cox proportional hazard model with the survival time (censored at 90 days) as outcome and the four longitudinal enzyme activities as independent variables. In addition, a logistic regression model with survival up to day 90 as binary outcome was fitted. The first model takes the days alive into account, while the second model only looks at whether the patient is still alive at day 90, yes or no. We describe these two models in the Materials and Methods, 2.4 Statistical analysis and the results thereof in lines 340-348 (only for DPP4 a significant association was seen at day 1 in both models). We hope that this is an answer to the first question.

In our association analysis we included a multitude of parameters, unfortunately we did not identify any strong associations and we therefore decided to describe the statistical analysis and the results in the supporting information. For clarity, we decided for the parameters measured once to only explicitly mention those parameters for which significance was reached (here defined as a p-value below 0.01, due to the multitude of analyses performed). Therefore, only the association between DPP4 and the length of hospital stay is discussed in 2.3. in the supporting information. The p-values of the parameters measured longitudinally are visually represented in S4 Fig.

Regarding vasopressor treatment, we evaluated the effect of the enzymatic activity on the duration of vasopressor treatment and whether this needed to be restarted. However, none of the enzymes had a significant effect on these parameters and are therefore not mentioned in the Results section of the supplementary materials.

We included the effect of the enzymes on the PaO2/FiO2 ratio, but again, no significance was reached here (see S4 Fig).

The data on whether the patients required dialysis or ventilation were available and were therefore included into the manuscript. See Table 1 and the supporting information S1 Appendix. In 1.3. Statistical analysis we have added (line 38) :

“The parameters ventilation and dialysis were analyzed using generalized linear mixed models.” and in 2.2. the following sentence was added (lines 86-87):

“Additionally, we fitted generalized linear mixed models to test if the enzymes were associated with the necessity of dialysis or ventilation, but no significant associations were found.”

In the main document a reference to the S1 Appendix has been added in line 353.

3.) The sample size is relatively small and there no validation cohort. This is a critical limitation of believing the credibility of the results herein.

We agree with the reviewer that our sample size is rather small. However, as also mentioned by the reviewer in the summary above, it only opens the door for these enzymes as biomarkers in septic shock. We believe that the data presented here is already of significant importance to be distributed within the scientific community, as these enzymes have never been studied before in this setting. We hope that it can be used as a kick-start for further research, including larger studies.

4.) The statistical analysis for the diagnostic aim should examine in addition only the day 1 values, since there is survival/non-survivor risk of bias due to early deaths after day 1. There could be non-random drop out due to death that could bias the results herein. The authors state in the discussion:” Additionally, it should be taken into consideration that only patients, of whom samples on days 1, 3, 5 and 7 were available, were included in the study.”

We fully agree with the reviewer and we have indeed mentioned this in the Discussion. Unfortunately, we do not have the samples of other patients than the 40 samples studied here. We acknowledge that this is a limitation of our study and we will explicitly mention this in the text (lines 414-417).

“A limitation of this study is that only samples from patients who stayed in the ICU during the entire week, were available. Follow-up studies should include all patients diagnosed with sepsis. This may even increase the biomarker value of the enzymes studied here.”

However, we believe that our results should already be disseminated within the scientific community, as we stated above.

5.) PRCP modifies/ presumably lowers angiotensin II levels (as per Yang HYT, Erdös EG, Chiang TS. New Enzymatic Route for the Inactivation of. Angiotensin. Nature. 1968;218(5148):1224–6) and angiotensin II is a recently approved therapy for septic shock(1, 2) so that has greater relevance for possibly discovering a predictive biomarker for response to angiotensin II. PRCP levels were increased in septic shock and higher levels were associated with a poor prognosis in your study. Furthermore, there are good studies of abnormalities of the renal-angiotensin-aldosterone (RAAS) system in sepsis(3). Accordingly, it would be useful to examine the PRCP association with markers of RAAS actively (plasma rein activity etc.). Of note, PRCP was not a great diagnostic (Low AUC) but was a good biomarker of prognosis – this suggests that the RAAS is a better marker of recovery or not from septic shock. Also, pre-sepsis use of ACE inhibitors could modify the PRCP activity – do you have that information available?

We agree with the reviewer that PRCP can be a promising predictive biomarker for response to angiotensin II treatment, as high PRCP levels (seen in septic shock patients) can theoretically reduce circulating Ang II levels. Therefore, more developed research is necessary to investigate whether high circulating PRCP levels have an influence on Ang II levels and administrated Ang II. In a recent study in mice, it was seen that after administration of Ang II, the Ang II conversion to Ang 1-7 in the circulation was independent of PRCP and ACE2, but dependent of PREP (Serfozo et al. Hypertension: 2020;75:173–182). As the reviewer suggested, it is therefore of great relevance to examine the PRCP association with markers of the renin-angiotensin system and definitely with Ang II and Ang 1-7 levels. In this study, we do not have these data available and there are no longer samples available to measure these parameters. We thank the reviewer for this interesting feedback and suggestion, but research into predictive biomarkers falls outside the scope of this study. However, we included an extra paragraph in the discussion to state this interesting hypothesis (lines 374-380):

“We report that PRCP activity is increased in the septic shock patient group on day 1. As PRCP is an angiotensin II degrading enzyme and angiotensin II is recently approved as therapy for septic shock [27,28], PRCP has the potential to become a promising predictive biomarker for response to angiotensin II. Angiotensin II treatment probably will not have the desired effect in patients with high PRCP activity. A dedicated study, including associations between PRCP activity and angiotensin II levels, angiotensin (1-7) levels and ideally other markers of the renin-angiotensin system is needed to confirm this hypothesis.”

Also, in the introduction we mention that Ang II is recently approved as therapy for septic shock, including the references suggested by the reviewer (Senatore, 2019 and Khanna, 2017) (lines 98-101):

“PRCP is involved in the regulation of blood pressure and hypotension is common in sepsis and septic shock, moreover, angiotensin II is recently approved as therapy for distributive shock. Therefore, a possible role for PRCP in the pathogenesis of sepsis and septic shock is conceivable.”

The overall association between PRCP and survival time (p=0.043), formally tests the null hypothesis that all 4 regression coefficients (for day 1, day 3, day 5 and day 7) from the Cox Proportional Hazard model are simultaneously zero. However, none of the regression coefficients from the separate days is significant, and there is no clear trend in the effect of PRCP on survival when going from day 1 to day 7. Therefore, the effect of PRCP on survival remains unclear at the moment. Due to the multitude of hypotheses tested in this study, it cannot be excluded that this association represents a false positive finding.

We stated in the text: “An overall effect on survival time was found for PRCP and PREP (p = 0.043 and p = 0.048, respectively) in the Cox proportional hazard model. Despite this overall effect, none of the enzymes at the individual time points was significantly associated with mortality before day 90”: “An overall association with PREP and PRCP was seen in the Cox proportional hazard model. However, finding such p-values is not surprising given the large number of p-values tested here.”

To be more clear, we adapted the text to (lines 425-428):

“An overall association with PREP and PRCP was seen in the Cox proportional hazard model. However, given the large number of associations tested here, the reported p-values are not very promising and therefore we cannot conclude from this data that PRCP or PREP are good prognostic biomarkers for sepsis.”

We do not have data available on the pre-sepsis use of ACE-inhibitors. Therefore, we cannot investigate whether the use of ACE inhibitors has an effect on the PRCP activity. Searching the literature, we cannot find information regarding the effects of ACE-inhibitors on PRCP activity, except that a polymorphism (E112D) in the PRCP gene has been associated with blood pressure response to benazepril (Zhang et al. Chin Med J 2009;122:2461-5). Because both, ACE and PRCP, are involved in the (alternative) renin-angiotensin system, ACE inhibitors can possibly have an effect on the PRCP activity, therefore more research is needed which falls outside the scope of this study.

MINOR

1.) Results. Study populations paragraph: would be useful to state the mortality rates in the Table 1 and that paragraph.

We have added a sentence describing the mortality at day 90 to the Study populations (line 231).

“Additionally, 13 patients had died at day 90.”

This information has also been included in Table 1.

We kept the sentence “At day 90, 13 people had died (32.5%).” in the Survival analysis section, to remind the reader.

2.) 30 healthy controls aren’t mentioned until the results section. These should be introduced and described in methods (re how and where they were recruited). They should also be described re age, gender etc. in Table 1.

To avoid confusion with the ICU control group, we decided to provide the information on the healthy controls in the supporting information. We acknowledge that they should be mentioned earlier in the main text. Therefore, we have added a sentence to the Materials and Methods, referring the reader earlier to the supporting information (lines 135-137).

“Additionally, 30 healthy controls were included, more information on this study group can be found in S1 Appendix and S1 Fig.”

Here we describe their age, gender distribution and how they were processed. A sentence describing their recruitment has been added (lines 6-8).

“These controls were recruited among employees of the University of Antwerp and their families and friends.”

3.) The figure legends appear in the text but should be moved to the end.

As stated in the submission guidelines and confirmed by the editor in his communication the figure legends should be inserted in the text, immediately following the paragraph in which the figure is first cited. Therefore, we have not moved the figure legends.

4.) Results, page 15. “For the septic shock patients day 1 was used.” I presume this refers to the ROC for diagnostic biomarker. Please clarify and introduce the rationale for this in methods.

We have changed the title of Table 2 to “Cutoff values based upon the ROC curves of DPP4, FAP, PRCP and PREP” and have added the rationale for the use of day 1 in the Materials and Methods section describing the ROCs (lines 200-203).

“ROC curves were computed to assess the four enzymes as possible biomarkers for septic shock. For the septic shock patients (n = 40) only day 1 was used. The patient’s condition on day 1 is clinically the most relevant, since it is of utmost importance to identify patients as soon as possible to improve their outcome.”

5.) Discussion:” Since septic shock has a high mortality, another cutoff value was calculated in which no patients would be missed. In this case, each of the enzymes gives an insufficient specificity as more than half of the ICU control patients would be falsely diagnosed with septic shock.” This is not clear and requires better explanation.

In the Materials and Methods where we described the generation of the ROCs, we explain now how we defined a cutoff that would include all septic shock patients. This would mean that the sensitivity had to be 100% (lines 203-207).

“ROC curves were generated using the ICU control group. Cutoff values were determined using two methods, one in which equal weight is given to false positives and false negatives, also called the Youden index, and another one in which all patients with sepsis are identified, meaning that the sensitivity reaches 100%.”

We have also decided to repeat this difference between the two methods in the Results section, so it is certainly clear for the reader (lines 278-282).

“Cutoff values, either determined using the Youden index (giving equal weight to false positives and false negatives) or to obtain 100% sensitivity (to identify all patients with sepsis), are listed in Table 2 together with the sensitivity, specificity, negative and positive predictive value and the negative and positive likelihood ratios.”

In the discussion, we finally also added that we mean that the sensitivity had to be 100% (line and we also added that a lot of false positives would lead to unnecessary treatment (lines 396-400).

“Since septic shock has a high mortality, another cutoff value was calculated in which no patients would be missed (sensitivity is 100%). In this case, each of the enzymes gives an insufficient specificity as more than half of the ICU control patients would be falsely diagnosed with septic shock and would receive unnecessary treatment.”

6.) The Kaplan-Meier 90-day survival curves for the best PSE activity cutoffs should be shown as a figure in the main manuscript. The reader would like to see visually how PSE levels are associated with prognosis in septic shock.

Visualizing how PSE levels are associate with prognosis would technically be possible using Kaplan-Maier curves (splitting groups according to some cutoff for the PSE). Cutoffs could be based upon (for example) a ROC curve that uses PSE as predicting variable. However, when it comes to predicting the survival, the current data suffer from an ascertainment bias. Patients quickly recovering and leaving the ICU within one week, are not included. Patients who passed away within one week are also not included. As previously stated, we have included this as a weakness of the study (line 414-415). Therefore, although it would be technically possible to generate these curves, they may generate a biased view on the relation between PSE and survival at 90 days. Therefore, we chose not to show them.

References suggested by the reviewer:

1. Khanna A, English SW, Wang XS, Ham K, Tumlin J, Szerlip H, et al. Angiotensin II for the Treatment of Vasodilatory Shock. N Engl J Med. 2017;377(5):419-30.

2. Senatore F, Jagadeesh G, Rose M, Pillai VC, Hariharan S, Liu Q, et al. FDA Approval of Angiotensin II for the Treatment of Hypotension in Adults with Distributive Shock. Am J Cardiovasc Drugs. 2019;19(1):11-20.

3. Emdin M, Fatini C, Mirizzi G, Poletti R, Borrelli C, Prontera C, et al. Biomarkers of activation of renin-angiotensin-aldosterone system in heart failure: how useful, how feasible? Clin Chim Acta. 2015;443:85-93.

Reviewer #3:

We would like to thank the reviewer for the careful revision. Please find our answer below.

The manuscript highlights markers that have the potential to become biomarkers for sepsis, for which currently no concrete association exists (as mentioned by authors as well). However, the associations described for most of the proline-specific enzymes were not very strong. As the authors mentioned further research is warranted. This manuscript will greatly benefit from more detailed analysis of age, gender, ethnicity to determine where the enzymes have any stronger correlation in any specific group. It is understandable data from human patients are not easily available however, to correlate the enzymes with progression in sepsis, this study requires control known inflammatory/anti-inflammatory patterns demonstrated side by side. This analysis will reveal the correlation of the inflammatory status, disease progression of sepsis 2 patient along with the enzymes can be potential biomarkers.

In our association study we included several (anti-) inflammatory parameters, such as IL-1β, IL-1RA, IL-6, IL-8, IL-10, TNFα, IFNγ and leucocytes. These were all the parameters regarding the immune system that were recorded for these 40 patients. As mentioned in the text (lines 353-355), we did not identify strong associations with the above mentioned parameters, except for PREP with IL-1RA.

Further dividing our 40 patients into groups for a detailed analysis of age, gender or ethnicity will result in small groups that will not have enough statistical power and we have therefore opted not to include these analyses.

In future efforts other parameters can be included as well such as IL-17, IL-27, IL-33, which recently have been described to contribute to the immunological dysfunction in sepsis [5]. In addition, as we have mentioned in the manuscript, larger and more diverse patients should be recruited to validate the data presented here.

Regarding the diagnostic value (adjusted upon recommendation of reviewer 2) (lines 404-417):

“Therefore, in the future, it would be very interesting to study larger groups of patients and to prospectively include all patients at high risk of developing sepsis and measure, for example, their DPP4- and FAP-activities on a daily basis. … A limitation of this study is that only samples from patients who stayed in the ICU during the entire week, were available. Follow-up studies should include all patients diagnosed with sepsis. This may even increase the biomarker value of the enzymes studied here.”

Regarding the prognostic value (lines 430-431):

“A stronger association might be identified when the DPP4-activity on day 1 is measured in a larger and more diverse sepsis patient population.”

References:

1. Shih C-J, Wu Y-L, Chao P-W, Kuo S-C, Yang C-Y, Li S-Y, et al. Association between use of oral anti-diabetic drugs and the risk of sepsis: a nested case-control study. Sci Rep. 2015;5:15260.

2. Van Der Velden VH, Naber BA, Van Hal PT, Overbeek SE, Hoogsteden HC, Versnel MA. Peptidase activities in serum and bronchoalveolar lavage fluid from allergic asthmatics--comparison with healthy non-smokers and smokers and effects of inhaled glucocorticoids. Clin Exp Allergy. 1999;29(6):813–23.

3. Lefebvre J, Murphey LJ, Hartert T V., Shan RJ, Simmons WH, Brown NJ. Dipeptidyl peptidase IV activity in patients with ACE-inhibitor-associated angioedema. Hypertension. 2002;39(2 II):460–4.

4. Dellinger R, Levy M, Rhodes A. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med. 2013;41(2):580–637.

5. Morrow KN, Coopersmith CM, Ford ML. IL-17, IL-27, and IL-33: A Novel Axis Linked to Immunological Dysfunction During Sepsis. Front Immunol. 2019;10:1982.

Attachment

Submitted filename: Response to reviewers.docx

Click here for additional data file. (39.7KB, docx)

Decision Letter 1

Paul Proost

26 Mar 2020

Dysregulated activities of proline-specific enzymes in septic shock patients (sepsis-2)

PONE-D-19-30816R1

Dear Dr. De Meester,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Paul Proost, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thanks for addressing my concerns. My concerns have been resolved. I judged this revised manuscript 'accept'.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Acceptance letter

Paul Proost

7 Apr 2020

PONE-D-19-30816R1

Dysregulated activities of proline-specific enzymes in septic shock patients (sepsis-2)

Dear Dr. De Meester:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Paul Proost

Academic Editor

PLOS ONE

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    S1 Fig. Plasma activity of DPP4, FAP, PRCP and PREP in healthy and ICU controls.

    The plasma activity of DPP4 (A), FAP (B), PRCP (C) and PREP (D), expressed in U/L (median with interquartile range), measured in a healthy control group (n = 29–30) and the ICU control group (n = 22). Differences between the two control groups were tested using Mann-Whitney U tests. Abbreviations used: DPP4: dipeptidyl peptidase 4; FAP: fibroblast activation protein α; NS: non-significant; PRCP: prolylcarboxypeptidase; PREP: prolyl oligopeptidase; U/L: units per liter.

    (DOCX)

    Click here for additional data file. (201.5KB, docx)
    S2 Fig. Principal component analysis.

    Plot of the first two principal components scores for the ICU control group and the healthy control group, based upon PREP, PRCP, FAP and DPP4. Abbreviations used: ICU; intensive care unit; PC: principal component.

    (DOCX)

    Click here for additional data file. (91.8KB, docx)
    S3 Fig. Receiver Operating Characteristic (ROC) curves of the combination FAP and DPP4 and FAP with PREP.

    (DOCX)

    Click here for additional data file. (97.9KB, docx)
    S4 Fig. Visual representation of the statistical analysis of the studied associations.

    (A) Dot plot of the log-transformed p-value for the interaction (pInt) for the longitudinally measured parameters. The dashed line indicates a p-value of 0.01. Significant interactions were obtained for PRCP with total bilirubin and TNFα. P-values are based upon linear mixed models. (B) Dot plot of the log-transformed p-value for the main effect (pMain) for the longitudinally measured parameters. In case of a significant interaction between time and the parameter, no main effect p-value is shown in the dot plot. The dashed line indicates a p-value of 0.01. Linear mixed models were fitted to obtain the p-values. (C) R² for the longitudinally measured parameters. R2 is only depicted when pInt ˃ 0.01. P-values are based upon linear mixed models. ○ DPP4; + PRCP; △ FAP; ● PREP. Abbreviations used: DPP4: dipeptidyl peptidase 4; FAP: fibroblast activation protein α; I-FABP: intestinal fatty acid-binding protein; IFNγ: interferon γ; IL: interleukin; IL-1RA: interleukin-1 receptor antagonist; MAP: mean arterial pressure; NIR, noradrenalin infusion rate; PaO2/FiO2 ratio: ratio of arterial oxygen partial pressure to fractional inspired oxygen; PRCP: prolylcarboxypeptidase; PREP: prolyl oligopeptidase; TNFα: tumor necrosis factor α.

    (DOCX)

    Click here for additional data file. (207KB, docx)
    S1 Table. Characteristics of the non-septic shock ICU control group.

    A total of 22 patients were included for this study.

    (DOCX)

    Click here for additional data file. (15.6KB, docx)
    S2 Table. The enzyme activity of DPP4, FAP, PRCP and PREP.

    Enzymatic activities were determined in EDTA-plasma of septic shock patients on day 1, 3, 5 and 7 and non-septic shock ICU control patients. Data are represented as median (interquartile range). Abbreviations used: DPP4: dipeptidyl peptidase 4; FAP: fibroblast activation protein α; PRCP: prolylcarboxypeptidase; PREP: prolyl oligopeptidase; U/L: units per liter.

    (DOCX)

    Click here for additional data file. (18.3KB, docx)
    S3 Table. Cutoff values receiver operating characteristic curve of dipeptidyl peptidase 4 (DPP4).

    The 1-specificity, sensitivity, negative predictive value (NPV), positive predictive value (PPV), positive likelihood ratio (LR+), negative likelihood ratio (LR-) and Youden index for every cutoff value expressed in U/L for DPP4. For the septic shock patients day 1 was used. Blue indicates the maximum for the Youden index. Red indicates where the sensitivity reaches 1. In the case of DPP4, a value lower than the indicated cutoff would suggest a diagnosis with septic shock. ICU controls: n = 22; Septic shock patients: n = 40.

    (DOCX)

    Click here for additional data file. (24.6KB, docx)
    S4 Table. Cutoff values receiver operating characteristic curve of fibroblast activation protein α (FAP).

    The 1-specificity, sensitivity, negative predictive value (NPV), positive predictive value (PPV), positive likelihood ratio (LR+), negative likelihood ratio (LR-) and Youden index for every cutoff value expressed in U/L for FAP. For the septic shock patients day 1 was used. Blue indicates the maximum for the Youden index. Red indicates where the sensitivity reaches 1. In the case of FAP, a value lower than the indicated cutoff would suggest a diagnosis with septic shock. ICU controls: n = 22; Septic shock patients: n = 40.

    (DOCX)

    Click here for additional data file. (24.4KB, docx)
    S5 Table. Cutoff values receiver operating characteristic curve of prolylcarboxypeptidase (PRCP).

    The 1-specificity, sensitivity, negative predictive value (NPV), positive predictive value (PPV), positive likelihood ratio (LR+), negative likelihood ratio (LR-) and Youden index for every cutoff value expressed in U/L for PRCP. For the septic shock patients day 1 was used. Blue indicates the maximum for the Youden index. Two maxima were obtained for PRCP, in this case the cutoff that would yield the highest sensitivity was selected, as that would identify more patients with septic shock. Red indicates where the sensitivity reaches 1. In the case of PRCP, a value higher than the indicated cutoff would suggest a diagnosis with septic shock. ICU controls: n = 22; Septic shock patients: n = 40.

    (DOCX)

    Click here for additional data file. (23.7KB, docx)
    S6 Table. Cutoff values receiver operating characteristic curve of prolyl oligopeptidase (PREP).

    The 1-specificity, sensitivity, negative predictive value (NPV), positive predictive value (PPV), positive likelihood ratio (LR+), negative likelihood ratio (LR-) and Youden index for every cutoff value expressed in U/L for PREP. For the septic shock patients day 1 was used. Blue indicates the maximum for the Youden index. Red indicates where the sensitivity reaches 1. In the case of PREP, a value higher than the indicated cutoff would suggest a diagnosis with septic shock. ICU controls: n = 22; Septic shock patients: n = 40.

    (DOCX)

    Click here for additional data file. (23.1KB, docx)
    S1 Appendix. Dysregulated activities of proline-specific enzymes in septic shock patients(sepsis-2).

    (DOCX)

    Click here for additional data file. (30.2KB, docx)
    Attachment

    Submitted filename: Response to reviewers.docx

    Click here for additional data file. (39.7KB, docx)

    Data Availability Statement

    The SPSS file containing all raw data is available from Figshare (DOI 10.6084/m9.figshare.10252877.). Processed data are within the manuscript and its Supporting Information Files.

    Articles from PLoS ONE are provided here courtesy of PLOS

    RESOURCES