Blood RNA biomarkers and a point-of-care elastase assay for detecting host immune activation in suspected sepsis: Trajectory matters Authors: John E. Lafleur, Eduard Shaykhinurov, John Perkins, Richard Wargowsky, Kevin Jaatinen, Mary Pasquale, Grace Holloway, David Yamane, Akhil Patel, Daniel King, Andrew Meltzer, Ryan Heidish, Soroush Shahamatdar, Aditya Loganathan, Tarun Loganathan, Taylor Bolden, Michael Zane Hayden, Aditya Maddali, Jennifer Goldman, Zachary Falk, Tisha Jepson, Avery League, Timothy A. McCaffrey, Claudine Irles, Tomasz Kaminski, Tomasz Kaminski, Tomasz Kaminski Published: 2025 ABSTRACT Introduction/HypothesisThe chances of survival from sepsis are improved by early diagnosis and treatment. In the prospective SENSOR study, RNA biomarkers of innate immune neutrophil activation were examined in emergency department (ED) patients triggering an automated sepsis alert. We hypothesized that higher levels of blood RNA biomarkers related to neutrophil activation would be associated with progression to more severe forms of sepsis.MethodsAdult patients in the ED triggering a sepsis alert were consented, enrolled, and study samples were obtained during the ED visit. Additionally, study samples were collected from a convenience sample of 16 adult non-ED controls and 8 other adults with self-described infectious illnesses. Blood samples were drawn in RNA preservative (Tempus) and whole blood RNA was analyzed by droplet digital PCR (ddPCR) for RNA transcripts related to neutrophil response to infection by bacterial (DEFA1; ALPL, IL8RB/CXCR2), and viral (IFI27, RSAD2) pathogens. Bacterial burden in blood was quantitated by ddPCR of 16S ribosomal DNA. Separately, neutrophil elastase was measured in immunomagnetically captured CD66b+ neutrophils by a novel point-of-care device.ResultsPatients were grouped using both ‘Sepsis-2’ SIRS criteria, adjudicated by independent physicians, and ‘Sepsis-3’ criteria which uses a qSOFA score for categorizing the severity of illness. Across 72 enrolled sepsis alert patients, 62.5% showed positive RNA biomarkers for bacterial infection, and 8.3% were positive for viral markers, with only 2 cases that showed only a viral signal. Septic patients showed a 4-fold increase in RNA markers vs. those without infection (p < 0.05). However, no significant differences were observed in RNA levels between those with sepsis vs those with more severe forms of sepsis. Likewise, RNA biomarker levels did not discriminate patients with qSOFA≥2 from qSOFA = 1. In a subset of patients with zero and three hour blood samples (n = 27) it was found that changes in RNA levels (up or down), or neutrophil elastase activity, was strongly associated with progression to more severe forms of sepsis or a qSOFA score of ≥2.ConclusionsPatients progressing to more severe forms of sepsis did not have higher absolute levels of neutrophil activation RNA biomarkers (or higher bacterial burden in blood) compared to patients with sepsis. However, a change in RNA biomarkers or elastase between zero and three hours was strongly indicative of progression to more severe forms of sepsis. Introduction Historically, sepsis had been defined by physiologic parameters, including systemic inflammatory response syndrome (SIRS) criteria [1], which lacked specificity, and while relatively sensitive, still failed to capture almost 13% of sepsis cases [2]. In 2016, sepsis was redefined as ‘severe organ dysfunction resulting from immune dysregulation in response to infection’ [3]. This updated definition, known as Sepsis-3, led to greater diagnostic specificity, but loss of sensitivity, due to a more restricted list of clinical and laboratory findings that qualify an infected patient as ‘septic’ [4,5]. Despite exhaustive research, improved means for screening for sepsis have not materialized. Recent algorithmic approaches based upon the electronic health record (EHR) have failed to improve diagnosis [6]. The Centers for Medicare & Medicaid Services (CMS) mandates the collection of quality metrics around sepsis care based on a pre-Sepsis-3 framework, and most medical systems continue to screen for sepsis using pre-2016 ‘SIRS’ criteria [7], The Sepsis-3 definition relies on parameters included in a clinical predictive tool known as the ‘Sequential Organ Failure Assessment’ (SOFA), and ‘SOFA’ requires lab values which are usually not available in the time-frame required in the ED for early recognition of patients who may be septic. Thus, an alternative measure, quick SOFA (qSOFA), has been adopted. However, past research shows that the sensitivity of the qSOFA score during ED triage is as low as 31% for predicting admission to the Intensive Care Unit (ICU) and 60% for predicting mortality at 48 hours [4]. ‘Severe sepsis’ which is important in ‘Sepsis-2’ (sepsis plus end organ dysfunction), was removed in ‘Sepsis-3’. For purposes of convenience, because results in the present study were analyzed according to criteria from both ‘Sepsis-2’ and ‘Sepsis-3’, in the following the term ‘more severe form of sepsis’ has been used to denote those with ‘severe sepsis/septic shock’ under ‘Sepsis-2’, and those with a qSOFA score ≥2 in ‘Sepsis-3’. There is a well-recognized need for novel biomarkers and a better understanding of the dynamics of the transition to more severe forms of sepsis [8]. There has been an extensive search for sepsis biomarkers, with possibly ~250 identified over the past few decades, but few have proven better than a functional measure of organ failure, such as the SOFA score [9–11], which is largely an ICU-based measure not suitable for application to emergency department patients where most sepsis is diagnosed. Adherence to the tenets of ‘early goal-directed therapy’ [12], had been considered beneficial, however its value compared to ‘usual care’ is now questioned [13]. About 20% of deaths worldwide [13] are due to sepsis, and about 20% of patients presenting to the ED trigger sepsis alerts based upon current screening practices—though many turn out to have no infection [6]. Aggressively treating all ED patients who trigger sepsis alerts is a costly, time-intensive means for improving survival. Thus, improved diagnostics are needed to identify the patients that require aggressive treatment for sepsis. A large and growing body of literature implicates excessive neutrophil activation and cell death in the progression of sepsis (reviewed in [14]). In previous work, we employed whole transcriptome profiling to identify RNA biomarkers in peripheral blood from patients with appendicitis or respiratory infections [15], and in various types of COVID-19 patients [16]. Relatively large fold-change increases (~20-fold) were identified in RNAs coding for neutrophil primary granule markers, especially neutrophil α-defensins (DEFA1) in infections such as pneumonia. In appendicitis patients, there was marked increases in RNAs coding for IL8 receptor-ß (IL8RB/CXCR2), and secondary granule proteins such as alkaline phosphatase (ALPL) [15]. Neutrophil RNA markers of viral infection, IFI27 and RSAD2, were identified from COVID19 cases and the literature to provide a more complete picture of the infection profile [16–18]. Recent work suggests the trajectory of sepsis biomarkers over time can provide more diagnostic value than measurements of these markers at a single time point [9,19]. In a subset of patients in the current study samples were drawn at time zero (patient enrollment in the study), and again three hours later. Additionally, to further understand the interactions between pathogens and host immunity, bacterial 16S and fungal 23S DNA were assayed, as were DEFA1 protein levels. Here we report on a set of promising RNA biomarkers in whole blood, and elastase activity of circulating CD66b neutrophils, to evaluate whether they can help identify ED patients at risk for developing more severe forms of sepsis early in the ED course, when resuscitative measures are most useful for preventing the worst outcomes. Materials and methods Enrollment and blood draw This prospective, observational study, shown schematically in Fig 1, was approved by the Institutional Review Board of The George Washington University (#NCR213645). All subjects, or their legally authorized surrogate, provided written informed consent, witnessed by a study team member in the period from 11/16/2022 through 8/21/2024. Patients presenting to The George Washington University Hospital ED were screened for inclusion criteria. Inclusion criteria consisted of a ‘sepsis alert’ trigger generated automatically by the Cerner EHR using modified SIRS criteria shown in Fig 1 [20]. The Cerner sepsis alert uses The St. John’s Sepsis alert that is similar to SIRS criteria: temperature >38.3˚C or <36˚C, heart rate > 95 bpm, respiratory rate ≥22 breaths/min, white blood cell count >12 or <4 K/ul, or glucose >141 and <200 mg/dl. The alert registers when 3 criteria are met, or 2 criteria and evidence of organ failure: hypotension, elevated lactate, elevated bilirubin or creatinine. Design of the SENSOR study. Patients presenting to the emergency department (ED) were screened manually and by an electronic health record algorithm to detect a risk of sepsis using fever, heart rate, respiration, white blood cell (WBC) count, and other criteria. Patients meeting criteria were consented, blood drawn into an RNA stabilizer for quantitation of RNA biomarkers of neutrophil activation using droplet digital PCR for the bacterial response markers RNA: DEFA1, ALPL and IL8RB/CXCR2, and the viral response marker RNAs: IFI27 and RSAD2. Blood was also collected in EDTA anticoagulant for determination of neutrophil elastase using a novel immunomagnetic isolation and kinetic elastase assay (CyBIS). The clinical course was not altered, but all relevant clinical results were recorded and used to adjudicate the diagnosis as involving No Infection, Sepsis, or more severe forms of sepsis (Sepsis/Shock). Normal controls and subjects with self-reported infections were sampled from medical center personnel. After informed consent was obtained, a sample of whole blood was drawn into 2 Tempus Blood RNA tubes (ThermoFisher, 3 ml each), and 1 BD Vacutainer K2 EDTA tubes (BD Biosciences, 5 ml) for neutrophil capture and elastase assay. A subset of patients (n = 27) had a second blood sample drawn 3 hours after the initial sample in order to assess temporal changes in the biomarkers. Study personnel subsequently processing the blood samples were blinded to the subjects’ clinical characteristics and outcomes by anonymous coding of the samples. Adjudication of infection and sepsis status The presence of infection and sepsis was adjudicated from the medical records by two physicians unrelated to the ED care of the patients, and in the case of discordance was reviewed by a third physician. Determination of infection was determined by microbiologic testing (culture and non-culture based). In cases of negative microbiologic testing, clinical or radiographic diagnoses consistent with an infection were considered positive, for example perforated diverticulitis or pneumonia. More severe forms of sepsis were determined by an active systemic infection with signs of end organ dysfunction, including hypotension, acute kidney injury, altered mental status, or lactic acidosis. In a second analysis, classification was determined by calculating the qSOFA scores using respiratory rate ≥22, systolic blood pressure <100, and altered mentation (Glasgow Coma scale, GCS < 15), consistent with Sepsis-3 criteria [3]. The ‘No Infection’ subgroup was defined as patients that triggered the sepsis alert, but post-hoc analysis adjudicated that no infection was present by all available testing. Concordance of infection and sepsis adjudication The inter-observer concordance for the presence of an infection was 87.5%, but the concordance for more severe forms of sepsis was only 69.4%, and concordance on both parameters was only 65.8%. While this seems low, it is consistent with other published data. In large clinical trials of other infection diagnostics, inter-observer concordance was found to be only 71.2% for both bacterial and viral infection adjudication [21]. Neutrophil-associated elastase activity by cytocapture (CyBIS) The method for automated cytocapture and neutrophil elastase assay is described previously [22]. Briefly, blood collected in EDTA anticoagulant (75 µl) is pelleted at 300 x g for 5 min at 4˚ C to remove soluble plasma ligands that can block the antibody beads. The pelleted cells are resuspended in bead binding buffer, containing EDTA, and are mixed with CD66b antibody-coated magnetic beads and then loaded into the CyBIS device. The device incubates the blood and bead mix for 20 minutes, collects magnetic beads with cells, and washes away unbound cells and proteins 3 times. The captured cells are lysed and assayed by time-dependent cleavage of a colorimetric elastase substrate, yielding a rate constant (Vmax, mOD/min) that is proportional to elastase levels in the captured cells. Cell counts were performed using 10 µl in a hemocytometer at 200X using phase contrast. The identities of the isolated cells were confirmed by fixing with 4% formaldehyde in 1X PBS, and using a May-Grünwald stain (Wright-Giemsa stain with eosin Y). Whole blood RNA extraction The Tempus Blood RNA preservation tubes (Applied Biosystems) were stored at −80˚ C and thawed for 1 hour at RT. RNA was purified using the Tempus Spin RNA Isolation Reagents (Applied Biosystems). The Tempus blood lysate was spun at 3000 x g for 30 minutes, and the pellet was resuspended and transferred onto purification columns. Following on-column DNAse treatment using TurboDNase (Ambion) and a series of washing and drying steps, purified RNA was eluted with nuclease-free water. RNA levels were quantified using the NanoDrop ND-1000 Spectrophotometer and quality was determined by capillary electrophoresis on the Agilent 2100 Bioanalyzer. RNA Biomarker Digital Droplet PCR (ddPCR) Prior microarray and RNAseq analysis of whole blood RNA from patients with various types of infections identified RNA markers of viral and bacterial infections [15,16,23]. Primers and internal fluorescent probes for the RNAs known to be responsive to bacterial infections (DEFA1, IL8RB/CXCR2 and ALPL) and to viral infections (IFI27 and RSAD2), and to ACTB as a reference control can be found in Supplementary Data S1 Fig. The DNAse-treated RNA from whole blood (200 ng) was reverse transcribed into cDNA using gene-specific primers (250 nM) with the RNAse H+ reverse transcriptase and 5X Supermix from the iScript Select cDNA Synthesis Kit (Bio Rad). Each of the markers were paired to make a total of three primer-probe master mixes (4 µM forward primer; 4 µM reverse primer, 1.12 µM quenched fluorescent probe) with water and the ddPCR Supermix (Bio Rad). The BioRad Automated Droplet Generator typically produced >15,000 nanoliter-sized droplets per well, the target transcripts were amplified in each droplet by PCR, and each droplet was read separately by the BioRad QX200 Droplet Reader. BioRad Quantasoft software was used to set the threshold for positive droplets. Transcript counts for the RNA biomarkers (DEFA1, ALPL, etc.) were normalized per sample by dividing by the number of ACTB copies to further minimize the effects of changes in cell counts or in RNA yield. Defensin 1 protein in plasma A subset of patients and controls had plasma from EDTA blood stored at −80˚C. Thawed plasma was analyzed for levels of the defensin-α1 protein by ELISA (RayBio ELH-DEFA1). Plasma was diluted 1:1 with assay buffer and reacted in the sandwich ELISA assay and compared to a standard curve of defensin-α1 protein (0–250 ng/ml). The observed values were corrected for dilution and reported in ng/ml. 16S Bacterial and 28S Fungal ribosomal DNA quantitation DNA was isolated from the Tempus tubes via a modified phenol/chloroform extraction procedure [24]. First, total lysis of any remaining viable cells was ensured by the addition of an 8% SDS solution to the Tempus blood/preservative mixture and multiple cycles of freezing in a dry/ethanol bath immediately thawing at 95˚C [25]. Two rounds of extraction with a phenol/chloroform/isoamyl mixture (25:24:1 ThermoFisher) and a final cleanup with a chloroform/isoamyl (24:1) mixture were then performed. The DNA was then precipitated overnight at −80˚C in the presence of 3M sodium acetate and glycogen (8 µg/µl). 16S (Gram positive/negative) and 28S RRNA fragments were then quantified by ddPCR using primers with minimized human cross-reactivity and high sensitivity and specificity in patients with proven central line infections [26]. Statistical methods Continuous variables, such as gene expression levels, were tested for normality using the Jarque-Bera test. Parametic tests, such as the Student’s t-test, were used for continuous variables, while non-parametric tests, such as chi-square, were used for categorial variables such as sex. Fisher’s exact test was used when categorical variables contained a group of less than 5. Error bars are shown using the standard error of the mean (S.E.M.). Intercorrelations of measures was computed using Pearson’s R. Results Clinical characteristics of sepsis patients and controls A total of 73 ED subjects were enrolled, but 1 subject failed to yield sufficient RNA for analysis, leaving 72 ED subjects that divided about evenly between subjects adjudicated with more severe forms of sepsis (n = 22), sepsis (n = 26), or judged to have no infection (n = 24). The clinical, demographic, and laboratory findings are shown in Table 1, and reflect the type of changes known for sepsis and infection including age, white blood cell count, neutrophil count, and lactate levels. While trends in these were identified, differences between the groups were mostly not statistically different. Average lactic acid levels were higher than expected in the No Infection group (3.13 mmol/L) because of 2 patients with very high values (11–13 mmol/L) due to seizures or alcohol withdrawal. Normal control subjects, and non-ED subjects with self-reported infections, were more likely to be younger and white, without significant differences in gender or ethnicity (S1 Table). 10.1371/journal.pone.0338012.t001Table 1Demographic and lab values.SEPSIS/SEMSEPSISSEMNOSEM SHOCK INFECTION N 22–26–24– Age (yrs) 57.953.73*47.193.9348.714.38 Weight (kg) 88.285.7384.566.7794.916.34 Sex (% male) † 36.40–50.00–37.50– Race (% white) † 27.27–38.46–29.17– Body Temperature (˚F) 99.730.50*98.850.3798.400.35 Systolic Blood Pressure (mmHg) 129.779.00131.884.95*146.525.72 Diastolic Blood Pressure (mmHg) 75.864.3977.403.1685.914.78 Heart Rate (beats per minute) 118.455.84105.324.11114.524.01 Respiratory Rate (breaths per minute) 23.761.8520.051.1322.051.70 White Blood Cells (K/µL) 14.871.3813.461.1112.361.91 Red Blood Cells (M/µL) 4.130.234.040.253.930.18 Hemoglobin (g/dL) 12.240.7211.730.6711.210.56 Hematocrit (% packed RBC) 37.501.9836.052.0035.261.59 Mean Corpuscular Volume (fL) 91.401.8790.181.5790.512.35 Mean Corpuscular Hemoglobin (pg) 29.600.6629.270.5928.770.98 MCH Concentration (g/dL) 32.410.3732.460.3931.680.49 Red Cell Distribution Width (mean %) 15.370.6814.530.4615.470.59 Platelet Count (K/µL) 260.5238.89327.6242.27*222.2135.96 Bands (%) 5.752.370.270.200.330.22 Neutrophils (%) 76.802.79*78.961.65*64.434.37 Lymphocytes (%) 13.502.4611.461.19*22.133.45 Monocytes (%) 7.470.807.670.568.611.06 Eosinophils (%) 0.380.15*0.880.331.730.53 Basophils (%) 0.070.060.080.06*0.450.18 Immature Granulocytes (%) 0.850.13*0.620.130.400.03 Urine Leukocyte Esterase (% +) ‡ 50.00–*25.00–7.690.19 Urine Nitrate (% +) ‡ 22.22–16.67–0.00– Urine White Blood Cells (% +) † 76.19–*41.67–39.13– Urine White Blood Cell Count 191.0695.3141.7013.991.780.33 Blood Culture (% +) ‡ 40.00–*5.00–0.00– C-reactive Protein (mg/L) 198.5635.32146.7622.51186.7328.05 Lactic Acid (mmol/L) 3.590.511.880.183.130.82Notes: Values are means unless otherwise indicated. Percent values calculated using the total number of patients with available data for each variable as the denominator. Asterisks indicate statistically significant differences between the sepsis/shock vs. sepsis and/or no infection groups and the sepsis vs. no infection groups based on p-values < 0.05 using equal variance two-tailed t-tests for continuous variables and either chi-squared tests or Fisher’s Exact Tests for categorical variables based on sample size. SEM = Standard Error of the Mean. Parameters for Controls are shown in S1 Table.† Chi-squared test‡ Fisher’s Exact Test (note: no infection group excluded in the blood culture and urine nitrate tests to avoid categories where N = 0) Demographic and lab values. Notes: Values are means unless otherwise indicated. Percent values calculated using the total number of patients with available data for each variable as the denominator. Asterisks indicate statistically significant differences between the sepsis/shock vs. sepsis and/or no infection groups and the sepsis vs. no infection groups based on p-values < 0.05 using equal variance two-tailed t-tests for continuous variables and either chi-squared tests or Fisher’s Exact Tests for categorical variables based on sample size. SEM = Standard Error of the Mean. Parameters for Controls are shown in S1 Table. † Chi-squared test ‡ Fisher’s Exact Test (note: no infection group excluded in the blood culture and urine nitrate tests to avoid categories where N = 0) Blood RNA biomarkers and neutrophil elastase as a function of infection, sepsis, and more severe forms of sepsis RNA Biomarkers. The blood RNA biomarkers were quantitated in each patient and then analyzed as a function of their clinically adjudicated status. Additionally, normal control subjects (n = 16) and non-ED volunteers with self-described infections (n = 8) were similarly assayed for neutrophil elastase activity and blood RNA biomarkers. As shown in Fig 2, the RNA biomarkers showed stepwise increases from normal healthy controls to normal controls with self-reported infectious illness to the ED patients. However, there were not significant differences in either the RNA or elastase measures between patients that developed severe sepsis versus those with sepsis and an adjudicated infection. In fact, DEFA1 RNA levels were similar between more severe forms of sepsis (Sepsis/Shock), sepsis, and non-ED patient controls with self-reported infections. These results were similar in the second analysis, where there were no significant differences in either the RNA or elastase measures across the qSOFA groups. While the average levels of RNA biomarkers were not statistically different, a higher percentage of patients with more severe forms of sepsis were positive for any one RNA biomarker using preset thresholds: more severe sepsis (81.8%), sepsis (57.7%), and uninfected (50.0%). The patient-level, anonymous raw data is available in S2 Table. RNA biomarkers and cytocaptured neutrophil elastase levels in controls and ED patients. Panel A: RNA stabilized blood samples from Controls (n = 16), self-diagnosed but non-ED infections (Non-ED Infect n = 8), and ED patients adjudicated to have no infection (n = 24), sepsis (n = 26), or severe sepsis/septic shock (n = 22), were purified and quantitated by ddPCR for 6 host RNAs: ACTB1, a reference; DEFA1, ALPL and IL8RB/CXCR2 (AL + IL%), bacterial response markers; and IFI27 and RSAD2 (Viral%), viral response markers. The host response markers are expressed as a percent of ACTB reference (mean ± s.e.m., * = p < 0.05) Panel B: EDTA-treated blood was assayed for neutrophil elastase activity by immunomagnetic capture of neutrophils using CD66b-antibody coated beads followed by a kinetic assay for elastase that measures elastase activity over a 15 min period (Vmax, mOD/min). Panel C: The patients were regrouped according to their qSOFA scores and the RNA scores were re-averaged and compared between groups. Panel D: Cytocaptured neutrophil elastase activity is plotted according to the same qSOFA groupings as Panel C. Correlation of RNA biomarkers and clinical laboratory values. The correlation of RNA biomarkers and other clinical variables can be seen in Fig 3. It should be noted that RNA biomarkers are measured in such a way that variations in WBC count would have little impact on the RNA levels. The reasons for this are: 1) whole blood RNA contains significant RNA from RBC and platelets, thus minimizing the impact of WBC variation on total RNA yield, 2) regardless of RNA yield from the 3 ml of blood, a fixed amount of RNA is taken for ddPCR, and 3) the RNA biomarkers are normalized per ß-actin (ACTB) copies in the sample. Intercorrelation matrix of clinical and experimental parameters. The relationship between clinical laboratory values, such as the WBC Count, and experimental biomarkers, such as DEFA1 RNA levels, was evaluated across all 72 patients in the cohort. It is possible that some parameters, such as CRP, could have smaller n due to that test not being ordered on all patients. Color scale highlights negative (green) to positive (red) correlations. Correlations above ±0.35 are boxed to highlight possible associations. Dynamic levels of immune activation in patients progressing to more severe forms of sepsis. To understand the temporal stability of the RNA and elastase biomarker signals, and whether temporal changes could be relevant to sepsis progression, a subset of patients had a second blood draw (n = 27). Blood samples were obtained at the time of enrollment (0 hr), and 3 hours later. As shown in Fig 4, the more severe sepsis patients showed a noticeable pattern of changes in the DEFA1 RNA level across the 3 hour period, with 4 patients showing abnormal scores with >25% change, up or down, in expression level. Likewise, ALPL and IL8RB/CXCR2 were quite variable in the more severe sepsis group with 3 additional patients showing abnormal scores with >25% change. Thus, 7/11 more severe sepsis patients had abnormal and changing scores. The sepsis group and non-infected group each had only 1 DEFA1 and 1 ALPL and IL8RB/CXCR2 score that was abnormal and changing (chi-square p < 0.05 more severe sepsis vs sepsis+uninfected). These results were similar in the second analysis, where 83% (10/12) of patients with a qSOFA score ≥ 2 had a > 25% change in their DEFA1 or ALPL and IL8RB/CXCR2 scores compared to 40% (6/15) of patients with qSOFA scores = 0 or 1 (chi-square p < 0.05 qSOFA ≥ 2 vs qSOFA = 1 and 0). Temporal changes in host immune biomarkers associated with more severe forms of sepsis. A subset of patients were enrolled in which 0 hr and 3 hr samples were obtained. Each line reflects the same patient at two time points for bacterial response markers: DEFA1 RNA, (left panels), and ALPL+IL8RB/CXCR2 (right panels). The panels show patients judged to have more severe forms of sepsis (Sepsis/Shock, n = 11), Sepsis (n = 8), and No infection (n = 8). Defensin protein in plasma vs DEFA1 RNA. Suspecting that rapidly decreasing DEFA1 RNA could be indicative of active translation leading to translational decay of the RNA, we analyzed the protein levels. In a subset of 15 patients and 6 controls, 29 samples (some at 0 and 3 hr) had both DEFA1 RNA by ddPCR and defensin-α1 protein in plasma by ELISA. There was a broad range of absolute values (DEFA1 RNA 0.69−87% of ACTB) and defensin-α1 protein in plasma (44−451 ng/ml), but there was only a modestly negative correlation between their levels (R = −0.27) suggesting that lower DEFA1 RNA does not directly translate to higher defensin protein levels in plasma (Fig 5). A possible explanation is that defensin-α1 protein could remain cell-associated, and especially tightly incorporated into peri-neutrophil extracellular traps (NETs), as has been observed for related proteins such as neutrophil elastase and histone antimicrobial peptides [27]. DEFA1 RNA versus defensin-α1 protein levels. In a subset of patients, it was possible to determine both the DEFA1 RNA levels by ddPCR (Y axis) and defensin-α1 protein levels by ELISA (X axis). In 29 samples, from 15 patients and 6 controls, the best linear fit is shown along with the Pearson’s R correlation. Blood burden of bacteria versus RNA biomarkers and sepsis. The blood levels of bacterial and fungal pathogens was quantitated by DNA purification from the Tempus nucleic acid preserved blood, followed by ddPCR using primers and probes optimized to minimize cross-over reactivity with human sequences. Separate primer/probe pairs detected Gram (+), Gram (-) 16S ribosomal bacterial sequences, and 28S fungal sequences. Blood DNA was available and successfully purified and amplified on a subset of 59 of 72 cases (all 0 hr). As shown in Fig 6 (Left panel), the pathogen DNA burden in blood showed step-wise increases from no infection, sepsis, though severe sepsis/septic shock, with Gram (-) and total reads significantly elevated above No Infection cases. Blood pathogen levels by ribosomal ddPCR versus clinical infection grouping. Blood samples in RNA preservative were extracted for DNA, quantitated, and 200 ng was assayed by ddPCR for bacterial 16S specific for Gram (+) or Gram (-), or 28S ribosomal DNA for fungus. The read counts per pathogen type is averaged across the specific groups, with error bars reflecting s.e.m. In the left panel, the patients are grouped by the adjudicated presence of No Infection (n = 20), Sepsis (n = 20), or more severe Sepsis/Septic shock (n = 19). In the right panel, the patients are grouped by those with positive blood cultures (n = 7), versus negative blood cultures (n = 43) or no blood culture was performed (n = 10). * indicates p < 0.05 for comparison in brackets. The patients were re-grouped according to the results of blood culture status, as shown in Fig 6 (Right panel). As expected, the 16S reads were highest in patients with positive blood cultures (n = 7), relative to culture-negative patients (p = ns, n = 43) or patients from whom blood cultures were not drawn (p < 0.05, n = 10). Notably, the blood culture-negative patients still had Gram (-) 16S counts more than 2-fold higher than patients not cultured, and not significantly different from blood culture-positive patients. Relationship between blood pathogen levels and host RNA biomarkers. It was then determined whether the levels of bacterial reads in blood are related to the levels of the neutrophil activation RNA biomarkers. This was evaluated in two ways. First, the correlation between microbial pathogen DNA and the neutrophil RNA biomarkers was examined. As shown in Fig 7 (Left Panel), there is a modestly positive correlation overall of R = 0.35 between the 2 parameters. Microbial pathogen levels in blood versus host RNA biomarkers. Whole blood in a nucleic acid preservative was extracted for both RNA and DNA. The RNA was quantitated by ddPCR for the host immune RNA biomarkers, and the DNA was quantitated for Gram (+) and Gram (-) 16S ribosomal reads and fungal 28S ribosomal reads using ddPCR. In the Left Panel, each patient is plotted for total DNA levels of pathogens on the Y axis, and the total host biomarker RNA levels on the X axis. Symbols reflect the assignment of the patient to adjudicated diagnosis. In the Right Panel, the patients are stratified by the presence of High (>100 reads/20 µl), Medium (>50 but <100 reads/20 µl), or Low (<50 reads/20 µl) levels of Total Blood Pathogen reads (bacterial and fungal). Average microbial pathogens (±s.e.m) are shown in the top bars divided by 10 for graphic fit. Levels of elastase activity (Vmax), DEFA1%, ALPL+IL8RB%, WBC count (k/ul), and lactic acid (X 10) are shown as grouped by pathogen burden. * indicates p < 0.05. Specific cases, highlighted in Fig 7, demonstrate examples of cases in which the two parameters are highly correlated or highly disconnected. One more severe sepsis case in particular (S37), demonstrates highly elevated levels of both bacterial Gram (+) and Gram (-) DNA, as well as host RNA biomarkers (DEFA% = 100%, ALPL and IL8RB/CXCR2 = 19%). However, two other cases readily demonstrate the potential disconnects: a more severe urosepsis case with obstructive stones (S57), demonstrated even higher Gram (+) and similar Gram (-) levels, but had markedly lower bacterial RNA biomarker levels (DEFA = 4%) and equivalent ALPL and IL8RB/CXCR2 levels (ALPL + IL8RB/CXCR2 = 19%) that were just above the abnormal threshold of 18%. Conversely, a non-severe sepsis pneumonia case (S44), showed very high DEFA1 levels (153%), and elevated ALPL and IL8RB/CXCR2 levels (25%), but had strikingly lower bacterial 16S reads (136), that were however, still above levels observed on average in the No Infection cases (mean = 61). Clearly however, there is an abundance of severe sepsis/septic shock cases with microbial DNA levels at or below those of patients judged to be not infected. Examined differently, the general relationship between the bacterial load and RNA biomarkers is evident. As shown in Fig 7 (Right Panel), the blood bacterial DNA levels were used to divide the patients into Low (<50, n = 16), Medium (50–100, n = 20), and High burden (>100, n = 23) patients (note levels/10 in figure for graphic purposes). As expected, in the subset of cases where both the bacterial burden and blood cultures were obtained (n = 50), there was a stepwise increase in positive blood cultures from 0/10 in the low burden group to 5/23 in the high burden group. While encouraging, it highlights that of the 23 patients with the highest 16S burden in blood, only 5 showed positive blood cultures. The neutrophil elastase activity also tracked the pathogen load, but did not reach statistical significance. Further, the DEFA1% is significantly higher in the High (30.4%) vs Low (7.8%) burden group, with the Medium burden showing an intermediate level (18.2%). The ALPL and IL8RB/CXCR2 scores trended similarly, but were not statistically significant. The WBC count and lactic acid levels showed no notable variation across the bacterial burden groups. Discussion Due to its rapid and often fatal course, sepsis remains a clinical challenge and active research topic. Contrary to our initial literature-based hypothesis, infected patients that developed more severe sepsis did not show levels of neutrophil activation-related RNA biomarkers significantly different from subjects that did not develop more severe forms of sepsis. Also contrary to our expectations, we observed rapid change in these RNA biomarkers between zero and three hours among subjects that went on to develop more severe sepsis. One possible explanation for increases is that this change in measurable RNA coincided with activation of the neutrophil NETosis program [28]. Declining neutrophil activation RNA levels in some subjects may be associated with more advanced stage NETosis in which neutrophil transcription is outweighed by translational decay of RNAs. There are two types of human defensins: α-defensins produced mostly by neutrophils and some cells in the gut; β-defensins are produced by various types of epithelial cells and leukocytes [29,30]. Both types defend against microbial infections including bacterial, fungal, viral and protozoal [31]. β-defensins have previously been reported to have reduced inducibility in leukocytes from severely septic patients [32], and β-defensin expression in whole blood from ICU patients with sepsis was found to be down-regulated [39]. Because of their direct mechanistic involvement in neutrophil function and sepsis progression, considerable work has focused on neutrophil α-defensins as biomarkers. Defensins are critical to fighting infections, but like many aspects of innate immunity can also have deleterious effects on endothelial tissues (reviewed in [33]). Studies have reported higher sepsis mortality among mice with higher copy numbers of alpha-defensin genes, and among mice treated with human alpha defensin in a mouse sepsis model [45]. Plasma levels of α-defensin proteins have been reported to be elevated in severe sepsis, and positively correlated with the SOFA scores [34]. Elevated circulating levels of defensin-α1 protein were significantly higher in COVID19 patients versus controls and associated with adverse outcomes [35]. Analysis of hemadsorption filters from septic patients suggests that defensins are major factors that could contribute to secondary endothelial damage that accelerates the syndrome [36,37]. We did not observe similar correlation between outcomes and raw defensin-α1 protein levels in our study cohort; however, changes in the levels of DEFA1 RNA over a 3-hour period was associated with clinical decompensation and emergence of more severe sepsis/septic shock. This observation in ED patients between 0 and 3 hours may be meaningful in an immunologic sense, and useful diagnostically. Prior time series analysis of septic patients identified three genes in the elastase family (ELANE, CTSG, PRTN3) and DEFA4 as candidate biomarkers for sepsis [38], while other studies likewise implicated NETosis genes including ALPL [39]. IL8, via IL8RB/CXCR2, can trigger NETosis of neutrophils [40]. We suspect that the DEFA1/3 locus would have also been identified in these studies, but it is prone to filtering due to its multi-copy number variants. A similar approach to the present study was used to build a seven gene expression model, including DEFA4, that predicted sepsis severity with reasonable accuracy (ROC = 0.88) [41]. The present results suggest that prior biomarker studies of sepsis would have been complicated by the quite variable time frames of sampling, whereby even a 3-hour difference in sampling could change the RNA expression level by 50%. Thus, we propose a key factor in understanding the disparate results with defensin RNA levels in sepsis is the timing of blood sample in relation to the course of the disease. The present samples were captured in the ED as close as is practically possible to the time of patient presentation, while most studies have used patients with established sepsis that may have been ongoing for days. A second major factor is the difference between the RNA levels and protein levels, which is emphasized by the observation that RNA and protein levels for defensin-α1 had a slightly negative correlation. We propose that the negative trajectory of DEFA1 RNA is part of the key pathological process of NETosis, whereby the defensins, and other proteins, are translated and secreted into a physical NET that adheres either to other circulating cells or to endothelium. The present studies are limited by a modest sample size in the trajectory analysis where only 27 patients were available with multiple time points. The enrollment criteria for sepsis was somewhat modified by the ‘sepsis alert’ used in the Cerner EHR, and post-hoc adjudication is subjective, although this was minimized by engaging multiple judges. Conclusions A change in RNA markers of neutrophil activation among septic ED patients between 0 and 3 hours is associated with progression to more severe forms of sepsis. A change in RNA markers of neutrophil activation among septic ED patients between 0 and 3 hours is associated with progression to more severe forms of sepsis. Patients that developed more severe forms of sepsis did not have higher neutrophil activation RNA levels than those with sepsis. Patients that developed more severe forms of sepsis did not have higher neutrophil activation RNA levels than those with sepsis. Patients with more severe sepsis had similar burdens of bacteria in blood compared to those with sepsis. Patients with more severe sepsis had similar burdens of bacteria in blood compared to those with sepsis. Supporting information Full demographics and lab values of patients and controls. (DOCX) Patient-level raw data for gene expression, diagnosis, and lab values. (XLSX) Temporal changes in RNA biomarkers grouped by Sepsis-3 Criteria. (DOCX) Transfer Alert This paper was transferred from another journal. As a result, its full editorial history (including decision letters, peer reviews and author responses) may not be present. 25 Feb 2025 25 Jun 2025 Dear Dr. McCaffrey, 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. Please submit your revised manuscript by Aug 09 2025 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org . When you're 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. A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter. If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols . Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols . We look forward to receiving your revised manuscript. Kind regards, Claudine Irles, 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 https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf 2. Thank you for stating the following financial disclosure: “The Ulvi and Reykhan Kasimov Family The St. Laurent Institute NIH S10 OD021622” Please state what role the funders took in the study. If the funders had no role, please state: "The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript." If this statement is not correct you must amend it as needed. Please include this amended Role of Funder statement in your cover letter; we will change the online submission form on your behalf. 3. Thank you for stating the following in the Acknowledgments Section of your manuscript: “The authors are very grateful to the patients and their families who kindly agreed to participate in this research study despite their difficult health situations. The authors are grateful to The Ulvi and Reykhan Kasimov Family and The St. Laurent Institute for generous financial support that made these studies possible. Other support was obtained from the NIH S10 OD021622 to TM.” We note that you have provided funding information that is currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form. Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows: “The Ulvi and Reykhan Kasimov Family The St. Laurent Institute NIH S10 OD021622” Please include your amended statements within your cover letter; we will change the online submission form on your behalf. 4. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please move it to the Methods section and delete it from any other section. Please ensure that your ethics statement is included in your manuscript, as the ethics statement entered into the online submission form will not be published alongside your manuscript. 5. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information. [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? Reviewer #1: Yes ********** 2. Has the statistical analysis been performed appropriately and rigorously? -->?> Reviewer #1: Yes ********** 3. Have the authors made all data underlying the findings in their manuscript fully available??> The PLOS Data policy Reviewer #1: Yes ********** 4. Is the manuscript presented in an intelligible fashion and written in standard English??> Reviewer #1: Yes ********** Reviewer #1: Introduction: The authors appropriately delineate the limitations of the prior definitions of sepsis that led to the current definition often referred to as ‘Sepsis-3.’ As noted, a foundational flaw in the prior definitions of sepsis is their requirement for SIRS criteria, which are neither sensitive nor specific. Further, while it is true that the SOFA score was developed to quantify and characterize critical illness, it is also, to various degrees, included in the more contemporary definition of sepsis. In any case, this discussion of the various definitions of sepsis – and their limitations – becomes important because, in the results and discussion, it appears that the authors have reverted to the prior definitions. The discussion regarding the treatment of sepsis, while true, is largely unnecessary and impertinent to the current discussion, which is one of diagnosis (not treatment). The results are ‘teased’ in the introduction, which is somewhat unusual and should be reserved for the results and discussion sections. Methods The authors point to the criteria used in the sepsis alert; however, these criteria are not clear. Please include the referenced figure (i.e., are these the critera noted as ‘enrollment criteria’ in figure 1? In using these criteria, it should be noted that these are not SIRS criteria, but they are a modified version of SIRS. For example, the Temp > 38.3 is higher than traditional SIRS criteria, as are both the Respiratory and Heart Rates. In this way, this modification of SIRS criteria will be less sensitive than previously noted; it is unclear the impact that the glucose will have. In essence, this will result in an included population that is somewhat different than any current definition of sepsis, and may impact the generalizability of the results. How was the subset of patients and controls selected for measurement of the defensin-1 level selected? Was this selection random, or was it based on specific criteria, or perhaps through more practical reasons (availability of sample for testing). With regards to controls … in many cases, understanding the control group becomes one of the most integral components to interpreting a study. There is mention that a normal control group was derived from medical center personnel, and, presumably, another group of subjects with self-reported infections was enrolled. How were these subjects selected? Who are they? A description is important because a 25 year-old medical resident is vastly different than a 65 year old ED patient; the mechanism of selection of the control group is important to disclose. Results Overall, the number of groups included in the study is somewhat difficult to follow (there are 5 – control, self-reported infection, no sepsis, sepsis, septic shock). There may be a way to simplify this, and the N’s of the control groups are small enough that it is unclear that they provide any meaningful information. The first paragraph of the results is the first mention of adjudication for sepsis and sepsis shock. Because there is no reference standard for sepsis, it is necessary to have such a process. This does raise several questions: how was the adjudication performed? Why did the authors chose to use the now antiquated terminology of ‘sepsis / severe sepsis / septic shock,’ which were abandoned in the current Sepsis-3 definition referenced in the Introduction. Also, as previously noted, the vital sign criteria are modified from the original SIRS criteria such that these patients may or may not meet any particular definition of sepsis. An explanation for these criteria and a summary of the adjudication process should be included in the Methods. “Average lactate levels were artificially high …” This statement is incorrect; there is nothing ‘artificial’ about the lactic acidosis associated with such physiologic derangements as alcohol withdrawal and seizure. Table 1: This table needs a legend. Are these means, medians? It should also be reformatted to read easier. Blood culture (+) – is this percent? Presumably so – but is this percent of cultures drawn or percent of patients? We could presume that all patients had blood cultures drawn, but this presumption may not be accurate. Where appropriate, units should be included. I note that, in race, the %white is on average, about 33%. Since this represents a minority of the study population, it would be helpful to have more information on the ethnic and racial distribution. Also, the number of patients in each group is important, given than many of the patients are unlikely to have had specific values tested (CRP, lactic acid, etc), given the data were derived from standard care. The control group is not included in any description. They should be included in Table 1, if not elsewhere. It is possible that they are in an original format, but, for the PDF available for review, I do not see a column for controls. Again, a more detailed explanation or description of the adjudication process would be appropriate. It is difficult to interpret inter-observer variability for infection and sepsis without understanding the process by which the determination was made, especially since this definition of sepsis is not consistent with any of the prior definitions that have been published. The term “simple sepsis” is used. Colloquially, this is often the term we use for ‘SIRS with a source.’ This is not a technical term, and it should be defined or, better yet, abandoned. There is some mention of ‘non-ED controls with self-reported infections’ in the methods; however, there is no detail in how these controls were selected. More clarity is necessary as mentioned in the above comments about the methods. For the investigation in to termporal change – did this include all individuals? This should be included in the Methods; the first mention of this should not be in Results. In examining Figure 4, it appears that the serial measurement was only performed in a subset of patients, and this is confirmed in the legend, so a description of why and how this occurred is helpful. The statement “suggesting that blood culture misses a significant percentage of bacteremic cases” is interpretive and should be moved to the Discussion. In Figure 7, specific cases are highlighted. How and why were these cases chosen? What is different about them as compared to the ones that were not included? Some mechanism to disclose this information would be appreciated. That said, there are considerable limitations given the quite restrictive inclusion criteria that were used. I would certainly welcome the opportunity to review this manuscript again after revision, because I do think, in light of the current advancement in diagnostics – including expression assays – such knowledge and understanding is helpful. In summary, though I find the concept and results intriguing and certainly worthy of publication, in light of the contemporary definition of sepsis, I find the results, as presented, difficult to interpret. Perhaps the results would be different if it were presented as ‘no infection,’ ‘infection,’ and ‘sepsis’ by the Sepsis-3 definitions. Unfortunately, it is difficult to assess given the small sample size and heterogeneity seen in the study population vs the control group, and without great understanding of the adjudication process. In any case, the findings suggest that temporal changes in gene expression are important; thus, reformatting of the manuscript may make these results more easily interpreted in the context of our current definitions. Whether or not this is clinically meaningful is in question, but this small (though rigorous) study is meant to be hypothesis generating. Thoughout the manuscript, then N’s are very different in each group. For example, in Figure 6, the N’s are 20 / 20 / 19 for No Infection / Severe Sepsis / Septic Shock while in Table 1, the N’s are 24 / 26 / 22. An explanation or description of the missing patients would be helpful. ********** 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 ********** [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.] 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 PLOS at figures@plos.org . Please note that Supporting Information files do not need this step. 1 Aug 2025 Response to Critique: LaFleur et al. Blood RNA biomarkers and a point-of-care elastase assay for detecting host immune activation in suspected sepsis: Trajectory matters. Reviewer #1: We are very grateful for your thoughtful review, which is constructive, detailed, and accurate. We believe that all of the concerns are addressed to the best of our abilities, and that the manuscript is improved because of the changes. We include a ‘track changes’ version so that it is easy to identify the extensive revisions, including new data analysis, that were made to address these concerns. Introduction: Concern 1: The discussion regarding the treatment of sepsis, while true, is largely unnecessary and impertinent to the current discussion, which is one of diagnosis (not treatment). Response: We have shortened this section to discuss only how diagnostic uncertainty drives needless treatments. Concern 2: The results are ‘teased’ in the introduction, which is somewhat unusual and should be reserved for the results and discussion sections. Response: We have also shortened this section to simply frame the question, without mentioning the outcomes. Methods: Concern 3: The authors point to the criteria used in the sepsis alert; however, these criteria are not clear. Please include the referenced figure (i.e., are these the criteria noted as ‘enrollment criteria’ in figure 1? In using these criteria, it should be noted that these are not SIRS criteria, but they are a modified version of SIRS. For example, the Temp > 38.3 is higher than traditional SIRS criteria, as are both the Respiratory and Heart Rates. In this way, this modification of SIRS criteria will be less sensitive than previously noted; it is unclear the impact that the glucose will have. In essence, this will result in an included population that is somewhat different than any current definition of sepsis, and may impact the generalizability of the results. Response: These are valid points. Yes, the values in Figure 1 are the criteria used for the Cerner ‘sepsis alert’ that identified subjects for screening and consent. The Cerner alert is based on the St. John’s sepsis alert, which is now cited for clarity and fully detailed in Methods. While it is slightly different than the SIRS criteria, it is a standardized system that many ED physicians are familiar with. We corrected all ‘SIRS criteria’ to ‘sepsis alert’, which is clearly defined as a modified SIRS criteria in the methods. Further, toward a similar point discussed later, we went back to the medical records and computed qSOFA scores used in the SEPSIS-3 criteria (see below). Concern 4: How was the subset of patients and controls selected for measurement of the defensin-1 level selected? Was this selection random, or was it based on specific criteria, or perhaps through more practical reasons (availability of sample for testing). Response: Yes, the selection of samples for defensin-a1 protein level was made purely on the basis of having plasma available for testing. This is now noted in Methods. Concern 5: With regards to controls … in many cases, understanding the control group becomes one of the most integral components to interpreting a study. There is mention that a normal control group was derived from medical center personnel, and, presumably, another group of subjects with self-reported infections was enrolled. How were these subjects selected? Who are they? A description is important because a 25 year-old medical resident is vastly different than a 65 year old ED patient; the mechanism of selection of the control group is important to disclose. The control group is not included in any description. They should be included in Table 1, if not elsewhere. It is possible that they are in an original format, but, for the PDF available for review, I do not see a column for controls. There is some mention of ‘non-ED controls with self-reported infections’ in the methods; however, there is no detail in how these controls were selected. More clarity is necessary as mentioned in the above comments about the methods. Response: The control group, including those with self-reported infections, are now more fully described in the text. We prepared summary statistics of age, sex, and race, and this is included as Supplementary Table 1. It would not be practical to include the controls in Table 1 because there are no lab values available on these non-patient controls. However, we have repeated the other groups in the supplementary table so that they can easily be compared. In short, the controls are younger and more likely to be White, but their main purpose is simply to provide a reference for non-ED patients, with or without self-reported infection, and they fulfill that need. We extensively revised Table 1 to include full titles for the variables, with additional details in the legend. Results: Concern 6: Overall, the number of groups included in the study is somewhat difficult to follow (there are 5 – control, self-reported infection, no sepsis, sepsis, septic shock). There may be a way to simplify this, and the N’s of the control groups are small enough that it is unclear that they provide any meaningful information. Response: There are 16 controls without any self-reported infection, and this adds a valuable frame of reference for the 8 sub-clinical infections, and the total of 72 ED patients. The n is sufficient to detect statistically significant differences in the RNA biomarker levels (Figure 2). Based on the new Sepsis-3 analysis, we provide new groupings that are simpler, and more amenable to the terminology moving forward. Concern 7: The first paragraph of the results is the first mention of adjudication for sepsis and sepsis shock. Because there is no reference standard for sepsis, it is necessary to have such a process. This does raise several questions: how was the adjudication performed? Why did the authors choose to use the now antiquated terminology of ‘sepsis / severe sepsis / septic shock,’ which were abandoned in the current Sepsis-3 definition referenced in the Introduction. Also, as previously noted, the vital sign criteria are modified from the original SIRS criteria such that these patients may or may not meet any particular definition of sepsis. An explanation for these criteria and a summary of the adjudication process should be included in the Methods. […] Again, a more detailed explanation or description of the adjudication process would be appropriate. It is difficult to interpret inter-observer variability for infection and sepsis without understanding the process by which the determination was made, especially since this definition of sepsis is not consistent with any of the prior definitions that have been published. Response: True, we used the prior Sepsis-2 terms because those are still terms relevant to billing in many US centers, and are still very familiar to most practicing physicians. We agree that Sepsis-3 terms will be more relevant moving forward, and so we reanalyzed the medical records to build the Sepsis-3 qSOFA scores and we present similar results using this new analysis. The adjudication process is now explained in greater detail, and it is supplemented by the new analysis using the Sepsis-3 criteria to group the patients. We think that with the clarification of the Cerner Sepsis alert, the clarified adjudication process, and the new qSOFA analysis that readers will be able to understand the patient groups adequately. We show the results of the qSOFA analysis as new panels in Figure 2, and as a new Supplementary Figure 1 for the trajectories. Further, to clarify the group names in this analysis, we changed “Severe Sepsis” to “Sepsis/Shock” to indicate that this group is more severe sepsis and includes patients with organ failure signs and vasopressors in some cases. Concern 8: “Average lactate levels were artificially high …” This statement is incorrect; there is nothing ‘artificial’ about the lactic acidosis associated with such physiologic derangements as alcohol withdrawal and seizure. Response: Apologies, we meant that the elevated lactate was not related to infection. We have corrected the wording to “higher than expected”. Concern 9: Table 1: This table needs a legend. Are these means, medians? It should also be reformatted to read easier. Blood culture (+) – is this percent? Presumably so – but is this percent of cultures drawn or percent of patients? We could presume that all patients had blood cultures drawn, but this presumption may not be accurate. Where appropriate, units should be included. I note that, in race, the %white is on average, about 33%. Since this represents a minority of the study population, it would be helpful to have more information on the ethnic and racial distribution. Also, the number of patients in each group is important, given than many of the patients are unlikely to have had specific values tested (CRP, lactic acid, etc), given the data were derived from standard care. Response: Table 1 has been completely revised with the variables spelled out and a legend that clearly denotes that they are means, sems, and how the statistics were conducted. Concern 10: The term “simple sepsis” is used. Colloquially, this is often the term we use for ‘SIRS with a source.’ This is not a technical term, and it should be defined or, better yet, abandoned. Response: We removed all instances of ‘simple sepsis’. Concern 11: For the investigation into temporal change – did this include all individuals? This should be included in the Methods; the first mention of this should not be in Results. In examining Figure 4, it appears that the serial measurement was only performed in a subset of patients, and this is confirmed in the legend, so a description of why and how this occurred is helpful. Response: The 0 and 3 hour measures were only available on a subset of patients (n=27), largely because the patient declined the second draw or was otherwise unavailable due to transfer or discharge. We have clarified this subset as patients in which their clinical course and consent allowed a 2nd draw. Concern 12: The statement “suggesting that blood culture misses a significant percentage of bacteremic cases” is interpretive and should be moved to the Discussion. Response: Agreed. This is now deleted. Concern 13: In Figure 7, specific cases are highlighted. How and why were these cases chosen? What is different about them as compared to the ones that were not included? Some mechanism to disclose this information would be appreciated. Response: Agreed. We noted that these examples are chosen because they represent ‘non-synonymous’ cases of pathogen load and RNA biomarker levels. We explain more clearly that they are cases where pathogen levels and the RNA biomarkers are “highly correlated or highly disconnected”. Concern 14: Throughout the manuscript, then N’s are very different in each group. For example, in Figure 6, the N’s are 20 / 20 / 19 for No Infection / Severe Sepsis / Septic Shock while in Table 1, the N’s are 24 / 26 / 22. An explanation or description of the missing patients would be helpful. Response: True. The changing N’s are due to the specific criteria involved in each of the figures and reflect our best efforts to be accurate about the size of the groups in each comparison. In Figure 6, for example, we did not always have excess blood for the isolation of whole blood DNA, and thus the N is smaller. We have added text to clarify this. Final Issues: That said, there are considerable limitations given the quite restrictive inclusion criteria that were used. I would certainly welcome the opportunity to review this manuscript again after revision, because I do think, in light of the current advancement in diagnostics – including expression assays – such knowledge and understanding is helpful. In summary, though I find the concept and results intriguing and certainly worthy of publication, in light of the contemporary definition of sepsis, I find the results, as presented, difficult to interpret. Perhaps the results would be different if it were presented as ‘no infection,’ ‘infection,’ and ‘sepsis’ by the Sepsis-3 definitions. Unfortunately, it is difficult to assess given the small sample size and heterogeneity seen in the study population vs the control group, and without great understanding of the adjudication process. In any case, the findings suggest that temporal changes in gene expression are important; thus, reformatting of the manuscript may make these results more easily interpreted in the context of our current definitions. Whether or not this is clinically meaningful is in question, but this small (though rigorous) study is meant to be hypothesis generating. Submitted filename: 0 SENSOR Response to Reviewers.docx 6 Nov 2025 Dear Dr. McCaffrey, 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. Please submit your revised manuscript by Dec 21 2025 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org . When you're 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. A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols . Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols . We look forward to receiving your revised manuscript. Kind regards, Tomasz W. Kaminski Academic Editor PLOS ONE Journal Requirements: If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise. Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice. Additional Editor Comments: Dear Authors, Thank you for your thorough and thoughtful revision. The revised version of your manuscript shows substantial improvement and clearly addresses the major reviewer concerns from the previous round. I appreciate the extensive effort invested in refining the analyses and clarifying the methods and results. At this stage, I have only a few minor points that should be addressed before the manuscript can proceed to acceptance. These are primarily editorial and clarificatory in nature: -State statistical approach in more detail. Briefly describe normality testing, justify the use of parametric or non-parametric tests, -Ensure consistent figure and table formatting. Verify that figure numbering matches the text; add missing units to all axes and continuous variables in tables. - Add or confirm data availability information. Please provide a DOI or repository link for the underlying data, as required by journal policy. -Clarify the “No Sepsis” group definition. Please explicitly define this cohort as sepsis-alert positive but non-infectious inflammatory controls to avoid confusion regarding elevated CRP and inflammatory markers. - Expand the limitations paragraph slightly. Explicitly acknowledge the small trajectory subset, modified enrollment criteria, and adjudication variability. - Proofread for minor typographical and formatting inconsistencies. (e.g., spacing, capitalization in figure labels, consistent terminology such as “flu” vs “influenza.”) All the best. [Note: HTML markup is below. Please do not edit.] Reviewers' comments: Reviewer's Responses to Questions Comments to the Author Reviewer #1: All comments have been addressed ********** 2. Is the manuscript technically sound, and do the data support the conclusions??> 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 Reviewer #1: Yes ********** 5. Is the manuscript presented in an intelligible fashion and written in standard English??> Reviewer #1: Yes ********** Reviewer #1: My concerns have been adequately addressed. I appreciate the author's attention and effort. My only comment is that some of the tables need appropriate formating, but this may be in the rendering of the document. ********** 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 ********** [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 ensure your figures meet our technical requirements, please review our figure guidelines: https://journals.plos.org/plosone/s/figures You may also use PLOS’s free figure tool, NAAS, to help you prepare publication quality figures: https://journals.plos.org/plosone/s/figures#loc-tools-for-figure-preparation. NAAS will assess whether your figures meet our technical requirements by comparing each figure against our figure specifications. 13 Nov 2025 Response to Editorial Concerns: (LaFleur et al. PONE-D-25-10033R1) Overall: At this stage, I have only a few minor points that should be addressed before the manuscript can proceed to acceptance. These are primarily editorial and clarificatory in nature: Concern: State statistical approach in more detail. Briefly describe normality testing, justify the use of parametric or non-parametric tests. Response: A brief “Statistical Methods” paragraph was added to address this. Concern: Ensure consistent figure and table formatting. Verify that figure numbering matches the text; add missing units to all axes and continuous variables in tables. Response: We verified that figure numbering matches the text. We added several units that were missing in the figures, especially Figure 4. We could not detect any missing units in the only table, Table 1. Concern: Add or confirm data availability information. Please provide a DOI or repository link for the underlying data, as required by journal policy. Response: Due to the US government shutdown, it has been impossible to get an accession at Gene Expression Omnibus (GEO), and it may take months to resolve a 40 day backlog. Because there are only 6 gene expression levels per patient, we have included the patient level raw data, in anonymous form, in Supplementary Table 2. Concern: Clarify the “No Sepsis” group definition. Please explicitly define this cohort as sepsis-alert positive but non-infectious inflammatory controls to avoid confusion regarding elevated CRP and inflammatory markers. Response: Correct. We inserted a sentence to this effect in the methods. Concern: Expand the limitations paragraph slightly. Explicitly acknowledge the small trajectory subset, modified enrollment criteria, and adjudication variability. Response: True, these limitations were added to the end of the discussion. Concern: Proofread for minor typographical and formatting inconsistencies. (e.g., spacing, capitalization in figure labels, consistent terminology such as “flu” vs “influenza.”) Response: We corrected figure labels for consistent capitalization, and made several corrections for consistent terminology. Submitted filename: SENSOR Response #2.docx 18 Nov 2025 Blood RNA biomarkers and a point-of-care elastase assay for detecting host immune activation in suspected sepsis: Trajectory matters. PONE-D-25-10033R2 Dear Dr. McCaffrey, We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements. Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication. An invoice will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager® and clicking the ‘Update My Information' link at the top of the page. For questions related to billing, please contact billing support . If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- 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. Kind regards, Tomasz W. Kaminski Academic Editor PLOS ONE PONE-D-25-10033R2 PLOS ONE Dear Dr. McCaffrey, I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now being handed over to our production team. At this stage, our production department will prepare your paper for publication. This includes ensuring the following: * All references, tables, and figures are properly cited * All relevant supporting information is included in the manuscript submission, * There are no issues that prevent the paper from being properly typeset You will receive further instructions from the production team, including instructions on how to review your proof when it is ready. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few days to review your paper and let you know the next and final steps. Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, 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. You will receive an invoice from PLOS for your publication fee after your manuscript has reached the completed accept phase. If you receive an email requesting payment before acceptance or for any other service, this may be a phishing scheme. Learn how to identify phishing emails and protect your accounts at https://explore.plos.org/phishing. If we can help with anything else, please email us at customercare@plos.org. Thank you for submitting your work to PLOS ONE and supporting open access. Kind regards, PLOS ONE Editorial Office Staff on behalf of Dr. Tomasz W. Kaminski Academic Editor PLOS ONE