Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

# Identification of novel blood biomarkers of treatment response in cystic fibrosis pulmonary exacerbations by label-free quantitative proteomics

## Abstract

Pulmonary exacerbations (PEx) are clinically impactful events for individuals with CF. Unfortunately, many CF individuals with PEx fail to regain their baseline lung function despite treatment. The objective of this study was to use unbiased proteomic technology to identify novel blood protein biomarkers that change following intravenous (IV) antibiotic treatment and to explore if changes correlate with clinical response by the end of treatment. Blood samples from 25 PEx events derived from 22 unique CF adults were collected within 24 hours of hospital admission, day 5, day 10, and IV antibiotic completion. Three-hundred and forty-six blood proteins were evaluated with label-free liquid chromatography-tandem mass spectrometry (LC-MS/MS) quantitative proteomics and immunoassays. Forty-seven plasma proteins changed significantly following 5 days of IV antibiotic treatment (q-value ≤ 0.10). Early change in IGF2R from hospital admission to day 5 correlated with overall change in symptom score (CFRSD-CRISS) by the end of treatment (r = −0.48, p-value = 0.04). Several plasma proteins identified and quantified by label-free LC-MS/MS changed early following treatment with IV antibiotics and many of these proteins are involved in complement activation and inflammatory/immune-related pathways. Early change in IGF2R correlated with symptom response following IV antibiotic treatment and requires further validation as a predictive biomarker of symptom response.

## Introduction

Cystic fibrosis (CF) is a life-limiting autosomal recessive disease affecting over 70,000 people worldwide1. Individuals with CF experience recurrent pulmonary exacerbations (PEx) that are characterized by intermittent worsening in respiratory signs and symptoms2,3,4 and are associated with increased morbidity and irreversible loss in lung function5,6. CF PEx are commonly triggered by respiratory viruses, clonal shifts of colonizing bacteria, and sometimes non-infectious causes (air pollution, medication non-adherence) and are typically treated with antibiotics and increased airway clearance therapies (i.e. hypertonic saline and dornase alfa)2,4. Based on data from the 2017 Cystic Fibrosis Foundation Patient Registry, over 40% of CF adults were diagnosed with at least one PEx that required treatment with intravenous (IV) antibiotics7. Unfortunately, many individuals with CF fail to regain their baseline lung function despite IV antibiotic treatment8.

Suboptimal PEx outcomes may be due to delayed recognition and treatment, widely varied treatment decisions, and differences in the approach to monitoring treatment response, including recovery in lung function and/or resolution in signs and symptoms9,10. A novel adjunctive strategy that can provide an additional dimension to the objective monitoring of PEx treatment response is desired as it has the potential to improve clinical outcomes for CF individuals.

Biomarkers that are measured objectively and reproducibly have been studied in order to help guide therapeutic interventions in diseases such as COPD11,12,13. In CF, blood-based biomarkers that reflect systemic inflammation, such as C-reactive protein and calprotectin, decrease significantly following PEx treatment14,15,16,17,18. However, for a biomarker to aid in clinical decision making, an early change is potentially more informative in assisting treatment decisions as it provides the opportunity for CF physicians to modify treatments earlier than if they had waited for the patient to respond or not clinically, which can take longer in some patients. Based on the results of a prior study performed by our group, admission but not early change in CRP was found to be useful in predicting treatment outcomes and therefore CRP does not appear to be a useful marker of early response to PEx treatment in CF19.

In this study, we recruited adult CF subjects who were diagnosed with a PEx and required hospitalization for IV antibiotic treatment. We prospectively collected blood, symptom diaries, and spirometry (e.g. FEV1) within 24 hours of admission, day 5, day 10, and IV antibiotic treatment completion. Blood samples from these CF subjects have previously been evaluated in a prior study using multiple-reaction monitoring mass spectrometry (MRM-MS), a targeted proteomics approach20. The objective of this study was to use label-free (untargeted) liquid chromatography-tandem mass spectrometry (LC-MS/MS) to identify novel blood protein biomarkers that are associated with early response to IV antibiotics (i.e. from hospital admission to day 5) and to determine if early changes correlate with clinical outcome by the end of IV antibiotic treatment, in terms of improvement in lung function and symptoms.

## Results

### Clinical characteristics at hospital admission for PEx

A total of 25 PEx events from 22 unique CF subjects were eligible for this study. Clinical characteristics of participating subjects at hospital admission (V1) are summarized in Table 1. In brief, mean baseline lung function was 63.4 (21.5) % predicted and 60% of CF subjects had moderate to severe airflow obstruction (FEV1 < 70% predicted). Mean FEV1% predicted and CF Respiratory Symptom Diary-Chronic Respiratory Infection Symptom Score (CFRSD-CRISS) at V1 were 53.2 (SD 20.7) % predicted and 50.5 (SD 7.5), respectively. The modified Fuchs exacerbation score ranged from 4 to 8 with a median score of 6. Fourteen PEx events (56%) were characterized by a > 10% relative drop in FEV1% predicted when compared to baseline. Over half of PEx events (13/25, 52%) were non-acute with the patient describing a change in symptoms for at least two weeks prior to hospitalization and 4 of the PEx events were characterized by the receipt of oral antibiotics prior to hospitalization.

### Clinical outcomes of PEx treatment

Clinical outcomes of PEx treatment are summarized in e-Table 1. For 25 PEx events, the median duration of IV antibiotic treatment was 14 days (ranged from 13 to 24 days). The FEV1% predicted increased from V1 to each of the subsequent time points (V2, V3, and V4) but overall changes were not significant (e-Table 1, Fig. 1a). Majority of the PEx events (n = 21, 84%) recovered to 90% of baseline lung function but fewer recovered to ≥100% of their baseline lung function (n = 9, 36%). Twenty out of 25 PEx events had symptom questionnaires completed and the mean CFRSD-CRISS decreased significantly from V1 to each of the following time points (V2, V3, V4; e-Table 1, Fig. 1b). Fifteen (60%) events were characterized by >11-point decrease, which has been defined as the minimum clinically important difference10,21.

### Correlation between protein levels at hospital admission (V1) with clinical and demographic factors

Serum IL-6 levels at hospital admission (V1) measured with immunoassay inversely correlated with baseline FEV1% predicted (r = −0.40, p-value = 0.046; e-Table 2). The levels of 16 plasma proteins measured with LC-MS/MS also significantly correlated with baseline FEV1% predicted at V1 (e-Table 2). Moreover, significant correlations were identified between age and many plasma protein abundances at V1 and the correlations varied by sex (e-Table 3).

### Longitudinal changes in candidate and LC-MS/MS blood proteins

Serum IL-6 levels significantly decreased from hospital admission (V1) to each of the following time points (V2, V3, V4; e-Table 1, Fig. 1c), whereas significant changes were not identified between subsequent time points, which is consistent with IL-6 changing early in response to IV antibiotic therapy. Serum calprotectin levels significantly decreased from hospital admission (V1) to treatment completion (V4) but not at earlier time points (V2, V3). Significant change in the levels of serum IL-8 and TNF-$$\alpha$$ were not identified between V1 and any of the subsequent time points.

Following adjustment for baseline lung function, sex, age, and the interaction between sex and age, 47 proteins changed significantly from V1 to V2 with a FDR cut-off q-value ≤ 0.10 (e-Table 4, Fig. 2a) but just 6 proteins changed significantly from V1 and V4 with a FDR cut-off q-value ≤ 0.10 (e-Table 5, Fig. 2b).

### Protein-protein interaction (PPI) network and pathway enrichment analysis by STRING algorithm

The online STRING database (version 11.0) was applied to identify the protein-protein interaction (PPI) network using differentially expressed (DE) proteins and their most enriched molecular pathways following IV antibiotic treatment. The PPI network was constructed by the STRING algorithm after analyzing 47 proteins that changed significantly (q-value ≤ 0.10) between V1 and V2 (Fig. 3) and the top 10 enriched GO biological process terms (Table 2) and Reactome pathways (Table 3) are presented. Twenty-two of the 47 DE proteins from V1 to V2 were involved in the following immune/inflammatory-related GO biological processes: regulation of complement activation, regulation of acute inflammatory response, regulation of inflammatory response, immune effector process, and complement activation/classical pathway (Table 2, e-Fig. 2). Based on the Reactome database, 24 of the 47 DE proteins were involved in 5 immune related pathways (Table 3, e-Fig. 3). Similar analyses were not applied to DE proteins between V1 and V4 since only 6 proteins were identified.

### Correlation between early change in blood protein levels from hospital admission (V1) to Day 5 (V2) with changes in clinical outcomes from hospital admission (V1) to IV antibiotic completion (V4)

Among 47 DE proteins identified and quantified with LC-MS/MS, early change (V1 to V2) in the levels of just one protein, insulin like growth factor 2 receptor (IGF2R), inversely correlated with overall change in CFRSD-CRISS from V1 to V4 (r = −0.48, p-value = 0.04; e-Table 6). No significant correlations were identified between early change in the levels of 47 DE proteins and relative change in FEV1% predicted from V1 to V4. Additionally, early change in the levels of candidate blood proteins did not correlate with overall change in either FEV1% predicted or CFRSD-CRISS from V1 to V4.

A post-hoc sensitivity analysis indicated that the correlation between early change in IGF2R with overall change in CFRSD-CRISS from V1 to V4 was generally consistent when we randomly selected one PEx from each of the three subjects who had repeat PEx (e-Table 6).

### Correlation between overall change in blood protein levels from hospital admission (V1) to IV antibiotic completion (V4) with changes in clinical outcomes from hospital admission (V1) to IV antibiotic completion (V4)

Six plasma proteins identified and measured with LC-MS/MS significantly changed from V1 to V4 but no significant correlations with relative change in FEV1% predicted or absolute change in CFRSD-CRISS from V1 to V4 were observed. Similarly, overall change in the levels of candidate blood proteins did not correlate with overall change in either FEV1% predicted or CFRSD-CRISS from V1 to V4.

## Discussion

This is the first study to apply untargeted LC-MS/MS quantitative proteomics to identify blood proteins that change in response to IV antibiotics during the treatment of CF PEx. In addition to confirming blood protein biomarkers previously reported to change following IV antibiotics, including serum IL-6, calprotectin and plasma C-reactive protein (CRP), several novel plasma proteins were also identified with LC-MS/MS16,22. Interestingly, more proteins exhibited changes early during the treatment course (i.e. by day 5 of treatment) and relatively fewer towards the end of the treatment. Based on pathway enrichment analysis many of these proteins are involved in complement activation and regulation of the inflammatory response/immunity.

Proteins that changed early and remained significant by the end of IV antibiotic treatment included serum IL-6 and plasma CRP. However, early and overall change in serum IL-6 and plasma CRP did not correlate with changes in clinical outcomes, including FEV1% predicted and CFRSD-CRISS scores, by the end of treatment. Calprotectin is a candidate marker that also changed significantly by the end of treatment but also did not correlate with changes in clinical outcomes. In contrast, early change (V1 to V2) in insulin like growth factor 2 receptor (IGF2R; also known as cation-independent mannose-6-phosphate receptor or CI-M6PR) levels significantly correlated with symptom response (i.e. CFRSD-CRISS) by the end of IV antibiotic therapy. IGF2R is a multi-functional binding protein capable of binding insulin growth factor 2 (IGF2), mannose-6-phosphate (M6P), and retinoic acid23. Depending on the ligand to which it binds, IGF2R is involved in modulating a number of biological pathways including cell migration, wound healing, angiogenesis, apoptosis and the response to viral infection. Although it has not been studied in the context of CF previously, it is also induced by inflammatory mediators and has been studied as a candidate marker for systemic inflammation in other patient populations such as HIV24.

Many of the proteins that were found to be downregulated by day 5 of treatment are involved in complement activation (i.e. complement proteins C1r, C1q, C4b, C4b-binding protein, C8, C9). The complement proteins are key components of the innate immune system which promote neutrophilic inflammation and defend against pathogens25. Unregulated or persistent complement activation triggers a destructive inflammatory cascade which may lead to lung tissue damage and cause progressive loss of lung function25,26. Elevated levels of pro-inflammatory complement proteins have been observed in the sputum of individuals with CF26. Despite the potential lung protective effects of downregulating the complement system early during PEx treatment, we did not observe a greater recovery of lung function in such individuals but this study was small and therefore this warrants further study.

This study has a number of important limitations. Only 25 PEx events were analyzed in this exploratory study and therefore this study could have been underpowered to examine relationships between biomarkers and treatment outcomes. Furthermore, there was minimal decrease observed in FEV1% predicted at the time of hospitalization in comparison to stable baseline as only half of the PEx events were characterized by a >10% relative drop in FEV1% predicted from baseline. As a result, the improvements in FEV1% predicted in response to treatment were relatively modest and may have limited the potential to identify significant correlations with changes in blood proteins levels. As this was an untargeted discovery study, a large number of proteins were identified with LC-MS/MS proteomics and found to change significantly following treatment but the multiple statistical comparisons performed could have inflated the type 1 error. As such, the Benjamini-Hochberg method was applied to adjust for multiple-testing with a cut-off q-value of ≤0.10. However, this approach may have been too stringent and resulted in false negatives at this discovery stage.

In conclusion, by using label-free LC-MS/MS quantitative proteomics, we identified several blood proteins involved in complement activation and inflammatory/immune-related pathways that changed in response to IV antibiotic treatment. Early change in IGFR2 correlated with symptom improvement by the end of treatment and requires further validation as an early marker of symptomatic treatment response in individuals with CF.

## Methods

### Study ethics

The research protocol was approved by the University of British Columbia Providence Health Care Research Institute Research Ethics Board (UBC-PHC REB number H12-00835). Informed written consent was obtained from participating subjects. All methods were performed in accordance with the relevant guidelines and regulations to protect human subjects and ensure that participants remain de-identified during samples analysis and data reporting.

### Study cohort

Adult CF subjects were recruited prospectively when diagnosed with a PEx and admitted to St. Paul’s Hospital (Vancouver, BC) for intravenous (IV) antibiotic therapy between July 1, 2013 and June 30, 2015. The PEx events were defined according to changes in respiratory symptoms with a modified Fuchs PEx score of at least 4/12 requiring hospitalization for IV antibiotic treatment3. Subjects were excluded if they had a history of solid organ transplantation or were receiving chronic immunosuppressive treatment. Participating subjects received IV antibiotic treatment in conjunction with airway clearance therapies but the IV antibiotic regimen and duration of PEx treatment were left to the discretion of the most responsible CF physician.

### Blood samples and clinical outcomes

Blood samples and clinical outcomes were obtained from subjects within 24 hours of hospitalization for IV antibiotic therapy (V1), treatment day 5 (V2), treatment day 10 (V3), and IV antibiotic treatment completion (V4). Blood samples (serum, EDTA-treated) were collected and processed following standard operating procedures and then stored at −80 °C until thawing for batched analysis.

Clinical characteristics including age, sex, baseline lung function (i.e. FEV1% predicted), PEx requiring IV antibiotic treatment in the prior year, sputum microbiology, were collected at hospital admission (V1). Baseline lung function was defined as the best FEV1% predicted in the 6 months prior to the index PEx. CF Respiratory Symptom Diary-Chronic Respiratory Infection Symptom Score (CFRSD-CRISS) and FEV1% predicted were recorded at V1 and each of the following time points (V2, V3, V4). CFRSD-CRISS is scaled from 0 to 100 points with a higher score indicating more severe respiratory symptoms21. Clinical outcomes of interest included absolute change in CFRSD-CRISS and relative change in FEV1% predicted from admission (V1) to treatment completion (V4). Absolute change in FEV1% predicted was not analyzed as it is largely influenced by baseline lung function with larger increases seen with higher baseline lung function9,10.

### Untargeted proteomic profiling of blood proteins

Plasma samples were analyzed with label-free liquid chromatography-tandem mass spectrometry (LC-MS/MS) at the University of British Columbia Proteomics Core Facility. In brief, to facilitate the analysis of less abundant plasma proteins, fourteen of the most highly abundant proteins were first immunodepleted using the Human 14 Multiple Affinity Removal Spin Cartridge (Agilent Technologies, Santa Clara, CA), which removes albumin, immunoglobulin (Ig) G, alpha 1-antitrypsin, IgA, transferrin, haptoglobin, fibrinogen, alpha 2-macroglobulin, alpha 1-acid glycoprotein, IgM, apolipoprotein AI, apolipoprotein AII, complement C3 and transthyretin27. Remaining plasma samples were trypsin-digested overnight as previously described27. Resulting peptides were desalted and purified with C-18 STop And Go Extraction (STAGE) Tips28. Purified peptides were fractionated using the Agilent 1100 HPLC system at 50 μL/min flow rate29. The analytical column was operated at 50 °C using an in-house packed 75 μm C18 column heater. The trap column that was added onto the analytical column was a 2 cm-long, 100 μm-inner diameter fused silica, packed with 5 μm-diameter Aqua C-18 beads (Phenomenex, Torrance, CA). Analytical gradient was set at 75 minutes: changing from 10% to 35% Buffer B (80% acetonitrile, 0.1% formic acid) for the first 60 minutes and then wash with 100% Buffer B for 15 minutes. The sample was initially fractionated into 96 wells (45 seconds per well) then pooled in a noncontiguous manner (every 6th well was pooled) resulting into 6 fractions for further LC-MS/MS analysis29. Six fractions were then loaded into the Impact II Q-ToF mass spectrometer (Bruker, Germany)30. Peptides identified by LC-MS/MS were searched with MaxQuant software (version 1.5.3.30) with default label-free-quantitation setting and match-between runs options enabled31. All plasma samples were evaluated in duplicate and mean values were used for analyses. The LC-MS/MS data were deposited in the PRoteomics IDEntifications (PRIDE) database under accession number PXD016089.

### Analysis of candidate blood proteins

Five low-abundance candidate blood proteins, including interleukin (IL)-$${\rm{\beta }}$$, IL-6, IL-8, tumor necrosis factor (TNF)-$$\alpha$$, and calprotectin, that are beyond the detection limits of LC-MS/MS were analyzed in serum samples with multiplex electrochemiluminescence immunoassays (Meso Scale Discovery, Carlsbad, CA). Among these five low-abundance blood proteins, IL-1 $${\rm{\beta }}$$ was below the detection limits for most of the samples, and therefore, was excluded from subsequent analyses. All assays were performed in duplicate with mean coefficient of variation (CV) < 5% and mean values were used for analyses.

### Protein-protein interaction (PPI) network and pathway enrichment analysis

The differentially expressed (DE) proteins identified by LC-MS/MS were applied as inputs for PPI network and pathway enrichment analysis. The STRING database (version 11.0) was utilized to assess the protein functional association32. The active protein-protein interactions were identified based on experimentally determined interactions and curated databases, such as Gene ontology, KEGG, Reactome databases. The predicted protein-protein interactions included gene neighborhood, gene fusion, gene co-occurrence, text-mining, co-expression, and protein homology32. Enriched Gene Ontology (GO) biological process term and Reactome pathways were reported and proteins involved in the immune and inflammation related pathways were highlighted in the constructed PPI network.

### Statistical analyses

Statistical analyses were performed using R (version 3.5.0, the R Foundation for Statistical Computing, Vienna, Austria) and Prism 8 (GraphPad, La Jolla, CA). Continuous variables were presented as mean ± standard deviation (SD). Categorical variables were reported as number with proportions. Longitudinal changes in the levels of candidate blood proteins and changes in clinical outcomes (i.e., FEV1% predicted and CFRSD-CRISS) were analyzed with non-parametric Kruskal-Wallis test followed by the post-hoc Dunn’s test to correct for multiple comparisons.

LC-MS/MS data was pre-processed as described in the flow diagram (e-Fig. 1). Proteins were excluded if they were identified as reversed and/or contaminated by MaxQuant software during peptide searching or detected in less than 25% of blood samples. Three-hundred and forty-one proteins passed the quality control matrix. Missing values from these 341 proteins were imputed with half of the minimum abundance of each protein across the analyzed samples and then, the levels of proteins were log2 transformed before subsequent analyses. Fold-changes of blood proteins from V1 to V2 and V1 to V4 measured with LC-MS/MS were analyzed with the limma R software package and adjusted for baseline lung function, sex, age, and an interaction term between sex and age. The Benjamini-Hochberg method was applied to correct for multiple comparisons and false discovery rate (FDR) adjusted p-values (q-value) ≤ 0.10 were reported for differentially expressed (DE) proteins.

To assess how blood protein levels at V1 might be confounded by baseline disease severity and demographic factors, correlation between blood protein levels at V1 and age, sex, and baseline lung function were evaluated with Spearman’s correlation. Additionally, correlations between early change (V1 to V2) and overall change (V1 to V4) in blood proteins with relative change in FEV1% predicted and absolute change in CFRSD-CRISS from V1 to V4 were examined with Spearman’s correlation. Statistical significance was reported when two-sided p-values were ≤0.05.

To ensure our findings were robust, we performed a post-hoc sensitivity analysis that randomly selected one PEx from three subjects who had repeat PEx.

## Data availability

The datasets generated during the current study are available in the PRoteomics IDEntifications (PRIDE) database under accession number PXD016089. Protein quantification results from MaxQuant software (version 1.5.3.30), study cohort characteristics, and R scripts used for statistical analyses are included in the Supplementary materials files.

## References

1. 1.

Ratjen, F. et al. Cystic fibrosis. Nature Reviews Disease Primers 1, doi:ARTN 15010. https://doi.org/10.1038/nrdp.2015.10 (2015).

2. 2.

Goss, C. H. & Burns, J. L. Exacerbations in cystic fibrosis. 1: Epidemiology and pathogenesis. Thorax 62, 360–367, https://doi.org/10.1136/thx.2006.060889 (2007).

3. 3.

Bilton, D. et al. Pulmonary exacerbation: Towards a definition for use in clinical trials. Report from the EuroCareCF Working Group on outcome parameters in clinical trials. Journal of Cystic Fibrosis 10, S79–S81, https://doi.org/10.1016/S1569-1993(11)60012-X (2011).

4. 4.

Wagener, J. S. et al. Oral, inhaled, and intravenous antibiotic choice for treating pulmonary exacerbations in cystic fibrosis. Pediatric Pulmonology 48, 666–673, https://doi.org/10.1002/ppul.22652 (2013).

5. 5.

Bradley, J. M., Blume, S. W., Balp, M. M., Honeybourne, D. & Elborn, J. S. Quality of life and healthcare utilisation in cystic fibrosis: a multicentre study. Eur Respir J 41, 571–577, https://doi.org/10.1183/09031936.00224911 (2013).

6. 6.

Dill, E. J., Dawson, R., Sellers, D. E., Robinson, W. M. & Sawicki, G. S. Longitudinal trends in health-related quality of life in adults with cystic fibrosis. Chest 144, 981–989, https://doi.org/10.1378/chest.12-1404 (2013).

7. 7.

Cystic Fibrosis Foundation Patient Registry 2017 Annual Data Report. (Cystic Fibrosis Foundation, 2018).

8. 8.

Sanders, D. B. et al. Failure to Recover to Baseline Pulmonary Function after Cystic Fibrosis Pulmonary Exacerbation. Am. J. Respir. Crit. Care Med. 182, 627–632, https://doi.org/10.1164/rccm.200909-1421OC (2010).

9. 9.

Sanders, D. B. et al. Standardized Treatment of Pulmonary Exacerbations (STOP) study: Observations at the initiation of intravenous antibiotics for cystic fibrosis pulmonary exacerbations. J Cyst Fibros 16, 592–599, https://doi.org/10.1016/j.jcf.2017.04.005 (2017).

10. 10.

West, N. E. et al. Standardized Treatment of Pulmonary Exacerbations (STOP) study: Physician treatment practices and outcomes for individuals with cystic fibrosis with pulmonary Exacerbations. J Cyst Fibros 16, 600–606, https://doi.org/10.1016/j.jcf.2017.04.003 (2017).

11. 11.

Biomarkers Definitions Working, G. Biomarkers and surrogate endpoints: preferred definitions and conceptual framework. Clin Pharmacol Ther 69, 89–95, https://doi.org/10.1067/mcp.2001.113989 (2001).

12. 12.

Prins, H. J. et al. CRP-guided Antibiotic Treatment in acute exacerbations of COPD admitted to Hospital. Eur Respir J, https://doi.org/10.1183/13993003.02014-2018 (2019).

13. 13.

Butler, C. C. et al. C-Reactive Protein Testing to Guide Antibiotic Prescribing for COPD Exacerbations. N Engl J Med 381, 111–120, https://doi.org/10.1056/NEJMoa1803185 (2019).

14. 14.

Shoki, A. H., Mayer-Hamblett, N., Wilcox, P. G., Sin, D. D. & Quon, B. S. Systematic review of blood biomarkers in cystic fibrosis pulmonary exacerbations. Chest 144, 1659–1670, https://doi.org/10.1378/chest.13-0693 (2013).

15. 15.

Gray, R. D., Downey, D. & Taggart, C. C. Biomarkers to monitor exacerbations in cystic fibrosis. Expert Rev Respir Med 11, 255–257, https://doi.org/10.1080/17476348.2017.1307739 (2017).

16. 16.

Gray, R. D. et al. Sputum and serum calprotectin are useful biomarkers during CF exacerbation. J Cyst Fibros 9, 193–198, https://doi.org/10.1016/j.jcf.2010.01.005 (2010).

17. 17.

Sagel, S. D. et al. Effect of treatment of cystic fibrosis pulmonary exacerbations on systemic inflammation. Ann Am Thorac Soc 12, 708–717, https://doi.org/10.1513/AnnalsATS.201410-493OC (2015).

18. 18.

Horsley, A. R. et al. Changes in physiological, functional and structural markers of cystic fibrosis lung disease with treatment of a pulmonary exacerbation. Thorax 68, 532–539, https://doi.org/10.1136/thoraxjnl-2012-202538 (2013).

19. 19.

Sharma, A. et al. Clinical utility of C-reactive protein to predict treatment response during cystic fibrosis pulmonary exacerbations. PLoS One 12, e0171229, https://doi.org/10.1371/journal.pone.0171229 (2017).

20. 20.

Roberts, J. M. et al. Multiple reaction monitoring mass spectrometry to identify novel plasma protein biomarkers of treatment response in cystic fibrosis pulmonary exacerbations. J Cyst Fibros 17, 333–340, https://doi.org/10.1016/j.jcf.2017.10.013 (2018).

21. 21.

Cystic Fibrosis Respiratory Symptom Diary (CFRSD) including the Chronic Respiratory Infection Symptom Scale (CRISS) user manual U.S. version 2.0. (2016).

22. 22.

Nixon, L. S., Yung, B., Bell, S. C., Elborn, J. S. & Shale, D. J. Circulating immunoreactive interleukin-6 in cystic fibrosis. Am J Respir Crit Care Med 157, 1764–1769, https://doi.org/10.1164/ajrccm.157.6.9704086 (1998).

23. 23.

Gary-Bobo, M., Nirde, P., Jeanjean, A., Morere, A. & Garcia, M. Mannose 6-phosphate receptor targeting and its applications in human diseases. Curr Med Chem 14, 2945–2953 (2007).

24. 24.

Suh, H. S., Lo, Y., Choi, N., Letendre, S. & Lee, S. C. Insulin-like growth factors and related proteins in plasma and cerebrospinal fluids of HIV-positive individuals. J Neuroinflammation 12, 72, https://doi.org/10.1186/s12974-015-0288-6 (2015).

25. 25.

Pandya, P. H. & Wilkes, D. S. Complement system in lung disease. Am J Respir Cell Mol Biol 51, 467–473, https://doi.org/10.1165/rcmb.2013-0485TR (2014).

26. 26.

Sass, L. A. et al. Complement Effectors of Inflammation in Cystic Fibrosis Lung Fluid Correlate with Clinical Measures of Disease. PLoS One 10, e0144723, https://doi.org/10.1371/journal.pone.0144723 (2015).

27. 27.

Foster, L. J., De Hoog, C. L. & Mann, M. Unbiased quantitative proteomics of lipid rafts reveals high specificity for signaling factors. Proc Natl Acad Sci USA 100, 5813–5818, https://doi.org/10.1073/pnas.0631608100 (2003).

28. 28.

Ishihama, Y., Rappsilber, J., Andersen, J. S. & Mann, M. Microcolumns with self-assembled particle frits for proteomics. J Chromatogr A 979, 233–239 (2002).

29. 29.

Udeshi, N. D., Mertins, P., Svinkina, T. & Carr, S. A. Large-scale identification of ubiquitination sites by mass spectrometry. Nat Protoc 8, 1950–1960, https://doi.org/10.1038/nprot.2013.120 (2013).

30. 30.

Gibbs, M. R. et al. Conserved GTPase LepA (Elongation Factor 4) functions in biogenesis of the 30S subunit of the 70S ribosome. Proc Natl Acad Sci USA 114, 980–985, https://doi.org/10.1073/pnas.1613665114 (2017).

31. 31.

Tyanova, S., Temu, T. & Cox, J. The MaxQuant computational platform for mass spectrometry-based shotgun proteomics. Nat Protoc 11, 2301–2319, https://doi.org/10.1038/nprot.2016.136 (2016).

32. 32.

Szklarczyk, D. et al. STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res 47, D607–D613, https://doi.org/10.1093/nar/gky1131 (2019).

## Acknowledgements

We would like to thank Proteomics Core Facility from the University of British Columbia for performing LC-MS/MS proteomic experiments, as well as Jiah Jang for assistance with sample processing. Also, thanks to our patients for their participation. B.C. Lung Association provided grant support for this study.

## Author information

Authors

### Contributions

Dr. Quon had full access to all the data in the study and had final responsibility for the decision to submit for publication. Mr. Dong: contributed to data extraction, data analysis and interpretation, and drafting and review of the manuscript for important intellectual content. Ms. Moon: contributed to LC-MS/MS proteomic experiment, data extraction, and review of the manuscript for important intellectual content. Ms. Chen: contributed to LC-MS/MS data analyses and review of the manuscript for important intellectual content. Drs. Foster, Tebbutt, and Ng contributed to drafting and review of the manuscript for important intellectual content. Dr. Quon contributed to the study design, data extraction and interpretation, and drafting and review of the manuscript for important intellectual content.

### Corresponding author

Correspondence to Bradley S. Quon.

## Ethics declarations

### Competing interests

Mr. Dong, Ms. Moon, Ms. Chen, and Drs Foster, Tebbutt, Ng, and Quon report no potential conflicts of interest with any of the companies/organizations whose products or services were discussed in this article.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

## Rights and permissions

Reprints and Permissions

Dong, K., Moon, KM., Chen, V. et al. Identification of novel blood biomarkers of treatment response in cystic fibrosis pulmonary exacerbations by label-free quantitative proteomics. Sci Rep 9, 17126 (2019). https://doi.org/10.1038/s41598-019-53759-1

• Accepted:

• Published:

• ### Lipidome Alterations Induced by Cystic Fibrosis, CFTR Mutation, and Lung Function

• Adriana Zardini Buzatto
• , Mai Abdel Jabar
• , Imran Nizami
• , Majed Dasouki
• , Liang Li
•  & Anas M. Abdel Rahman

Journal of Proteome Research (2021)

• ### Proteomics and Metabolomics for Cystic Fibrosis Research

• Nara Liessi
• , Nicoletta Pedemonte
• , Andrea Armirotti
•  & Clarissa Braccia

International Journal of Molecular Sciences (2020)

• ### Laboratory biomarkers for lung disease severity and progression in cystic fibrosis

• Zsolt Bene
• , Zsolt Fejes
• , Milan Macek
• , Margarida D. Amaral
• , István Balogh
•  & Béla Nagy

Clinica Chimica Acta (2020)

• ### Exploring the Relationship between FEV 1 Loss and Recovery and Aminoglycoside Pharmacokinetics in Adult Patients with Cystic Fibrosis: Implications for Clinical Dosing Strategies

• Brian M. Hoff
• , Marc H. Scheetz
• , Manu Jain
• , Joanne F. Cullina
•  & Nathaniel J. Rhodes

Pharmacotherapy: The Journal of Human Pharmacology and Drug Therapy (2020)

• ### Serum-Based Proteomics Profiling in Adult Patients with Cystic Fibrosis

• Hicham Benabdelkamel
• , Hanadi Alamri
• , Meshail Okla
• , Afshan Masood
• , Mai Abdel Jabar
• , Ibrahim O. Alanazi
• , Assim A. Alfadda
• , Imran Nizami
• , Majed Dasouki
•  & Anas M. Abdel Rahman

International Journal of Molecular Sciences (2020)