**Preface to "Research in Metabolomics via Nuclear Magnetic Resonance Spectroscopy: Data Mining, Biochemistry and Clinical Chemistry"**

It is our pleasure to present this Special Issue entitled "Research in Metabolomics via Nuclear Magnetic Resonance Spectroscopy: Data Mining, Biochemistry and Clinical Chemistry", which broadly addresses the applications of nuclear magnetic resonance (NMR) spectroscopy in the metabolomics field.

Metabolomics is defined as the comprehensive characterization of the ensemble of endogenous and exogenous metabolites present in a biological specimen. Metabolites represent, at the same time, the downstream output of the genome and the upstream input from various exogenous factors, such as the environment, lifestyle, and diet. Even though the first scientific paper explicitly dealing with metabolomics is more than 20 years old, we think that this collection is still timely and of particular interest for the scientific community because this "-omic" science is still growing and novel practical applications in biomedicine and in the agricultural field continue to emerge. As researchers at the Magnetic Resonance Center of the University of Florence (Italy) we decided to focus this Special Issue on NMR-based metabolomics, which is our main field of research. The ensemble of the studies present in this volume offers a representative overview of the applications of NMR metabolomics raging from the biomedical fields to food science.

In the end, we want to thank the authors for their precious contributions as this Special Issue would not be possible without them. We would also like to express our sincere appreciation to the dedicated editorial team of Applied Sciences for their valuable contributions to this volume.

> **Alessia Vignoli, Gaia Meoni, and Leonardo Tenori** *Editors*

### *Editorial* **Applications and Challenges for Metabolomics via Nuclear Magnetic Resonance Spectroscopy**

**Alessia Vignoli 1,2,3,\*, Gaia Meoni 1,2,3,\* and Leonardo Tenori 1,2,3,\***


#### **1. Introduction**

Even though metabolomics is about 20 years old, the interest in this "-omic" science is still growing, and high expectations remain in the scientific community for new practical applications in biomedicine and in the agricultural field. Thus far, biomedical metabolomic studies have produced great advancements in biomarker discovery, identification of novel metabolites, and more detailed characterization of biological pathways involved in the manifestation and progression of diseases. In parallel, metabolomics has been shown to have an emerging role in monitoring the influence of different manufacturing procedures on food quality and food safety. In light of the above, this Special Issue was introduced to collect the latest research from various application fields of NMR-based metabolomics [1,2], ranging from biomedicine to data mining and food chemistry.

#### **2. NMR-Based Metabolomics**

Our collection comprises four research articles that report interesting applications of NMR metabolomics in the biomedical setting. In the first article published in our issue, Baranovicova et al. [3] present a longitudinal study that explores the dynamics of metabolomic changes in the plasma of 53 patients, diagnosed with SARS-CoV-2 infection, at three consecutive time points during their first week of hospitalization (days 1, 3, and 7 after admission to the hospital) to reveal the differences among patients with positive (survivors) and negative (worsening condition, non-survivors) outcomes. People with COVID-19, regardless their prognosis, presented alterations in their energy and amino acids metabolism. These changes were normalized by the seventh day in patients with positive outcomes; conversely, they were not reverted in patients with negative outcomes. These results indicate that the ability to respond to metabolomic alterations induced by severe inflammation due to SARS-CoV-2 infection is a key factor in determining patients' outcomes and that these metabolic changes can be tackled with individual pharmacological or diet interventions to support patient response.

In recent years, nanoscience and nanotechnology have been developing rapidly; at the same time, the increased use of nanoparticles has raised several concerns regarding human public health and occupational safety. In the article by Horník et al. [4], NMRbased metabolomics of exhaled breath condensate (EBC) and blood plasma is used to study the effects of occupational exposure to nanoparticles (NPs). The EBC and blood plasma samples from 20 workers exposed to NPs were collected pre-shift (i.e., before 2.5 h of exposure to NPs) and post-shift (i.e., after NP exposure). Moreover, 20 controls (not exposed to NPs) were enrolled for this study. Multivariate statistical analyses, performed both on EBC and plasma NMR data, showed clear discriminations among the three groups of interest (the pre-shift, post-shift, and control groups). The univariate metabolite analysis revealed several alterations in subjects exposed to NPs, in particular the acute effect of NP exposure is primarily reflected in the metabolic pathways involved in the production of

**Citation:** Vignoli, A.; Meoni, G.; Tenori, L. Applications and Challenges for Metabolomics via Nuclear Magnetic Resonance Spectroscopy. *Appl. Sci.* **2022**, *12*, 4655. https://doi.org/10.3390/ app12094655

Received: 29 April 2022 Accepted: 4 May 2022 Published: 6 May 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

antioxidants and of other protective species, whereas the chronic effect of NP exposure seems to be associated with alterations in glutamine and glutamate metabolism, and the purine metabolism pathways.

The paper authored by Vignoli et al. [5] characterizes the effects of surgery on the serum metabolomic profiles of colorectal cancer (CRC) patients and explores the possibility that metabolic variations among preoperative and postoperative serum samples could be informative on future cancer recurrence. A total of 41 patients diagnosed with early-stage CRC and scheduled for radical resection were enrolled for this study. Serum samples collected preoperatively (t0) and 4–6 weeks after surgery but before the start of any treatment (t1) were analyzed via 1H NMR spectroscopy. A clear discrimination between t0 and t1 emerged: after surgery, there are significant increases in pyruvate, HDL cholesterol, HDL phospholipids, HDL Apo-A1, and HDL Apo-A2 levels, coupled with significant decreases in acetone, 3-hydroxybutyrate, LDL-Chol/HDL-Chol ratio, and Apo-A1/Apo-B100 ratio. Taken together, these results point to a relevant rearrangement of the metabolic pathways related to lipoproteins, ketone bodies, and energy metabolism. Furthermore, several differences between post- and pre-operative serum samples, in particular those related to the HDL-Chol and VLDL-Chol subfractions, seem to be associated with cancer recurrence. These data pave the way for novel strategies for risk stratification in patients with early-stage CRC.

The paper by Georgiopoulou et al. [6] is the last research article related to biomedical applications of NMR metabolomics included in our issue. It proposes an analysis of urine samples of preterm infants with neonatal sepsis, a systemic infection difficult to diagnose in its early stages and thus reporting high rates of morbidity and mortality. In this study, the urine metabolomic profiles of 34 septic neonates, 14 preterm neonates without sepsis or other serious morbidity but hospitalized in the NICU, and 23 healthy preterm neonates were examined. Multivariate and univariate statistical analyses showed clear discriminations between septic and healthy newborns. In particular, alterations in the levels of gluconate, myo-inositol, hippurate, taurine, N, N-Dimethylglycine, betaine, creatinine, glucose and lactose emerged as the most significant. These data represent a promising basis for future large-scale multicenter studies and give new perspectives for clinical research in the field of neonatology.

We decided to address also foodomics in our issue, which refers to metabolomic approaches applied to foodstuff for investigating topics related to nutrition, fraud detection and traceability of the geographical origin and production/processing procedures of food. In this regard, in our issue, we decided to publish an NMR-based metabolomic study based on water extracts of green and roasted coffee beans of different cultivars from three distinct Nicaraguan farms [7]. We think that this study can show well the potential and versatility of NMR metabolomics. Here, the authors demonstrate the potential of NMR metabolomics not only to define the geographical origin and the farm of provenance but also to characterize the effect of the environment (microclimates, irrigation, fertilizers, etc.) and the post-harvest practices (e.g., drying and fermentation) that are responsible for different aroma precursors in coffee and thus affect its distinct taste.

The ensemble of these studies offers a representative overview of the applications of NMR metabolomics raging from the biomedical fields to food science.

The capabilities of NMR, coupled with an ever-growing list of statistical chemometric techniques, make NMR-based metabolomics a versatile technique. Applying correct and suitable statistical techniques has become of fundamental importance for metabolomics studies. For this reason the review of Corsaro et al. [8], which lists some of the most commonly used and useful statistical techniques in metabolomics, explaining their advantages and disadvantages, has been included in our issue. In this work, the authors give an overview of the wide range of statistical opportunities for NMR-based metabolomics, ranging from conventional approaches (e.g., unsupervised and supervised methods, and pathway analyses) to less frequently applied deep learning and artificial neural networks. We found this review

beneficial not only for fledgling metabolomic students approaching chemometrics but also for experts in the field looking for a more suitable approach to their studies.

In conclusion, the current Special Issue of *Applied Sciences* offers a variety of examples on how NMR-based metabolomics can potentially be used in several and varied settings. Although this Special Issue has been closed, more in-depth research on this topic is expected in the years to come, and future research will no doubt continue to explore the possibility of translating metabolomics into real-life applications.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **The Ability to Normalise Energy Metabolism in Advanced COVID-19 Disease Seems to Be One of the Key Factors Determining the Disease Progression—A Metabolomic NMR Study on Blood Plasma**

**Eva Baranovicova 1, Anna Bobcakova 2, Robert Vysehradsky 2, Zuzana Dankova 1, Erika Halasova 1, Vladimir Nosal <sup>3</sup> and Jan Lehotsky 4,\***


**Abstract:** Background: COVID-19 represents a severe inflammatory condition. Our work was designed to monitor the longitudinal dynamics of the metabolomic response of blood plasma and to reveal presumable discrimination in patients with positive and negative outcomes of COVID-19 respiratory symptoms. Methods: Blood plasma from patients, divided into subgroups with positive (survivors) and negative (worsening condition, non-survivors) outcomes, on Days 1, 3, and 7 after admission to hospital, was measured by NMR spectroscopy. Results: We observed changes in energy metabolism in both groups of COVID-19 patients; initial hyperglycaemia, indicating lowered glucose utilisation, was balanced with increased production of 3-hydroxybutyrate as an alternative energy source and accompanied by accelerated protein catabolism manifested by an increase in BCAA levels. These changes were normalised in patients with positive outcome by the seventh day, but still persisted one week after hospitalisation in patients with negative outcome. The initially decreased glutamine plasma level normalised faster in patients with positive outcome. Patients with negative outcome showed a more pronounced Phe/Tyr ratio, which is related to exacerbated and generalised inflammatory processes. Almost ideal discrimination from controls was proved. Conclusions: Distinct metabolomic responses to severe inflammation initiated by SARS-CoV-2 infection may serve towards complementary personalised pharmacological and nutritional support to improve patient outcomes.

**Keywords:** NMR metabolomics; human plasma; COVID-19

#### **1. Introduction**

COVID-19, which develops after SARS-CoV-2 infection, represents a severe inflammatory condition. Over the past two decades, a close link between metabolism and immunity has emerged [1,2]. The immune reaction in severe inflammation is intimately associated with a dependency on amino acids included in the proteosynthesis and specific metabolism of immunocompetent cells [3]. In addition, the immune response of the organism is also closely related to glucose energetical metabolism [1,2,4–6]. Synergic interactions between metabolism and immune processes serve as a tool to monitor the particular state of an organism relating to immunological response via metabolomics analysis. The increasing

**Citation:** Baranovicova, E.;

Bobcakova, A.; Vysehradsky, R.; Dankova, Z.; Halasova, E.; Nosal, V.; Lehotsky, J. The Ability to Normalise Energy Metabolism in Advanced COVID-19 Disease Seems to Be One of the Key Factors Determining the Disease Progression—A Metabolomic NMR Study on Blood Plasma. *Appl. Sci.* **2021**, *11*, 4231. https://doi.org/ 10.3390/app11094231

Academic Editor: Alessia Vignoli

Received: 8 April 2021 Accepted: 4 May 2021 Published: 7 May 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

number of studies confirms the great potential of the metabolomic approach in the evaluation of COVID-19 disease, its course, and its outcome [7–10]. A comprehensive metanalysis of COVID-19 patients showed several key metabolic characteristics for disease progression and clinical outcome [11]. Untargeted metabolomics on patients' serum via mass spectroscopy revealed potential prognostic markers of both severity and outcome [10,12]. Interestingly, metabolomics may also predict antiviral drug efficacy in COVID-19 [13], and metabolomic analysis of patients' exhaled air can identify patients with COVID-19 in acute respiratory distress syndrome. NMR-based metabolomic profiling of blood samples has been already used to monitor COVID-19 patients' response to tocilizumab [14].

We focused herein on the dynamics of metabolomic changes in blood plasma at three successive time points during the first week of COVID-19 patient hospitalisation, with patients divided into two groups: (i) those with a positive outcome (survivors) and (ii) those with a negative outcome (non-survivors or obviously worsening condition). Hospitalised COVID-19 patients with clinically proven moderate-to-severe pneumonia with acute hypoxemic respiratory failure were included. We were interested to explore the metabolic changes in blood plasma that could be associated with immune cell response, as well as with energy metabolism, in comparison to control subjects representing a sample of the normal population, without any acute or chronic inflammatory or pulmonary diseases. Secondarily, it was of interest as to whether there are metabolomic features in blood plasma that could predict patient outcome, at which time point are they recognisable, and to what extent. Complementary to testing significant changes, we also employed a discriminatory algorithm in the search for metabolites that could serve alone or in combination as plasma biomarkers.

#### **2. Materials and Methods**

#### *2.1. Subjects*

Altogether, 53 patients with PCR-confirmed SARS-CoV-2 were included in the study. Patients were admitted to the Clinic of Pneumology and Phthisiology, Martin University Hospital, Slovakia, due to chest X-ray/CT signs of bilateral pneumonia and acute hypoxemic respiratory failure requiring oxygen supplementation (oxygen saturation at <94% in room air). In general, patients presented with typical symptoms of COVID-19: fever, cough, dyspnoea, weakness, fatigue, myalgia and arthralgia, loss of smell and taste, and loss of appetite. Some patients suffered from gastrointestinal symptoms (diarrhoea) as well. Laboratory results on admission showed increased inflammatory markers (CRP, IL-6, ferritin, fibrinogen) and hypoxemic respiratory failure, and changes in differential blood count included leucocytosis, lymphopenia, neutrophilia, and eosinopenia in most patients.

During the study, patients received either standard hospital enteral nutrition or a diabetic diet (patients with diabetes). Patients incapable of oral food intake received the equivalent for enteral nutrition via nasogastric tube. None of the included patients had percutaneous endoscopic gastrostomy/jejunostomy. Neither nutritional supplementation nor parenteral nutrition was administered. When necessary, but only sporadically, patients received crystalloid solutions to treat dehydration or mineral imbalance.

Oxygen was administered via nasal cannula, face mask, or face mask with a rebreathing bag with flow adjusted to achieve target oxygen saturation of 94%. Seven patients required high-flow nasal oxygen therapy (HFNO), and in case of hypoxemic–hypercapnic respiratory failure, three received non-invasive ventilation (NIV). In patients with severe and critical clinical condition requiring a very high flow of oxygen, saturation of 90% was considered sufficient. None of the included patients received mechanical ventilation during sample collection; however, two patients were later intubated and mechanically ventilated. Apart from oxygen supply, patients were treated with dexamethasone (all patients, dose of 6 mg/day for a duration of 10 days); antivirals (remdesivir or favipiravir if eligible according to local guidelines—duration of symptoms less than 7 days), n = 17; antibiotics (in case of bacterial superinfection or its suspicion), n = 53; LMWH, n = 49; vitamins: vitamin C, n = 17, vitamin D, n = 19; zinc, n = 14; and betaglucans, n = 44.

Patients were divided into two subgroups: Group A (n = 34) contained patients with a positive outcome (survivors), while Group B (n = 19) contained patients with a negative outcome, i.e., patients with a worsening condition during the sampling period, or those who died (10 were dead at the time of manuscript preparation). All known patient comorbidities at the time of study enrolment are listed in Table 1. To assess the patients' condition, the determining criterion was the need for increasing/decreasing oxygen flow or switch to HFNO, NIV, or mechanical ventilation to achieve target oxygen saturation, together with clinical evaluation and known clinical outcome. Due to various causes such as hospital discharge before Day 7, death, or even patient disagreement with other blood collections, the number of samples on Day 3 or Day 7 is slightly reduced. All details about subjects included in the study are summarised in Table 1.


**Table 1.** Characteristics of patients included in the study.

\* Patient with chronic hypoxemic–hypercapnic respiratory failure due to COPD on home NIV (non-invasive ventilation) with LTOT (long-term oxygen therapy).

As controls, plasma samples from age- and gender-matched subjects without any acute or chronic inflammatory diseases, any type of respiratory failure, or any pulmonary diseases, regardless of common highly age-related conditions (hypertensia, obesity, and others in the representative sample of the population) were used, representing a 'sample of the normal population', collected in a fasting state without any additional criteria. Included were 55 subjects: median age 64, IQR 18, female n = 25.

#### *2.2. Sample Preparation*

Blood was collected in EDTA-coated tubes, in the fasting state, after the first night in the hospital (Day 1) and then 2 and 6 days later (Day 3 and Day 7). Within 1 h after collection, blood was centrifuged to plasma at 4 ◦C, at 2000 rpm, for 20 min and stored at −80 ◦C until use. Plasma denaturation was carried out according to Gowda et al. [15]: 600 μL of methanol was added to 300 μL of blood plasma. The mixture was briefly vortexed and frozen at −24 ◦C for 20 min. After subsequent centrifugation at 14,000 rpm for 15 min, 700 μL of supernatant was taken, dried out, and stored at −24 ◦C. Before NMR measurement, the dried matter was mixed with 100 μL of stock solution (consisting of: phosphate buffer 200 mM pH 7.4 and 0.30 mM TSP-d4 (trimethylsilylpropionic acid -d4) as a chemical shift reference in deuterated water) and 500 μL of deuterated water. Finally, 550 μL of the final mixture was transferred into a 5 mm NMR tube.

#### *2.3. NMR Measurement*

NMR data were acquired on a 600 MHz Avance III NMR spectrometer from Bruker, Germany, equipped with a TCI CryoProbe at T = 310 K. Initial settings (basal shimming, receiver gain, and water suppression frequency) were performed on an independent sample and adopted for measurements. After preparation, samples were stored in a Sample Jet automatic machine, cooled at approximately 5 ◦C. Before measurement, each sample was preheated to 310 K for 5 min. An exponential noise filter was used to introduce 0.3 Hz line broadening before Fourier transform. All data were zero-filled. Samples were randomly ordered for acquisition.

We modified standard profiling protocols from Bruker as follows: denaturised plasma: noesy with presaturation (noesygppr1d): FID size 64k, dummy scans 4, number of scans 64, spectral width 20.4750 ppm; profiling cpmg (cpmgpr1d, L4 = 126, d20 = 3ms): number of scans 64, spectral width 20.4750 ppm. For 15 randomly chosen samples, 2D spectra were measured: cosy with presaturation (cosygpprqf): FID size 4k, dummy scans 8, number of scans 16, spectral width 16.0125 ppm; homonuclear J-resolved (jresgpprqf): FID size 8k, dummy scans 16, number of scans 32. Samples were randomly ordered for acquisition. For denaturised plasma samples, we kept the half-width of the TSP-d4 signal under 1.0 Hz. All experiments were conducted with a relaxation delay of 4 s.

#### *2.4. Data Processing*

Spectra were solved using the human metabolomic database (www.hmda.ca, accessed on 23 March 2021) [16], chenomics software free trial version, internal metabolite database, and research in the metabolomic literature [15]. The proton NMR chemical shifts are reported relative to the TSP-d4 signal assigned a chemical shift of 0.000 ppm. The peak multiplicities were confirmed in J-resolved spectra, and homonuclear cross peaks were confirmed in 2D cosy spectra. Peak assignments are listed in Table 2.

All spectra were binned to bins of size 0.001 ppm. No normalisation method was applied to the data. Then, the intensities of selected bins were summed only for spectra subregions with only one metabolite assigned or minimally affected by other co-metabolites. Metabolites showing weak intensive peaks or strong peak overlap were excluded from the evaluation. The obtained values were used as relative concentrations of particular metabolites.

Besides principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA), we applied the random forest (RF) discriminatory algorithm on the data. We ran nonparametric ANOVA (Kruskal–Wallis) and the nonparametric Mann–Whitney U-test to test significance. For data processing and analyses, we used the online tool metaboanalyst 5.0 [17], Origin Pro 2019, PASW Statistics software, and Matlab 2018b.


**Table 2.** Chemical shifts (in ppm), J couplings (in Hz), and multiplicities (s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; dd, doublet of doublets; dq, doublet of quartets) for the pool of metabolites identified in blood plasma. Signals marked with # were not suitable for quantitative analyses.

#### **3. Results**

Altogether, 24 metabolites were identified in denatured plasma in both patients and healthy subjects, where the signals from 19 compounds were sufficient for quantitative evaluation (Table 2). Further in the text, we use the trivial names of 2-ketoacids derived from leucine, isoleucine, and valine (IUPAC names are in Table 2). Besides molecular metabolites, we also evaluated the lipoprotein fraction, which, as described by Liu et al., contains very-low-density lipoproteins (VLDL), low-density lipoproteins (LDL), and high-density lipoproteins (HDL), including up to one-third of triacylglycerides [18]. For multivariate analyses, we used the relative concentrations of plasma metabolites (expressed as the integral of a particular spectral region) as an input in order to target biologically informative value. We avoided feeding the algorithms with binned NMR spectra as is common in metabolomic studies, since there may be regions of NMR spectra marked as important that are not straightforward and unambiguously related to biological relevance.

Firstly, the data of all patients were analysed (Group A and Group B together) on Day 1 against controls by PCA and PLS-DA (Figure 1). In contrast to patients, controls were relatively clustered together. The loading values were the highest for glucose, 3 hydroxybutyrate, and leucine in PC1 and alanine, lactate, and glutamine in PC2. The situation was very similar after the PLS-DA run. The 10-fold cross-validated PLS-DA algorithm performed with accuracy of 0.954, R2 of 0.7926, and Q2 of 0.6749 for eight components. The variables with the highest VIP scores were: glucose, 3-hydroxybutyrate, alanine, leucine, valine, and glutamine (performance measured in accuracy). The incorporation of additional variables did not improve the performance.

**Figure 1.** PCA (**left**) and PLD-DA analyses (**right**) of the system: patients in the hospital on Day 1 versus controls; algorithms were fed by the relative concentrations of plasma metabolites, and analyses were run in metaboanalyst [16].

The PCA and PLS-DA analyses of the ternary system comprising Group A and Group B on Day 1 and the controls showed a very similar result to those for the previous binary system, where the patients were clustered together relatively well and patients were scattered among themselves without obvious differentiation between patient groups (results shown in Figure S1 in the supplement). PLS-DA analyses were further used to differentiate patient data on a given day. The results from these analyses can be summarised as follows (the best result, performance measured in accuracy): Day 1, accuracy of 0.73, R2 of 0.138 (one component); Day 3, accuracy of 0.76, R2 of 0.3905 (five components); and Day 7, accuracy of 0.72, R2 of 0.387 (four components). In all cases, Q2 values were negative, which suggests an overfitted model.

In the next step, we employed the random forest (RF) discriminatory algorithm to obtain a more realistic estimation of the discriminatory power of the system since RF is relatively robust to overfitting and outliers [19]. The RF algorithm used included crossvalidation via balanced subsampling. It worked with two-thirds of the data for training and the rest for testing for regression, and about 70% of the data for training and the rest for testing during classification to overcome the negative aspects of training and testing on the same data. This approach partially substitutes the validation on an independent data set. As input variables also for this algorithm, we used relative concentrations of metabolites in plasma expressed by the spectral integrals of particular NMR regions. In the case of highly correlating predictors, RF may label some of them as unimportant, so RF was launched 10 times. Within the RF re-runs, metabolites slightly permuted in the importance order. As an output from these analyses, receiver operating characteristic curve (ROC) curves were created. The ROC is defined only for binary systems, and it is created by plotting the true-positive rate against the false-positive rate at various threshold settings. An important output is the area under the curve (AUC), which represents ranking quality. The AUC of a ranking is 1 (the maximum AUC value) when all samples are truly assigned into the groups. An AUC of 0.5 is equivalent to randomly classifying subjects as either positive or negative (i.e., the classifier is of no practical utility) [20]. We ran RF discriminatory analyses for the systems of patients versus controls, Group A versus controls, Group B versus controls, and Group A versus Group B on Days 1, 3, and 7. The results of RF classifications are summarised in Table 3.


**Table 3.** Outputs from random forest discriminatory analyses for selected systems.

For significance testing among relative concentrations of plasma metabolites in patients against controls and patients' dynamic data, we used nonparametric ANOVA, known as the Kruskal–Wallis test. Due to the relatively low sample sizes, we continued with nonparametric testing via the Mann–Whitney U-test for the combination of binary data sets. The details are listed in Table 4. The Phe/Tyr ratio was also used as one variable. As the threshold to claim significance, the p-value was set to 0.05, as established. In the discussion, we did not strictly adhere to p-values, but we focused rather on the data behaviour visualised in the box plots.

**Table 4.** Results from statistical tests; *p*-value derived from nonparametric ANOVA and Mann–Whitney U-test.



**Table 4.** *Cont.*

#### **4. Discussion**

#### *4.1. Discriminatory Analyses*

PCA and PLS-DA analyses are well-established tools when evaluating multidimensional data. PCA analysis serves rather as a 2D visualisation of data sets indicating group proximity. PLS-DA includes a discriminatory algorithm and may be used also to differentiate among groups. PCA analysis of the patient data collected on Day 1 against controls showed controls clustered together, whilst patients were scattered in 2D space. This suggests the great data variability in patient samples, which was more or less confirmed by PLS-DA. As PLS-DA is known to overfit the data [19], for biomarker discovery, we employed a cross-validated RF algorithm. As an output, the ROC curve was created. For the system of patients on Day 1 and controls, RF performed very well with an AUC of 0.995 for five variables with an out-of-bag error of 3/108. The variables Phe/Tyr ratio, phenylalanine, 3-hydroxybutyrate, acetate, and glucose were of the highest importance. The corresponding ROC curve is shown in Figure 2.

Very similar performance—almost ideal discrimination—was achieved for the systems of Group A on Day 1 against controls and Group B on Day 1 against controls (details in Table 3). The five metabolites of the highest importance were identical to those before: phenylalanine, Phe/Tyr ratio, acetate, 3-hydroxybutyrate, glucose, permuted with glutamine, and proline.

The possibility to discriminate between acute COVID-19 patients and healthy controls has been proven in previous studies [7,10,11]. These studies covered another spectrum of metabolites evaluated by different analytical tools as NMR spectroscopy. Here, we also note that metabolites that were marked as the most important in the discrimination algorithm may not be specific to COVID-19 disease, since as discussed in the next text, they are generally related to inflammation, immune response, and energy metabolism.

**Figure 2.** ROC curve with AUC values for systems of COVID-19 patients on Day 1 vs. controls, determined by random forest algorithm with relative concentrations of metabolites in blood plasma as input variables; analysis run in metaboanalyst [16].

It was of interest to see whether there are any metabolites in blood plasma that could serve as potential predictors of disease progress/outcome. We ran RF discrimination for binary systems of patients' groups on collection days. Here the performance was weaker, with AUC values of 0.67 on Day 1 and 0.78 on Day 3 for common, permuting variables: Phe/Tyr ratio, alanine, lysine, glutamine, leucine, and phenylalanine. A further increase in the number of variables did not improve the performance of the discrimination analysis. For the data set of Group A versus Group B on Day 7, the system did not show any discriminatory potential, with an AUC value of 0.503, in other words, the classification was not relevant. Based on this, the biochemical changes observed were rather indicative, not defining unambiguous biomarkers for patient outcome.

#### *4.2. Metabolomic Changes*

Patients hospitalised due to a severe course of COVID-19 showed a significantly increased glucose level on Day 1. All patients were equally treated over the whole time period with dexamethasone, which is known to impair glucose metabolism [21] via the stimulation of gluconeogenesis from amino acids released from muscles, and even one dose of 10 mg dexamethasone may lead to a temporarily increased blood glucose level [22]. The hyperglycaemia in COVID-19 patients treated with dexamethasone is presumably caused by 'triple insult': dexamethasone-induced impaired glucose metabolism, COVID-19-induced insulin resistance, and COVID-19 impaired insulin production [23]. Prolonged uncontrolled hyperglycaemia, regardless of diabetes mellitus, seems to be important in the pathogenesis of COVID-19 [24]. In our study, the hyperglycaemia normalised in Group A, but not in patients with unfavourable outcome included in Group B (Figure 3). This observed result is in line with general knowledge that hyperglycaemia is an unfavourable state in many clinical conditions, i.a., in severe inflammation [25], and is one of the important risk factors of COVID-19 disease progression [26]. The plasma levels of glycolytic products pyruvate and, eventually, lactate were not significantly changed in any group of patients. The relative plasma level of alanine, a metabolite that contributes significantly to liver gluconeogenesis, was decreased on Day 1 in both groups but normalised in patients with a positive outcome on Days 3 and 7; however, it stayed decreased in Group B (figure not shown).

**Figure 3.** Relative concentrations of selected metabolites in blood plasma for patient Groups A and B on Day 1, Day 3, and Day 7. Values are relativised to the median of controls set to 1.

In the blood plasma of COVID-19 patients, we observed a significantly increased level of 3-hydroxybutyrate, a ketone bodies representative. Besides serving as an energy source for the brain, heart, and skeletal muscle, ketone bodies play pivotal roles as signalling mediators, drivers of protein post-translational modification, and modulators of inflammation and oxidative stress [27]. 3-hydroxybutyrate exerts a predominantly anti-inflammatory response [28–30], but can also be pro-inflammatory [31]. A recent study on COVID-19 patients already showed dysbalance in ketone bodies [32]. In our study, the initially increased plasma level of 3-hydroxybutyrate decreased over Day 3 and Day 7 in Group A, but it stayed at an elevated level in Group B on the third and seventh days (Figure 3). Interestingly, the glucose level in this patient group also remained high. As we did not analyse the level of C peptide as a representative of the insulin level, we can hypothesise that the proposed glucose resistance or insufficient glucose utilisation is compensated by ketone bodies. The increase in the 3-hydroxybutyrate level in COVID-19 patients is accompanied by a decreased amount of lipoprotein fraction in blood plasma in patients suffering from COVID-19, containing up to one-third of triacylglycerides [18] as one of the additional substrates for ketone body synthesis (boxplot not shown).

We observed a decreased citrate level in the blood plasma in COVID-19 patients, suggesting alteration of the TCA cycle (Figure 3), similar to the results of a recent study by Pang et al. [11]. Besides including α-ketoglutarate, an essential substrate for endogenous glutamate/glutamine synthesis, there is evidence that TCA cycle intermediates also have an epigenetic impact by influencing DNA and histone methylation, including immune cells [33]. Further, the metabolite creatine, a part of muscle energy metabolism, was significantly increased in the blood plasma of COVID-19 patients compared with controls in both groups, rising with the time of hospitalisation (Figure 3). Patients forced to lie in bed for a sustained period lack spontaneous movement utilising muscle energy, which is probably the reason for the increase of plasma creatine.

BCAAs (branched chain amino acids), including leucine, isoleucine, and valine, share a common pattern of extrahepatic metabolism, and their relative plasma concentrations were represented similarly in both groups of patients. In Figure 3, we show only the dynamics of leucine since isoleucine and valine behaved almost identically. As a representative of ketoacids derived from BCAAs, we show only the course of ketoleucine, as the dynamics was repeated for the other two ketoacids: ketovaline and ketoisoleucine. Increased leucine in COVID-19 patients was reported by Dierckx et al. [34]. There is an

established association between elevated circulating BCAAs and their deleterious effects, as their increased concentration may promote oxidative stress and inflammation [35], having also a neurological impact [36,37]. By monitoring dynamic changes for two different patient subgroups, we observed that initially increased plasma levels of BCAAs in both groups slowly decreased in Group A but not in Group B (Figure 3). Interestingly, the mean values of BCAAs in Group B obviously follow the course of the plasma glucose levels. The increase of BCAAs at time of impaired glycolysis and increased use of fatty acids were very recently discussed in a comprehensive review by Holecek [38], showing the important role of BCAAs in energy metabolism.

Taking the above discussed results together, severe inflammation induced by COVID-19 caused changes in energy metabolism, where we observed increased blood glucose that implies lowered glucose utilisation (the influence of dexamethasone treatment cannot be omitted). In balance, the body, including immune cells, uses ketone bodies (observed increased 3-hydroxybutyrate together with decreased triacylglycerides) as an energy source, and, alternatively also amino acids released by accelerated protein catabolism (increased levels of essential amino acids BCAAs). Interestingly, although all patients in both groups received the dexamethasone treatment during the follow-up period, the above mentioned changes normalised only in patients with a positive outcome; however, they persisted in patients with a negative outcome (more than half of them had died at the time of writing). This course was independent of the patients' diet (Figure S2 in Supplement).

In acute inflammatory conditions, the demand on glutamine increases [39] which may lead to its plasma decrease if the endogenous synthesis of glutamine does not fulfil the requirements of the body [39]. Glutamine serves besides others as a fuel for immune cells—lymphocytes, neutrophils, and macrophages [39–42]—and plays a crucial role in cytokine production [42]. In our study, we noticed a decrease in the glutamine plasma level in COVID-19 patients on Day 1, observed to a lower extent in Group A, which is in accordance with another study where glutamine deficiency may have contributed to disease severity [43]. The glutamine plasma level normalised in both groups, but this was faster in Group A (Figure 4). On Day 7, both groups of patients showed plasma glutamine levels very similar to the level in control subjects, where probably the balance between glutamine production and utilisation had stabilised (Figure 4). Accelerated spontaneous stabilisation of glutamine levels in patients with better outcome supports the results from another study, where the administration of glutamine in the early period of infection suggested a shortened hospital stay and decreased the need for ICU stay [40].

Another significant metabolic parameter associated with immune activation and inflammation is the Phe/Tyr ratio [44,45]. Perturbations in phenylalanine and tyrosine biosynthesis were recognised in SARS-CoV-2 patients by Barberis et al. [46]. In our study, both groups showed initially increased plasma phenylalanine levels, as observed in another study [34], and the level tended to decrease in Group A but not in Group B (Figure 4). The plasma tyrosine level did not show any substantial change. The Phe/Tyr ratio was calculated by dividing the relative concentrations of both metabolites. The obtained value is only the relative ratio, but for comparison, it has the same informative value. The Phe/Tyr ratio was increased in both groups, obviously higher in patients with unfavourable outcome, where a course towards control levels was slowed down in Group B against Group A (Figure 4). Positive relationships between the Phe/Tyr ratio and immune activation markers have been described earlier in several papers [44,45]. It was suggested that suppression of body inflammation can, to a certain extent, improve abnormalities in Phe metabolism within associated neuropsychiatric symptoms [44], among which, e.g., depression and fatigue are some of the most recognised post-COVID-19 difficulties [47].

**Figure 4.** Relative concentrations of selected metabolites related to immunity in blood plasma for patient Groups A and B on Day 1, Day 3, and Day 7. Values are relativised to the median of controls given a value of 1.

#### **5. Conclusions**

Metabolomic changes in blood plasma analysed by NMR in patients suffering COVID-19 were strong enough to obtain almost ideal discrimination from controls, where the ROC derived from random forest showed an AUC of 0.995 for the variables 3-hydroxybutyrate, phenylalanine, acetate, glucose, and Phe/Tyr ratio. The inflammation by COVID-19 caused changes in the body's energy metabolism, where we observed increased blood glucose that implies lowered glucose utilisation, balanced with increased production of 3-hydroxybutyrate as an alternative energy source. Besides that, increased essential BCAAs are a sign of accelerated protein catabolism, offering a further energy source. Interestingly, although all COVID-19-positive patients received dexamethasone treatment during the follow-up period, the above mentioned changes (increased glucose, 3-hydroxybutyrate, and BCAAs levels in blood plasma) normalised only in patients with positive outcome by the seventh day; however, they persisted for over one week in patients with negative outcome (more than half of them had died at the time of writing). Further, patients suffering COVID-19 showed decreased plasma glutamine that normalised faster in patients with a positive outcome. With the length of hospital stay, plasma levels of creatine increased in patients in both groups. Increased Phe/Tyr ratio, which is closely connected with neuropsychiatric morbidities, often reported as post-COVID-19 symptoms, was more pronounced in patients with a negative outcome. Based on our results, the ability of patients to normalise energy metabolism seems to be one of the key factors determining the disease progression. This trend was observed independently of patient diet, which differed with respect to diabetic condition. This study documents evident differences in the course of the metabolomic response to COVID-19 in relation to patient outcome. However, the described changes may not be unique for COVID-19 since they reflect generalised immune response and alterations in body energy metabolism as well. The presented results may serve towards complementary personalised pharmacological and nutritional support in order to improve patient outcomes in severe inflammatory conditions.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/app11094231/s1, Figure S1: PCA (left) and PLD-DA analyses (right) of the system of patients divided into subgroups Group A and Group B on Day 1 versus controls; algorithms were fed relative concentrations of plasma metabolites, and analyses were run in metaboanalyst [16]. Figure S2. The relative changes in two metabolites closely related to energy metabolism—glucose and 3-hydroxybutyrate—where both Groups A and B were divided into subgroups according to patient diet (according to presence of diabetes) on Days 1, 3, and 7 after hospital arrival; not dia = non-diabetic patients on a normal diet, dia = diabetic patients on a diabetic diet. Values were relativized to median of controls set to 1.

**Author Contributions:** Conceptualization, E.B., A.B., R.V., V.N. and J.L.; methodology, E.B.; formal analysis, E.B.; investigation, E.B., Z.D. and J.L.; data curation, E.B., A.B., R.V., Z.D. and J.L.; writing original draft preparation, E.B..; writing—review and editing, A.B., E.H., V.N. and J.L.; supervision, J.L.; funding acquisition, J.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This publication has been produced with the support of the Integrated Infrastructure Operational Program for the project: New possibilities for laboratory diagnostics and massive screening of SARS-Cov-2 and identification of mechanisms of virus behaviour in human body, ITMS: 313011AUA4, co-financed by the European Regional Development Fund and grant VEGA No. 230/20 and APVV 15/0107

**Institutional Review Board Statement:** This study was approved by the Ethics Committee of the Jessenius Faculty of Medicine in Martin (registered under IRB00005636 at Office for Human Research Protection, U.S. Department of Health and Human Services) under the code EK82/2020.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The NMR spectra or evaluated data used in this study are available on request from the author: eva.baranovicova@uniba.sk.

**Acknowledgments:** We would like to thank the medical staff from the Clinic of Pneumology and Phthisiology, Martin University Hospital, Slovakia, for their helpfulness during sample collection. This publication was produced with the support of the Integrated Infrastructure Operational Program for the project: New possibilities for laboratory diagnostics and massive screening of SARS-CoV-2 and identification of mechanisms of virus behaviour in human body, ITMS: 313011AUA4, co-financed by the European Regional Development Fund and grant VEGA No. 230/20.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


### *Article* **Effects of Workers Exposure to Nanoparticles Studied by NMR Metabolomics**

**Št ˇepán Horník 1,2,†, Lenka Michálková 1,2,†, Jan Sýkora 1,\*, Vladimír Ždímal 1, Št ˇepánka Vlˇcková 3, Št ˇepánka Dvoˇráckov ˇ á <sup>4</sup> and Daniela Pelclová 3,\***


**Abstract:** In this study, the effects of occupational exposure to nanoparticles (NPs) were studied by NMR metabolomics. Exhaled breath condensate (EBC) and blood plasma samples were obtained from a research nanoparticles-processing unit at a national research university. The samples were taken from three groups of subjects: samples from workers exposed to nanoparticles collected before and after shift, and from controls not exposed to NPs. Altogether, 60 1H NMR spectra of exhaled breath condensate (EBC) samples and 60 1H NMR spectra of blood plasma samples were analysed, 20 in each group. The metabolites identified together with binning data were subjected to multivariate statistical analysis, which provided clear discrimination of the groups studied. Statistically significant metabolites responsible for group separation served as a foundation for analysis of impaired metabolic pathways. It was found that the acute effect of NPs exposure is mainly reflected in the pathways related to the production of antioxidants and other protective species, while the chronic effect is manifested mainly in the alteration of glutamine and glutamate metabolism, and the purine metabolism pathway.

**Keywords:** NMR metabolomics; human plasma; exhaled breath condensate; nanoparticles exposure

#### **1. Introduction**

Nanoscience and nanotechnology have been developing rapidly in recent years, especially in new materials for electronics and optoelectronics fields, for energy technology, and in technology fields related to medical products, particularly for diagnostics and drugs delivery systems. The increased use of nanoparticles has raised concerns in many areas including the environment, human public health, consumer safety, and occupational safety and health [1,2]. Nanoparticles (NPs) are defined as particles with one or more dimensions at the nanoscale, less than 100 nm. The physiological response to NPs and the potential adverse effect on human health requires further research since contact with NPs is becoming a common part of everyday life. In recent years, numerous toxicity studies have assessed the hazard of NPs exposure [2–14]. In general, several health issues were associated with NPs including allergy, injury of epithelial tissue, inflammation, and oxidative stress response [1–3,6,10,11,15,16]. The mechanisms of NPs' biological interaction may vary according to the chemical composition, size, shape, bulk chemical composition, solubility, dose, etc. Moreover, NPs may show an increased toxicity when compared

**Citation:** Horník, Š.; Michálková, L.; Sýkora, J.; Ždímal, V.; Vlˇcková, Š.; Dvoˇráˇcková, Š.; Pelclová, D. Effects of Workers Exposure to Nanoparticles Studied by NMR Metabolomics. *Appl. Sci.* **2021**, *11*, 6601. https://doi.org/ 10.3390/app11146601

Academic Editor: Alessia Vignoli

Received: 7 June 2021 Accepted: 14 July 2021 Published: 18 July 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

to larger particles of the same chemical composition that are little or even non-toxic by themselves [2,10,16,17].

For humans, inhalation is probably the most common way of NPs access followed by oral and dermal routes of exposure. Inhaled NPs can be deposited throughout the human respiratory system including pharyngeal, nasal, transbronchial and alveolar regions, depending on the particle size. The fractional deposition efficiency of particles below 100 nm is in the range of 30–70% in pulmonary regions, and the alveolar deposition increases as the size of NPs decreases [2,16–18]. After deposition in the respiratory tract, NPs may penetrate through membranes and thus enter the blood, pulmonary interstitium, brain, liver, heart, spleen and possibly to the foetus in pregnant females. Since NPs can have the same dimensions as some biomolecules, such as proteins and nucleic acids, adsorption and subsequent disruption of their structure are also possible [2,3,10,11,19].

The existing toxicological methodology for NPs still requires further adjustment to properly assess the risks, including the transport and distribution of NPs in the human body and the mechanism of interaction at the subcellular and molecular level, and to extrapolate the results from in vitro and animal models experiments, which may explain the human health deterioration. Another challenge of this field is to find a fast, specific and sensitive way to evaluate occupational risk. So far, the number of human studies is very limited. As the main exposure to NPs takes place via inhalation and the respiratory system is the primary afflicted organ system, collection and analysis of exhaled breath condensate (EBC) is the most frequently used non-invasive technique for assessment of a subject's condition. EBC contains, besides water, a small proportion of inorganic ions, small organic molecules, proteins and other macromolecules. Analysis of EBC enables the determination of important biomarkers as a response to current physiological conditions [20].

Recently, two toxicological studies were performed on a cohort of 20 workers exposed to NPs during their occupation [12,13]. A detailed analysis of lung function parameters obtained by spirometry revealed a significant decline of forced expiratory volume (FEV1) and its ratio to forced vital capacity (FVC) when compared to the pre-shift values or to the control group. These data were accompanied by LC-MS analysis of inflammation markers in EBC. The levels of pro-inflammatory markers LTB4, LTD4, LTE4, IL 9 and TNF were found to be increased in the worker group relative to controls. On the other hand, the levels of anti-inflammatory LXB4 and IL 10 were lower in the worker group than in controls. Moreover, the levels of the TNF (tumour necrosis factor) found in the pre-shift samples were positively correlated with the duration of employment in the NPs processing workshop [13]. LC-MS analysis was also targeted at markers of oxidative stress. The oxidation of lipids was evaluated from the levels of malondialdehyde (MDA), 4-hydroxy-*trans*-hexenal (HHE), 4-hydroxy-*trans*-nonenal (HNE), C6–C13 aldehydes, and 8-isoprostane; oxidative damage of nucleic acids from levels of 8-hydroxyguanosine (8-OHG), 8-hydroxy-2-deoxyguanosine (8-OHdG), and 5-hydroxymethyl uracil (5-OHMeU); oxidation of proteins from levels of *o*-tyrosine (*o*-Tyr), 3-chlorotyrosine (3-ClTyr), and 3-nitrotyrosine (3-NOTyr). A statistically significant increase was observed for all markers of lipid oxidation in post-shift samples relative to pre-shift ones, while the markers of oxidation of nucleic acids and proteins were found already significantly elevated in the pre-shift EBC samples, and no further increase was observed in the post-shift samples [12]. Both studies suggested lung impairment at the molecular level induced by oxidative stress associated with NPs exposure. However, the adverse effects were attributed rather to NPs in general than to specific chemical composition of NPs.

In this study, the EBC and blood plasma samples of the same cohort were analysed by 1H NMR spectroscopy and processed by means of multivariate statistical analysis. It has already been shown that such NMR-based metabolomics can be advantageously used in NPs toxicology studies [1,11,21–28] reflecting the molecular changes induced by NPs inhalation. The samples studied here were examined as pre-shift and post-shift and were compared to controls. The main goal of this study was to assess the acute and chronic effect of NPs occupational exposure.

#### **2. Materials and Methods**

#### *2.1. Workplace and Process Description*

Subjects of the study were recruited at a research and development unit at a national research university, where a new thermoplastic or reactoplastic composite material was being developed. In the workplace, three different operations are performed, specifically, welding on metal surfaces, smelting of mixtures containing nanoadditives, and machining of the finished nanocomposite. A chemical analysis of aerosol generated in the working environment showed Fe, Mn and Si as the most abundant elements [12]. Aerosol mass concentration ranged from 0.12 to 1.84 mg/m3 during nanocomposite machining processes. Median particle number concentration ranged from 4.8 to 105 × 106 particles/m3 with the particle size ranging from 25 to 860 nm [12].

#### *2.2. Subject Recruitment and Sample Collection*

The samples were collected from 20 nanocomposite workers (15 men, 5 women; age 29–63, average 42 years; 1 smoker, 19 non-smokers) and from 20 control subjects living in the same district but working only in an office without any contact with NPs (13 men, 7 women; age 20–66, average 43 years; 2 smokers, 18 non-smokers).

The EBC and blood plasma samples from nanocomposite workers were collected twice during the workday, pre-shift (i.e., before 2.5 h exposure to NPs) and post-shift (i.e., after NPs exposure). The examinations are referred to as pre-shift and post-shift. Beside the NP exposed workplace, the rest of the 8-hour shift was spent in the office. The controls were examined only once during the same time frame as the workers.

The pre-shift samples were used to study the subacute/chronic effect on the subjects of exposures in previous days. Comparison of the pre-shift and post-shift samples was intended to evaluate the acute effect of exposure during the shift.

All subjects were asked questions from a standardized questionnaire which summarized information on personal and occupational history, medical treatments, dietary habits, smoking habits, and alcohol intake (Table S1, Supplementary Materials). Participants underwent a physical examination, followed by the collection of biological samples—exhaled breath condensate and blood plasma.

This study has been approved by the Ethics Committee of the 1st Medical Faculty, Charles University. All procedures were performed following the Helsinki Declaration and the Collection Law of the Czech Republic. All participants signed an informed consent.

#### *2.3. EBC Collection*

EBC samples were collected using an Ecoscreen Turbo DECCS device (Jaeger, Hochberg, Germany) equipped with a filter. All subjects breathed tidally for 15 min through a mouthpiece connected to the condenser (−20 ◦C) while wearing a nose-clip. A minimum volume of exhaled air of 120 L was monitored via the EcoVent device (Jaeger, Wurzburg, Germany). The sample collection took approximately 15 min. All samples were immediately frozen and stored at −80 ◦C.

#### *2.4. Blood Plasma Collection*

Venous blood (9 mL) from the subjects studied was collected using sterile blood collection tubes with heparin as an anticoagulant. The plasma fractions were obtained by centrifugation at 15,000×*g* for 10 min and immediately frozen and stored at −80 ◦C.

For more details on the subjects' cohort, working environment, analysis of NPs composition and properties, see previous publications [12,13]. A follow-up of the researchers in 2017 and 2018 confirmed the results from 2016 [29].

#### *2.5. Sample Preparation*

Samples were thawed at room temperature. For preparation of EBC and blood plasma samples for 1H NMR analysis, the following operation procedures were determined.

#### *2.6. EBC Sample Preparation*

An aliquot of 500 μL of EBC was mixed with 100 μL phosphate buffer (0.1 mol/L1, pH = 7.4, 0.1 mol/L<sup>1</sup> sodium salt of trimethylsilyl-2,2,3,3-d4-propionic acid (TSP), 38 mmol/L<sup>1</sup> NaN3). Thus, sufficient sample volume for NMR analysis was obtained and pH was adjusted to 7.7.

#### *2.7. Blood Plasma Sample Preparation*

Aliquots of 350 μL of blood plasma were centrifuged through an Amicon 3-kDa cut-off filter (Merck, Germany) for 30 min at 14,000 rpm to isolate low-molecular metabolites. Subsequently, the filtrate was mixed with 350 μL phosphate buffer in D2O (0.1 mol/L1, pH = 7.4, 0.1 mol/L<sup>1</sup> sodium salt of trimethylsilyl-2,2,3,3-d4-propionic acid (TSP), 38 mmol/L<sup>1</sup> NaN3). Thus, sufficient sample volume for NMR analysis was obtained and pH was adjusted to 7.4.

#### *2.8. Acquisition*

One dimensional proton NMR spectra for all EBC and plasma samples were acquired using a Varian INOVA 500 MHz spectrometer (Varian Instruments Inc., Palo Alto, CA, USA) operating at 499.87 MHz, equipped with Ultra Shim System II. A 5 mm probe with inner 1H coil was used to maximize the sensitivity. Prior to the analysis, samples were kept for at least 10 min inside the NMR probe for temperature equilibration (298.15 K). The 1H NMR spectra of EBC and plasma samples were obtained using wet1D and tnnoesy pulse sequence, respectively. Spectral width covered 8 kHz using 2.7 s acquisition time. A relaxation delay of 4 s and 2 s was used for EBC and plasma samples, respectively. The final spectrum resulted from an accumulation of 1000 scans. Representative 1H NMR spectra can be found in Figures S1 and S2 in the Supplementary Materials.

#### *2.9. Data Processing*

The Fourier-transform spectra were manually corrected for phase and baseline distortions using Chenomx NMR Suite 8.0 (NMR Suite program, Edmonton, Alberta, Canada [30]). The experimental spectrum was referenced to TSP. The solvent signal residuum was subtracted, TSP signal linewidth was determined, and pH was set.

Compound profiling was performed in the Chenomx Profiler by precise fitting of the compounds from the Chenomx library to the experimental spectrum. In EBC samples, 15 metabolites were identified, while 58 metabolites were identified in blood plasma samples. Since only 15 metabolites were quantified in the EBC, binning was used for EBC spectra to obtain more variables per sample. The binning was applied to each spectrum in the range 0.7–8.6 ppm, except for the region containing residual water signal (4.1–5.6 ppm). Standard bin size of 0.02 ppm was used, yielding 320 bins.

The concentration data from plasma samples were normalized to the total concentration sum to reduce the effects of sample dilution prior to statistical analysis. Total area normalization works well in biofluids, in which overall concentrations of metabolites are almost constant among the samples, such as blood plasma or urine [31]. However, normalization to the total area is not recommended in the case of EBC samples because of large differences in dilution [32]. Hence, PQN normalization was used for the EBC samples as a more robust type of normalization [33].

#### *2.10. Statistical Analyses*

All data analyses were performed using the open-source software R [34] and Metaboanalyst 5.0 [35]. Multivariate data analyses were conducted on processed concentration data and binned data separately. As a first step of statistical analysis, principal component analysis (PCA) was used to provide preliminary insight on the data complexity, trends of grouping or identifying outliers. Subsequently, orthogonal partial least squares discriminant analysis (OPLS-DA) was used for sample classification. Multilevel partial least squares analysis (mPLS) was used in the case of comparison of the pre-shift and post-shift

samples [36]. All reported values of accuracy, sensitivity, and specificity were assessed by means of 100 cycles of a Monte Carlo cross-validation scheme where 90% of the samples were randomly selected at each iteration as a training set to build the model; the remaining 10% were subsequently tested on performance characteristics for the classification.

In order to identify the most influential and statistically significant compounds, the Wilcoxon rank-sum test and its paired version, the Wilcoxon signed-rank test, were used. Obtained *p*-values were adjusted for multiple comparisons using the Benjamini and Hochberg correction [37]. The threshold of adjusted *p*-values was set to <0.05 for statistical significance. Fold change was performed following the general formula defined as a logarithm of base 2 of a division of a median concentration of an individual compound in one group by a median concentration of an individual compound in the other group. The result is projected in logarithm to base 2 scale.

Altered metabolic pathways were detected using Metaboanalyst 5.0 using the metabolite ID taken from the Human Metabolome Database. Metabolic pathway analysis was performed on blood plasma metabolic profiles to reveal the biological impact of NPs inhalation. A plot of affected pathways contained 43 nodes, each representing one pathway, with colour and size coding corresponding to pathway significance and its impact, respectively. The significance was generated from betweenness centrality and out-degree centrality measurements. The pathway impact was generated by the summation of importance measures of matched metabolites to all metabolites present within the pathway.

#### **3. Results and Discussion**

In this study, 1H NMR spectra of 60 exhaled breath condensate (EBC) samples and 60 blood plasma samples were analysed. The samples originate from a research nanoparticles-processing unit at a national research university. The samples were taken from three groups of subjects: (i) samples from workers exposed to nanoparticles (NPs) collected before shift (pre-shift, 20 EBC and 20 blood plasma) and (ii) after shift (post-shift, 20 EBC and 20 blood plasma), and (iii) a control group of subjects not exposed to NPs (controls, 20 EBC and 20 blood plasma). The pre-shift and post-shift samples were collected from the same individuals. Individual groups are defined in Materials and Methods. A comparative study of the pre-shift and control samples was applied to reveal a subacute/chronic effect of NPs exposure, while the comparison of the pre-shift and post-shift samples should reflect the acute effect on the workers' health.

#### *3.1. Exhaled Breath Condensate*

Since exhaled breath condensate is composed of 99.9% water, the other constituents are rather diluted. For this reason, quantitative analysis of 1H NMR spectra using the Chenomx reference library provided only 15 metabolites. Due to the limited number of metabolites quantified, a meaningful multivariate statistical analysis cannot be performed on such a dataset. However, univariate statistical analysis identified some of the metabolites as statistically significant for discrimination of the groups studied. A Wilcoxon rank-sum test showed that pre-shift and post-shift EBC samples are mainly characterized by significantly elevated levels of acetoin and propionate, and decreased acetone, isopropanol and lactate levels when compared to control samples (Table S2, Supplementary Materials). On the other hand, an increase in dimethylamine and decrease in acetoin are the most significant changes induced by NPs exposure as observed in comparison of pre-shift and post-shift EBC samples (Figure 1).

Final group discrimination analysis was performed using binning data. Fingerprinting of individual 1H NMR spectra provided 320 bins which subsequently served as an input into multivariate statistical analysis. Principal component analysis (PCA) of all binned spectra did not show any significantly outlying sample. It also indicated certain trends in group separation; however, a clear discrimination was not achieved (Figure S3, Supplementary Materials). Satisfactory group separation was achieved by orthogonal partial least squares discriminant analysis (OPLS-DA), which was applied to the pre-shift and

control group to reveal the chronic effect of NPs exposure and to the pre-shift/post-shift and post-shift/control group to uncover the acute effect.

**Figure 1.** Fold change projections depicting differences in levels of individual metabolites observed in EBC samples between individual groups.

An excellent separation between the pre-shift and control group was achieved using three components. The model was characterized by 81.4% sensitivity, 94.8% specificity and 88.1% accuracy after Monte Carlo cross-validation (Figure 2a). Similarly, the separation of post-shift and the control group was achieved using a seven-component model yielding 88.7% accuracy, 93.9% sensitivity and 83.5% specificity after Monte Carlo cross-validation (Figure 2b).

**Figure 2.** OPLS-DA of pre-shift subjects (yellow circles) and healthy controls (blue diamonds) using 320 bins from EBC samples; Acc. 88.1%, Sen. 81.4%, Spe. 94.8% (**a**). OPLS-DA of post-shift subjects (red squares) and healthy controls (blue diamonds) using 320 bins from EBC samples; Acc. 88.7%, Sen. 93.9%, Spe. 83.5% (**b**).

The bins contributing significantly to the group separation were identified from OPLS-DA loadings provided by Metaboanalyst. These bins show increased EBC concentration of acetoin, acetate and propionate in the pre-shift and post-shift samples when compared to the controls. Mainly increased signal intensities of alcohols were found in controls. These findings correspond well with the statistically significant compounds identified by univariate statistics as discussed above (Figure 1).

The comparison of the pre-shift and post-shift groups should reveal the acute effect of NPs exposure. The performed OPLS-DA provided a very good discrimination of the two groups using six components with accuracy of 83.1%, sensitivity of 84.1% and specificity of 82.1% after Monte Carlo cross-validation (Figure 3a). As both groups consist of the same 20 subjects whose samples were collected before and after the shift, a pairwise multilevel partial least squares (mPLS) analysis can be applied [36]. Compared to other PLS analyses, mPLS does not focus on investigation of the studied groups as a whole, but rather observes changes in each individual before and after the stimulus of the change and reflects the changes occurring within the same subject. The mPLS analysis showed a satisfactory discrimination of the two groups using three components with 82.0% accuracy after Monte Carlo cross-validation (Figure 3b). Although the OPLS and the mPLS models show similar accuracy, the mPLS model requires fewer components.

**Figure 3.** OPLS-DA of pre-shift (yellow circles) and post-shift subjects (red squares); Acc. 83.1%, Sen. 84.1%, Spe. 82.1% (**a**). Multilevel partial least squares (mPLS) analysis of pre-shift (yellow circles) and post-shift subjects (red squares); Acc. 82% (**b**). Both using 320 bins in each EBC sample.

The bins responsible for the group separation correspond to acetoin, which was found increased in the pre-shift group, and to lactate, formate and unsaturated chains of higher carboxylic acids increased in the post-shift group. This is in agreement with the statistically significant compounds identified by univariate statistics (Figure 1).

Acetoin is a commonly identified metabolite in EBC [38–40] as a product of the detoxification process of acetaldehyde [41].

Since dimethylamine was found increased only in the post-shift group, it is probable that it may be associated with the acute effect of NPs exposure.

The increased levels of short-chain fatty acids such as acetate, propionate and butyrate in NPs exposed groups in comparison to the control group could be attributed to involvement in the regulation of several leukocyte functions such as eicosanoids and cytokines/chemokines production [38]. Propionate is associated with lipid metabolism [39], which was also found affected by chronic exposure to NPs [12]. Boxplots of selected metabolites affected by NPs exposure are depicted in Figure 4.

**Figure 4.** Boxplots of selected metabolites affected by NPs exposure.

#### *3.2. Analysis of Blood Plasma*

Using the Chenomx reference library, 58 metabolites were identified and quantified in each 1H NMR spectrum of blood plasma samples. The concentration data of all quantified metabolites were used as an input for both multivariate and univariate statistical analyses to reveal important features of each group. The homogeneity of the groups was tested by principal component analysis (PCA) as an unsupervised statistical method. According to PCA, no sample was found significantly outlying. Nevertheless, group discrimination was not achieved (Figure S4, Supplementary Materials).

Subsequently, a supervised statistical method (OPLS-DA) was employed to pre-shift and control samples. A very good separation between these two groups was achieved using three components. The model was characterized by 88.2% sensitivity, 73.2% specificity and 80.7% accuracy after Monte Carlo cross-validation (Figure 5a).

**Figure 5.** OPLS-DA of healthy controls (blue diamonds) and pre-shift subjects (yellow circles); Acc. 80.7%, Sen. 88.2%, Spe. 73.2% (**a**). OPLS-DA of post-shift subjects (red squares) and healthy controls (blue diamonds); Acc. 86.0%, Sen. 86.4%, Spe. 85.7% (**b**). Both using 58 normalized metabolites from blood plasma samples.

The nonparametric Wilcoxon rank-sum test was used to reveal statistically significant compounds that should reflect the effect of chronic exposure to NPs. Only acetone was found under the threshold for statistical significance (adjusted *p*-value ≤ 0.05). Four other metabolites were close to this threshold, specifically glutamate, glutamine, cystine and hypoxanthine (Table S3 and Figure S5 in Supplementary Materials). Levels of acetone, glutamine and cystine were found increased in the control group, whereas glutamate and hypoxanthine show higher levels in the pre-shift group.

OPLS-DA was also used for differentiation between the post-shift subjects and the healthy controls. A very good separation of the two groups was obtained using a sixcomponent model with an accuracy of 86.0%, sensitivity of 86.4% and specificity of 85.7% after Monte Carlo cross-validation (Figure 5b). The nonparametric Wilcoxon rank-sum test was used to reveal statistically significant compounds (Table S3 and Figure S6 in Supplementary Materials). Seven metabolites were found under the threshold for statistical significance (adjusted *p*-value ≤ 0.05). Levels of propylene glycol, glutamate and pyruvate were found increased in the post-shift group, whereas acetone, mannose, 2-oxoisocaproate and *O*-acetylcarnitine showed higher levels in the control group.

Analogically, the acute effect of NPs exposure was also studied on plasma samples of the pre- and post-shift groups using OPLS-DA (Figure 6a). This discrimination analysis showed a certain potential to distinguish between the two groups with a model based on eight components characterized by 75.4% accuracy, 74.0% sensitivity and 76.9% specificity. Subsequently, mPLS was performed with a remarkable discrimination of the two groups using five components with 89.0% accuracy after Monte Carlo cross-validation (Figure 6b). In this case, mPLS reflects the intra-individual differences within each subject; therefore, it provides better group separation than discrimination analysis based on OPLS.

Subsequent analysis by the nonparametric pairwise Wilcoxon signed-rank test revealed 11 statistically significant compounds (Table S3 and Figure S7 in Supplementary Materials). Only compounds with adjusted *p*-value ≤ 0.05 after Benjamini-Hochberg correction were deemed statistically significant. Out of the 11 statistically significant compounds, increased levels for eight metabolites were found in the pre-shift group, specifically isobutyrate, 2-hydroxybutyrate, 2-oxoisocaproate, lactate, 3-hydroxybutyrate, isopropanol, tryptophan and 3-methyl-2-oxovalerate. Levels of three statistically significant metabolites were elevated in post-shift groups, namely propylene glycol, glycolate and myo-inositol.

The stress induced by NPs exposure is well documented by increased levels of certain metabolites in post-shift samples when compared to the pre-shift samples or controls. In particular, increased levels were found for propylene glycol, glycolate, myo-inositol, pyruvate and glutamate (Figure 7). Propylene glycol is known to be predominantly of exogenous origin as a part of various vitamins and other dietary supplements. The

increased concentrations of propylene glycol in the post-shift group corresponds to the increased consumption of such supportive products in the morning before the shift.

**Figure 6.** OPLS-DA of pre-shift (yellow circles) and post-shift subjects (red squares); Acc. 75.4%, Sen. 74.0%, Spe. 76.9% (**a**). mPLS analysis of pre-shift (yellow circles) and post-shift subjects (red squares); Acc. 89% (**b**). Both using 58 normalized metabolites from blood plasma samples.

**Figure 7.** Boxplots of metabolites with increased levels in the post-shift samples.

On the other hand, the NPs exposure also induced a depletion of other metabolites in post-shift samples; namely of 3-methyl-2-oxovalerate, 2-oxoisocaproate, 2-hydroxybutyrate, 3-hydroxybutyrate, isobutyrate, isopropanol, mannose, *O*-acetylcarnitine and tryptophan (Figure 8). All the changes found in the post-shift group can be attributed to the acute effect of the NPs on workers' health.

The long-term effect of the NPs on workers' health can be deduced from the simultaneous changes in the pre- and post-shift group when compared to the control group. The levels of acetone, glutamine and cystine were found to decrease, while the levels of lactate and hypoxanthine increased in both groups when compared to the controls (Figure 9). The changes in levels of lactate and hypoxanthine were found to be more pronounced in the pre-shift group, which indicates involvement of these compounds in several metabolic pathways and mixing of acute and chronic effects.

**Figure 9.** Boxplots of metabolites showing simultaneous changes in the pre- and post-shift group in comparison to the controls.

Metabolic pathway analysis was performed in MetaboAnalyst to assess involvement of the statistically significant metabolites in the individual metabolic pathways. The alterations between individual groups found by the pathway analysis are depicted in Figure S8 (Supplementary Materials) and the most affected pathways are summarized in Table S4 (Supplementary Materials).

The increased levels of lactate found in the pre-shift group can be partially associated with the metabolism of propylene glycol, which is contained in the food supplements administrated before the shift, as mentioned above. On the other hand, lactate is also involved in several other metabolic pathways, including pyruvate metabolism, according to pathway analysis (Figure S6 in Supplementary Materials) or glucose-alanine metabolism, according to the literature [42]. These pathways play an important role as energy pathways where lactate is usually produced from pyruvate. The highest concentrations of pyruvate were found in the post-shift group, indicating that the energy pathways were affected and the transformation between lactate and pyruvate is impacted by the NPs exposure. Increased levels of lactate have been observed in several studies on the impact of NPs exposure to rats [21,43,44]. Additionally, decreased levels of mannose, which can serve as an additional energy source, were observed in the post-shift group [45].

Elevated levels of lactate can lead to metabolic acidosis, similarly to increased levels of glycolate, which were found in the post-shift group. The formation of acid metabolites can induce inhibition of other metabolic pathways [46]. Glycolate is mainly involved in glyoxylate metabolism where it is oxidized to glyoxylate, which is further transformed into glycine [47]. Glyoxylate can be also transformed into oxalate, which is then caught and secreted by the renal tubules. Excessive concentrations of oxalate cause urolithiasis and nephrocalcinosis [48]. The excessive oxidation of glyoxylate to oxalate by lactate dehydrogenase is prevented by reduction of cytosolic and mitochondrial glyoxylate to glycolate by cytosolic glyoxylate reductase [48].

The post-shift group also manifested increased concentrations of myo-inositol. This metabolite has an important osmoregulatory role and is involved in the running of a wide range of cell functions, including cell growth and survival [49,50], which could explain why myo-inositol is increased in the acute state. Several studies have reported alterations in the myo-inositol levels after exposure to NPs in rats [43] or mouse fibroblast cells L929 [51].

The decreased tryptophan concentration in the post-shift group could be explained by its transformation to kynurenic acid via the kynurenine pathway (pathway tryptophan metabolism in Figure S8c in Supplementary Materials). Kynurenic acid has a protective effect against oxidative stress and lung inflammation induced by exposure to NPs [52]. Furthermore, the kynurenine pathway was previously associated with the elevated levels of cytokines [53], which is consistent with the results of our previous study [12,13]. A decreased concentration of tryptophan was also observed in rat blood serum after exposure to TiO2 NPs [52], which is in agreement with the findings in workers exposed to nanoTiO2, and similar oxidative stress effects [54,55].

The levels of several metabolites associated with the synthesis of glutathione were found altered, namely cystine, glutamate and 2-hydroxybutyrate. Glutathione as a major antioxidant is synthesized from cysteine, glutamate and glycine. Cysteine is transformed to glutathione in response to oxidative stress. This is reflected in decreased levels of cystine, an oxidized dimer form of cysteine. The availability of cysteine was reported as the rate-limiting step in the glutathione synthesis where cysteine is supplied via the cystineglutamate antiporter system [49,56]. Elevated levels of glutamate in the pre- and post-shift group indicate that glutamate is exchanged for cystine in the antiporter system [49]. The elevated levels of glutamate also affect several other metabolic pathways, such as glutamine and glutamate metabolism, alanine, aspartate and glutamate metabolism, arginine and proline metabolism, histidine metabolism and butanoate metabolism, as is shown in the metabolic pathway analysis via MetaboAnalyst (Figure S8a,b in Supplementary Materials). Alterations in glutamate levels have also been observed in several studies focusing on NPs' impact on rats [26,43,51].

The utilization of glutamate in glutathione biosynthesis leads to higher demands on glutamate and increases glutamine transformation into glutamate via the glutamine and glutamate metabolism pathway. This is documented by decreased levels of glutamine in both pre- and post-shift groups compared to the control group. Together with glutamate, glutamine is also involved in the pathway of alanine, aspartate and glutamate metabolism, which was also found altered according to metabolic pathway analysis (Figure S8a,b in Supplementary Materials). Glutamine and glutamate were also found affected in the study of Kitchin et al. [57], in which the effect of TiO2 and CeO2 nanomaterials on human liver HepG2 cells was examined. A significant decrease in glutamine was observed, similar to observations from our study. Nevertheless, a decrease in glutamate was observed, on the contrary to our study, indicating that glutamate was involved at least partially in a different way.

2-Hydroxybutyrate is a reduction product of 2-ketobutyrate, which is produced during the transformation of cystathionine to cysteine within the methionine degradation pathway [58]. Since the concentration of 2-hydroxybutyrate is decreased in the acute state, 2-ketobutyrate is probably transformed into other metabolites, including propionyl-CoA [59], which is also associated with degradation of branched-chain amino acids, as discussed below.

The concentrations of two ketone bodies metabolites, 3-hydroxybutyrate and acetone, were found altered. Both compounds are closely connected to acetoacetate, another ketone body, which was, however, found unaltered. Nevertheless, decreased concentrations of 3-hydroxybutyrate and acetone after exposure to NPs suggest that the metabolic pathway of ketone body metabolism is affected. This was also revealed in the metabolic pathway analysis via MetaboAnalyst (Figure S8b in Supplementary Materials). 3-Hydroxybutyrate is also involved in butanoate metabolism, and it is also a degradation product of branched-chain amino acids, mainly of leucine [60]. The decreased concentrations of 3-hydroxybutyrate and acetone after NPs exposure are in contrast with other studies, which reported elevated levels of 3-hydroxybutyrate [21,23,61]. However, these studies were performed on rats exposed to high NPs doses. 3-Hydroxybutyrate is also an end product of β-oxidation of fatty acids [21]. The impairment of this metabolic pathway is also reflected in a decreased concentration of *O*-acetylcarnitine. This molecule serves as a carrier of acetyl from acetyl-CoA derived from fatty acids to mitochondria [62], thus taking part in energy metabolism. Moreover, decreased levels of *O*-acetylcarnitine can also be associated with oxidative stress. Similarly to this study, decreased levels of *O*-acetylcarnitine were also observed in zebrafish and mice embryos after Fe2O3 NPs exposure [62]. Accordingly, workers exposed to NPs during iron oxide pigment production showed elevated markers of lipid, nucleic acid, and protein oxidation in their EBC [63].

Decreased levels of 3-methyl-2-oxovalerate and 2-oxoisocaproate (4-methyl-2-oxoval erate) were found in the post-shift group. These compounds are produced as direct metabolites of isoleucine and leucine during their degradation by branched-chain amino acid aminotransferase [64]. Since the concentrations of leucine and isoleucine were found almost unaffected, leucine and isoleucine are involved in other pathways or processes, and the direct degradation pathway of these amino acids is inhibited. Isobutyrate is another metabolite associated with the metabolism of branched-chain amino acids, mainly of valine [65]. The decreased concentration of isobutyrate in the post-shift group also indicates that branched-chain amino acids' degradation is inhibited in the acute state. However, this inhibition was not manifested in the metabolic pathway analysis performed in MetaboAnalyst.

Hypoxanthine is an important part of purine metabolism [66], thus the elevated levels of hypoxanthine found mainly in pre-shift plasma samples indicate alterations in this metabolism. Such an observation is complementary to the findings of a previous study performed on the same cohort, where the markers of nucleic acids' oxidation (8-hydroxyguanosine and 8-hydroxy-2-deoxyguanosine) were identified in pre-shift EBC

samples. Similar markers were also found in other toxicological studies of occupational exposure to different NPs, indicating a general effect of chronic NPs exposure [12,54,55].

The metabolic pathway analysis performed mainly indicates that the induced oxidative stress activates anti-oxidative pathways, and antioxidants, such as glutathione, are extensively consumed. Higher demands in the supplement of the consumed antioxidants can be observed in decreasing levels of their intermediates, in particular glutamine and cystine. The decreased tryptophan levels may be related to the production of its metabolites like kynurenic acid, which have protective effects against oxidative stress and lung inflammation. Moreover, alterations of several other metabolic pathways were observed.

The changes induced in metabolic profiles by NPs exposure were associated predominantly to the organism's response to oxidative stress. Similar response has been observed in studies dedicated to evaluation of the oxidative stress induced by smoking. Despite the number of smoking subjects being rather small in the presented study, statistical analysis was also performed after exclusion of the smoking subjects. One subject was excluded from the pre-shift/post-shift group and two subjects from the control group. The obtained results were in correspondence with those found in the original study and only minor changes were observed. Adjusted *p*-values of several metabolites levels previously found as statically significant raised slightly above the designated threshold. On the other hand, the adjusted *p*-value of hypoxanthine in blood plasma descended below the threshold in the comparison of the pre-shift and control groups (Tables S5 and S6 in Supplementary Materials). It is worth noting that the changes found in the results of univariate statistical analysis can be partially attributed to the decreased number of samples. The group separation provided by multivariate statistical analyses remained unaffected (Figures S9–S12 in Supplementary Materials). Major limitation of this study is the small number of subjects reflecting the actual size of the workplace, as all available workers were included.

#### **4. Conclusions**

The EBC and blood plasma samples of a cohort of 20 workers exposed to NPs during their occupation were analysed by 1H NMR spectroscopy and processed by statistical analysis. Altogether, 15 metabolites were identified in EBC samples, while the analysis of plasma samples provided 58 metabolites. Subsequent multivariate statistical analyses performed on binning data from EBC and concentrations of 58 metabolites from plasma samples enabled clear discrimination between the pre-shift, post-shift and control groups. The univariate statistical analysis revealed statistically significant metabolites. Although plasma and EBC samples each showed changes in levels of different metabolites, the metabolic pathway analysis indicated, in both cases, mainly a reaction of the organism to oxidative stress and subsequent efforts for its protection.

The comparison of the pre-shift and post-shift group accompanied by comparison of the post-shift and control group provided insight into the acute effect of the NPs exposure. Altered levels of lactate, pyruvate, 3-hydroxybutyrate, mannose and *O*-acetylcarnithine indicated an energy balance impairment. The altered levels of glutamate, cystine, tryptophan, acetate, propionate and butyrate were associated to the pathways related to the production of antioxidants, mainly glutathione, and other protective species. The comparison of the pre-shift and control group revealed that the chronic effect of the NPs exposure manifested mainly in an alteration in glutamine and glutamate metabolism. The increased levels of hypoxanthine indicated an impairment of the purine metabolism pathway.

The presented results correspond well with similar studies performed on cohorts exposed to different types of NPs, indicating that the observed adverse effects can be attributed to nanoparticles in general, rather than to their chemical nature.

This work is one of the few dealing with the occupational exposure to NPs studied by the means of NMR metabolomics. Similar response to NPs exposure was observed for both types of samples indicating that either biofluid can be used for evaluation of adverse effects of nanoparticles inhalation. Potentially, blood derivatives could serve as an alternative to commonly used EBC samples.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/app11146601/s1, Table S1: Basic characteristics of the samples; Figure S1: 1H NMR spectrum of a representative EBC sample; Figure S2: 1H NMR spectrum of a representative blood plasma sample; Table S2: Wilcoxon test for EBC samples; Figure S3: Principal component analysis for EBC samples; Figure S4: Principal component analysis for plasma samples; Table S3: Wilcoxon test for blood plasma samples; Figure S5: Fold change projection of pre-shift subjects and healthy controls; Figure S6: Fold change projection of post-shift subjects and healthy controls; Figure S7: Fold change projection of pre-shift and post-shift subjects; Figure S8: Metabolic pathway analysis; Table S4: Overview of the most influenced metabolic pathways; Table S5: Wilcoxon test for EBC samples after exclusion of smoking subjects; Table S6: Wilcoxon test for blood plasma samples after exclusion of smoking subjects; Figures S9 and S10: Multivariate statistical analysis of EBC samples after exclusion of smoking subjects; Figures S11 and S12: Multivariate statistical analysis of plasma samples after exclusion of smoking subjects.

**Author Contributions:** Conceptualization, V.Ž., Š.D. and D.P.; methodology, Š.H., L.M. and J.S.; formal analysis, J.S. and D.P.; investigation, Š.H. and L.M.; resources, J.S., V.Ž. and D.P.; data curation, Š.H., L.M., J.S. and Š.V.; writing—original draft preparation, Š.H. and L.M.; writing—review and editing, J.S.; visualization, Š.H., L.M. and J.S.; supervision, J.S.; project administration, Š.V. and D.P.; funding acquisition, J.S., V.Ž. and D.P. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the Ministry of Youth, Education and Sports of the Czech Republic, project No. LM2018122 and by the Technology Agency of the Czech Republic (Grant No. TK02010035) and projects Progres Q25 and Q29 of the Charles University.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethical Committee of 1st Medical Faculty, Charles University.

**Informed Consent Statement:** All participants were informed of the study aim at least five days earlier, and signed an informed consent form before the study began.

**Data Availability Statement:** The NMR spectra or evaluated data used in this study are available on request from the author: sykora@icpf.cas.cz.

**Acknowledgments:** The authors are grateful to Andrew Christensen for proofreading.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

