Mechanisms linking metabolism of Helicobacter pylori to ¹⁸O and ¹³C-isotopes of human breath CO₂

Abstract

The gastric pathogen Helicobacter pylori utilize glucose during metabolism, but the underlying mechanisms linking to oxygen-18 (¹⁸O) and carbon-13 (¹³C)-isotopic fractionations of breath CO₂ during glucose metabolism are poorly understood. Using the excretion dynamics of ¹⁸O/¹⁶O and ¹³C/¹²C-isotope ratios of breath CO₂, we found that individuals with Helicobacter pylori infections exhibited significantly higher isotopic enrichments of ¹⁸O in breath CO₂ during the 2h-glucose metabolism regardless of the isotopic nature of the substrate, while no significant enrichments of ¹⁸O in breath CO₂ were manifested in individuals without the infections. In contrast, the ¹³C-isotopic enrichments of breath CO₂ were significantly higher in individuals with Helicobacter pylori compared to individuals without infections in response to ¹³C-enriched glucose uptake, whereas a distinguishable change of breath ¹³C/¹²C-isotope ratios was also evident when Helicobacter pylori utilize natural glucose. Moreover, monitoring the ¹⁸O and ¹³C-isotopic exchange in breath CO₂ successfully diagnosed the eradications of Helicobacter pylori infections following a standard therapy. Our findings suggest that breath ¹²C¹⁸O¹⁶O and ¹³C¹⁶O¹⁶O can be used as potential molecular biomarkers to distinctively track the pathogenesis of Helicobacter pylori and also for eradication purposes and thus may open new perspectives into the pathogen’s physiology along with isotope-specific non-invasive diagnosis of the infection.

Introduction

Helicobacter pylori (H. pylori) is a micro-aerophilic pathogen, which is known to be able to colonize the mucosal surfaces of the human stomach, where it gives rise to chronic gastritis, peptic ulcers^1,2,3 and is closely linked to the development of certain types of gastric cancer⁴. The gastric pathogen H. pylori uses glucose as the primary energy substrate^5,6, although the overall metabolism of H. pylori yet remains inadequately understood. Some early evidences, however, suggest that H. pylori has the ability to utilize glucose for metabolism through a glucokinase activity⁷ and enzymes of the pentose phosphate and glycolysis pathways^8,9. Carbon dioxide (CO₂) is usually produced as a by-product of glucose catabolism which is then transported to the lungs through the blood stream and finally it is excreted in human breath. However, the precise role of glucose metabolism, especially in the pathogenesis of the H. pylori infection is not currently known. A new insight into the role of glucose metabolism is essential to elucidate the pathophysiology of H. pylori for its successful colonization of the gastrointestinal tract. However, to our knowledge, so far there have been no studies focused on glucose uptake for individuals harboring H. pylori infections, exhibiting the time-dependent excretion dynamics of the metabolite CO₂ in exhaled breath. The purpose of this study was therefore, primarily to explore the potential links between breath CO₂ and H. pylori infections in response to unlabelled and labelled ¹³C-enriched glucose metabolism. A complete understanding of glucose metabolism during the H. pylori infection could be of significance in the development of novel therapies for the micro-organism alongside new and better approaches for treating the most common human bacterial infection.

Furthermore, an earlier study revealed that the oxygen-16 (¹⁶O) and the oxygen-18 (¹⁸O) isotopes in ¹²C¹⁶O₂ and water (H₂¹⁸O), respectively, are rapidly interchanged during the human respiration process mediated by the metalloenzyme carbonic anhydrase (CA)^10,11. It is also known that H. pylori encodes two different forms of the metalloenzyme carbonic anhydrase (α-CA and β-CA)¹² and this gastric pathogen plays a vital role in inter-conversion of carbon dioxide and bicarbonate (CO₂ + H₂O↔H⁺ + HCO₃⁻), catalyzed by the CA activity^12,13,14. This efficient activity suggests that investigations of breath ¹⁸O/¹⁶O isotopic fractionations of CO₂ may specifically track the gastric pathogen H. pylori and hence may introduce a novel non-invasive strategy in the diagnosis of H. pylori infections living in human stomach. As a consequence, we hypothesized that simultaneous monitoring the ¹⁸O/¹⁶O and ¹³C/¹²C stable isotope ratios of exhaled breath CO₂ associated with glucose metabolism in H. pylori may act as potential markers for the early detection of H. pylori infections or during the preclinical phase of the infections. In view of the fact that H. pylori is able to uptake and metabolize glucose confirmed as experimentally¹⁵ and also by analysing the genome sequence5 therefore, there is a pressing need to assess the clinical efficacy of the glucose utilization by H. pylori for large-scale screening individuals harboring the micro-organism. In addition, unravelling the precise metabolic pathways involved in causing the isotopic fractionations of ¹²C¹⁶O¹⁶O, ¹³C¹⁶O¹⁶O and ¹²C¹⁸O¹⁶O in human breath during the glucose uptake by H. pylori remains a major challenge, whenever an individual is at-risk of developing the disease.

In this study, we first report, the potential links of both ¹⁸O and ¹³C-stable isotopes of breath CO₂ with the gastric pathogen H. pylori in response to glucose ingestion. We investigated simultaneously the time-dependent excretion dynamics of the ¹²C¹⁸O¹⁶O/¹²C¹⁶O¹⁶O and ¹³C¹⁶O¹⁶O/¹²C¹⁶O¹⁶O isotope ratios of breath CO₂ from individuals with H. pylori positive and negative by employing a laser-based integrated cavity output spectroscopy (ICOS) method. We further explored the potential metabolic pathways underlying the glucose utilization in the pathogenesis of H. pylori infection and the mechanisms linking breath oxygen-18 and carbon-13 isotopic fractionations of CO₂ to the gastric pathogen H. pylori. Finally, we determined various diagnostic parameters such as optimal diagnostic cut-off values, diagnostic sensitivity and specificity of oxygen-18 and carbon-13 stable isotopes in breath CO₂ to gain a better insight into the diagnostic efficiency for the non-invasive detection of H. pylori infection in real-time.

Results and Discussion

To investigate the ¹⁸O and ¹³C isotopic fractionations of breath CO₂, we first studied the time-dependent excretion dynamics of both isotopes in exhaled breath after ingestion of an oral dose of ¹³C-enriched glucose for H. pylori positive (n = 72) and negative (n = 55) individuals, using a laser-based high-precision cavity-enhanced integrated cavity output spectrometer (ICOS). We explored the isotopic fractionation of CO₂ by simultaneous monitoring the ¹⁸O/¹⁶O and ¹³C/¹²C stable isotope ratios in breath, expressed as delta-over-baseline (DOB) relative to the Vienna Pee Dee Belemnite standard, i.e., δ_DOB¹⁸O‰ = [(δ¹⁸O‰)_t=t – (δ¹⁸O‰)_t=basal] and δ_DOB¹³C‰ = [(δ¹³C‰)_t=t – (δ¹³C‰)_t=basal], respectively, associated with glucose metabolism. In this investigation (Fig. 1a), individuals with H. pylori positive exhibited significantly higher isotopic enrichments of ¹⁸O in CO₂ compared with H. pylori negative during the 2h-glucose metabolism, while no significant enrichments of ¹⁸O in CO₂ were manifested in individuals without H. pylori infections. These findings suggest a potential link between H. pylori infections in human stomach and the ¹⁸O-isotopic fractionations in exhaled breath and hence may open a new route for the non-invasive assessment of H. pylori infections. Carbonic anhydrase (CA) activity of H. pylori has previously been proposed to interchange the oxygen isotopes of CO₂ (¹⁶O) and H₂O (¹⁸O) efficiently^10,11, suggesting that in our observations CA activity may play an important role in oxygen-isotope fractionations of breath CO₂ in the glucose-mediated bacterial environment. Hence a statistically significant difference in the δ_DOB¹⁸O‰ values in excretion dynamics established a marked distinction (Fig. 1b) between H. pylori infected and non-infected individuals. Taken together, these findings indicate that the monitoring of ¹⁸O-exchange between C¹⁶O₂ and H₂¹⁸O in response to CA activity may distinctively track the development of the gastric pathogen in human stomach and might be considered as a potential biomarker for the non-invasive detection of H. pylori infection.

Figure 1

Comparisons of δ_DOB ¹⁸O‰ and δ_DOB ¹³C‰ values of exhaled breath CO₂ associated with ¹³C-labelled glucose metabolism in presence [H. p. (+)] and absence [H. p. (−)] of H. pylori infection.

(a) Excretion kinetics of δ_DOB¹⁸O‰ and δ_DOB¹³C‰ values for H. pylori positive [H. p. (+)] and H. pylori negative [H. p. (−)] individuals up to 120 minutes. (b,c) The Box-Whiskers plots demonstrating a statistically significant differences of δ_DOB¹⁸O‰ [p < 0.05] and δ_DOB¹³C‰ [p < 0.01] values at 45 minutes between H. pylori positive [H. p. (+)] and H. pylori negative [H. p. (−)] individuals. *p < 0.05 and **p < 0.01. Data are means ± SE.

We then critically assessed the excretion dynamics of δ_DOB¹³C (‰) (Fig. 1a) values in exhaled breath samples in response to ¹³C-enriched glucose ingestion. The isotopic enrichments of ¹³C in breath CO₂ were significantly higher (Fig. 1c) in individuals with H. pylori positive compared with individuals with H. pylori negative. It was noticed that for H. pylori positive patients the δ_DOB¹³C (‰) values increased gradually with time at a faster rate in comparison with individuals without H. pylori infections. Several lines of evidence suggest that H. pylori can metabolize glucose in both oxidative and fermentative pathways^8,9 and as a consequence the catabolism of glucose resulted in higher isotopic enrichments of ¹³C in exhaled breath CO₂. Moreover, in some early evidences^9,16, it was demonstrated the biphasic characteristics of glucose utilization by H. pylori with a slower initial period, followed by a second phase with a higher rate of glucose uptake. The transport and utilization of glucose was previously investigated into the intact micro-organism employing the radioactive tracer analysis. Therefore, the gradual increase of the δ_DOB¹³C (‰) values in the time-dependent excretion dynamics is possibly attributed to the increased rate of glucose utilization through the biphasic activity of the micro-organism. Hence our results of the δ_DOB¹³C (‰) values in exhaled breath are coincidence with the previous study¹⁶, where the uptake of glucose into H. pylori cells exhibited the biphasic patterns. Our observations therefore, point to new perspectives into the physiology of H. pylori underlying the isotopic fractionations of ¹³C in breath CO₂ associated with glucose metabolism.

We next explored how the time-dependent excretion dynamics of isotopic breath CO₂ changes after administration of unlabelled glucose (i.e. with no ¹³C-enriched glucose), as the potential role of glucose metabolism in response to unlabelled glucose ingestion for individuals with H. pylori infection and the possible links underlying the ¹⁸O and ¹³C-isotopic fractionations of breath CO₂ remains unknown. To investigate this, we performed the 2-h excretion kinetics of δ_DOB¹⁸O‰ and δ_DOB¹³C‰ values simultaneously in breath samples for a number of 52 H. pylori infected and 45 non-infected individuals. When the unlabelled glucose was orally administered in positive H. pylori patients, the post-dose δ_DOB¹⁸O‰ values in breath samples manifested a significant change with time and depicted the similar excretion kinetics with that of ¹³C-enriched glucose (Fig. 2a,b), whereas no significant change of the post-dose δ_DOB¹⁸O‰ values in breath samples over time was evident for H. pylori negative individuals (Fig. 2a). These findings suggest that the mechanisms i.e. oxidation of glucose in the bacterial environment to form bicarbonate (HCO₃⁻) and subsequently the enzyme carbonic anhydrase-mediated rapid inter-conversion of HCO₃⁻ and CO₂, leading to the generation of ¹²C¹⁶O¹⁸O, exclusively depends on the substrate (glucose) regardless of its isotopic nature. Interestingly, although the isotopic nature of the substrate is vital to observe effectively the ¹³C-isotopic enrichments of breath CO₂ (i.e. enhancement of δ_DOB¹³C‰ values), yet the enrichments of δ_DOB¹³C‰ values, due to natural abundances of ¹³C, during glucose metabolism of H. pylori are significantly distinguishable for H. pylori positive patients (Fig. 2c,d), suggesting a new step towards characterizing the transport and utilization of unlabelled glucose into the human pathogen for better understanding of its physiology in the gastric niche. In view of the breath results, we have also established the previous hypothesis¹² that the bacterium requires high CO₂ for growth and the interconversion of ¹⁸O (H₂¹⁸O) and ¹⁶O (C¹⁶O₂) is vital, catalyzed by the enzyme carbonic anhydrase activity (α-CA and β-CA) of H. pylori and thus this activity might be a contributing factor for the development of the disease in the gastric environment. The summary of the detailed results has been provided in the Table 1.

Table 1 The summary of the detailed characteristics of the study subjects for H. pylori infected and non-infected groups. UBT and GBT correspond to the urea breath test and glucose breath test respectively.

Figure 2

Excretion kinetics of δ_DOB ¹⁸O‰ and δ_DOB ¹³C‰ values of exhaled breath CO₂ during unlabelled glucose (¹²C-G) metabolism of H. pylori positive [H. p. (+)] and H. pylori negative [H. p. (−)] individuals.

(a) The excretion kinetics illustrating the increased δ_DOB¹⁸O‰ values for H. pylori positive [H. p. (+)] patients. (b) depicts the similar excretion kinetics of δ_DOB¹⁸O‰ values with that of ¹³C-enriched glucose. (c,d) enhancement of δ_DOB¹³C‰ values for H. pylori positive [H. p. (+)] patients and comparisons with the ¹³C-enriched glucose. Values are means ± SE.

To distinctively track the H. pylori infection as well as for early detection prior to the onset of different gastric diseases, we subsequently determined numerous statistically sound optimal diagnostic cut-off points of δ_DOB¹⁸O‰ and δ_DOB¹³C‰ values in exhaled breath associated with ¹³C-labelled and unlabelled glucose metabolism, using receiver operating characteristics curve (ROC) analysis (Fig. 3). Individuals with δ_DOB¹⁸O‰ ≥ 1.44‰ and δ_DOB¹⁸O‰ ≥ 1.1‰ were considered to be H. pylori positive with and without ¹³C-enriched glucose metabolism respectively and these corresponded to the diagnostic sensitivity and specificity of 100% and ~98%, respectively. On the contrary, a different optimal diagnostic cut-off point of δ_DOB¹³C‰ ≥ 33.32‰ between individuals with H. pylori positive and negative, demonstrated the sensitivity and specificity of 100% and 100%, respectively, when ¹³C-labelled glucose is ingested, whereas without ¹³C-labelled glucose, δ_DOB¹³C‰ ≥ 1.51‰ precisely diagnosed the infected and non-infected persons corresponding to the similar levels of diagnostic sensitivity (100%) and specificity (98%). It is noteworthy to mention that the uncertainty of these cut-off values is associated with the less-sensitive techniques for isotope measurements and the variation of isotopic fractionations in the test meal. However, these findings suggest that the oxygen-18 and carbon-13 isotopic fractionations of the major metabolite CO₂ in human breath linked to glucose metabolism of H. pylori provide a new non-invasive approach to treat the world’s most common chronic bacterial infection of humans and hence may have a broad clinical efficacy for precise assessment of the gastric pathogen H. pylori.

Figure 3

Receiver operating characteristic (ROC) curves analysis for the optimal diagnostic cut-off points of H. pylori infection.

(a) δ_DOB¹⁸O values ≥1.44‰ and 1.1‰ are indicative of the H. pylori infection associated with ¹³C-labelled and unlabelled glucose metabolism at 45 minute, respectively, whereas (b) δ_DOB¹³C values ≥33.21‰ and 1.51‰ indicate the same for ¹³C-labelled and unlabelled glucose metabolism, respectively.

We next explored the efficacy of the glucose breath test in response to the standard eradication therapies of the infection. A marked depletions of both δ_DOB¹⁸O‰ and δ_DOB¹³C‰ values for H. pylori infected patients (n = 37 for ¹³C-glucose and n = 28 for ¹²C-glucose) (Fig. 4) after complete eradication of the infection were manifested, suggesting the widespread clinical significance of the glucose breath test. Our findings associated with the glucose metabolism by H. pylori infections thus point towards a considerable clinical advancement in the non-invasive diagnosis of H. pylori infection by contrast with the currently available ¹³C-urea breath test (¹³C-UBT), where ¹³C-enriched substrate (urea) is usually used. In view of this result, we therefore posit that the glucose breath test by ingestion of a natural substrate (unlabelled glucose) is a valid and potentially robust new-generation diagnostic tool and thus indicate great promise for comparatively less-expensive and non-toxic global technique, in comparison with the ¹³C-UBT, for the non-invasive assessment i.e. early detection and follow-up of patients after eradication of H. pylori infection.

Figure 4

Glucose breath test in response to the standard eradication therapies of the H. pylori infection.

(a–d) a marked distinction (*p < 0.01) for the δ_DOB¹⁸O and δ_DOB¹³C values at 45 minute was observed before and after the therapies in case of both ¹³C-glucose and ¹²C-glucose ingestion.

Finally, we elucidated the potential metabolic pathways (Fig. 5) underlying the mechanisms linking isotopic fractionations of breath CO₂ and glucose utilization by H. pylori infection. When a dose of glucose is orally administered to the patients, the ingested glucose disposal takes place in the cytoplasm of H. pylori through the HP1174 transporter (protein)7. After glucose enters into the cytoplasm, it is phosphorylated to produce glucose-6-phosphate which subsequently incorporated with three potential metabolic routes: glycolysis, pentose phosphate and the Entner-Doudoroff pathway^17,18. A part of the total glucose-6-phosphate, which goes into the pentose phosphate pathway, is predominantly oxidised into CO₂. The remaining part of glucose-6-phosphate enters into the other two metabolic pathways and may lead to the generation of pyruvate¹⁷ and eventually gives rise to CO₂ followed by the formation of acetate as the key metabolite through the intermediary oxidative and reductive fermentation pathways⁹. Another fate for pyruvate is the conversion of acetyl-CoA, which afterwards enters into the Krebs cycle and generates CO₂ as a by-product¹⁸. Therefore, the administration of ¹³C-glucose (either from ¹³C-enriched exogenous glucose or naturally abundant ¹³C-glucose) facilitates the production of ¹³CO₂ in the by-product CO₂ in presence of H. pylori infection. Thereafter, the major metabolite CO₂ (¹³CO₂ and ¹²CO₂) produced by all these metabolic processes is then transported through the blood streams and eventually excreted as ¹³C¹⁶O¹⁶O and ¹²C¹⁶O¹⁶O in exhaled breath. Conversely, the cytoplasmic β-carbonic anhydrase (β-CA) activity of H. pylori catalyzes the reversible interconversion between the major metabolite CO₂ and HCO₃⁻ (CO₂ + H₂O↔H⁺ + HCO₃⁻).Then the CO₂ diffuses rapidly through the inner member into the periplasm of H. pylori, where it forms carbonic acid (H₂CO₃), catalyzed by the α-CA. Because the isotopes ¹⁶O of ¹²C¹⁶O₂ and ¹⁸O of H₂¹⁸O are rapidly exchanged in response to periplasmic α-CA activity, it therefore leads to the generation of H₂C¹⁸O¹⁶O₂. This carbonic acid rapidly degasses to produce ¹²C¹⁸O¹⁶O, which is then transported to the lungs and is excreted through exhaled breath. As a result, individuals with H. pylori infections exhibit the preferential isotopic enrichments of ¹⁸O in breath CO₂, whereas no significant change of ¹⁸O in CO₂ was manifested in H. pylori-uninfected individuals.

Figure 5

Potential metabolic pathways for ¹³C and ¹⁸O-isotopic exchanges associated with glucose metabolism by H. pylori infection.

The ingested glucose in the cytoplasm of H. pylori through the HP1174 transporter is phosphorylated to produce glucose-6-phosphate which subsequently incorporated with three potential metabolic routes: glycolysis, pentose phosphate and the Entner-Doudoroff pathway to finally produce major metabolite CO₂.This major metabolite CO₂ is then transported through the blood streams and eventually excreted as ¹³C¹⁶O¹⁶O and ¹²C¹⁶O¹⁶O in exhaled breath. Conversely, the major metabolite CO₂ diffuses rapidly through the inner member into the periplasm of H. pylori, where it forms carbonic acid (H₂CO₃), catalyzed by the α-CA. The isotopic exchange between ¹⁶O of ¹²C¹⁶O₂ and ¹⁸O of H₂¹⁸O in response to periplasmic α-CA activity leads to the generation of ¹²C¹⁸O¹⁶O, which is then transported to the lungs and is excreted through exhaled breath.

In conclusion, our new findings point to a fundamental mechanism underlying both the ¹⁸O and ¹³C stable isotopic fractionations of the major metabolite CO₂ in human breath related to glucose metabolism of H. pylori infection in humans. Subsequently, we have taken a step towards unravelling the potential metabolic pathways linking the ¹⁸O and ¹³C-isotopic exchange of breath CO₂ and the glucose uptake by H. pylori, thus suggesting that breath ¹²C¹⁸O¹⁶O and ¹³C¹⁶O¹⁶O in response to glucose ingestion could be used as potential molecular biomarkers to distinctively track the pathogenesis of H. pylori infection in a non-invasive approach. Although many imperative gaps remain in our understanding of these processes and in the pathophysiology underlying the isotopic exchange and glucose metabolism, our studies may provide new perspectives in the isotope-specific molecular diagnosis of H. pylori infection and hence may pave the way for broad clinical applications along with eradication purposes following standard therapies. Furthermore, new insight into the mechanism linking the isotopic exchange in breath molecule CO₂ to glucose metabolism of H. pylori is fostering exploration of the molecular basis of this infection and new and better approaches together with new pharmacological targets to prevent or treat the deleterious effects of the world’s most common gastric pathogen.

Materials and methods

Subjects

Two hundred and twenty four individuals (135 male and 89 female with average age of 39 ± 10 yrs (SD)) were enrolled for this study with different gastrointestinal disorders such as active peptic ulcer disease (PUD), chronic gastritis and univestigated dyspepsia. We categorized all the human subjects in two distinct groups: infected with H. pylori (H. pylori positive patients: 124) and without the infection of H. pylori (H. pylori negative patients: 100) depending on the reports of gold standard invasive and non-invasive methods, i.e. endoscopy and biopsy based rapid urease test (RUT) and ¹³C-urea breath test (¹³C-UBT).The ¹³C-UBT was considered to be indicative of H. pylori positive when δ_DOB¹³C (‰) ≥ 3‰^19,20,21. There were no mismatches between the two test-reports of all the subjects enrolled in this study (Supplementary Table 1). Exclusion criteria included patients with previous history of diabetes and gastric surgery, taking antibiotics, proton pump inhibitors or H₂ receptor antagonists in the four week prior to endoscopy and ¹³C-UBT. We received the Ethical approval from the Ethics Committee Review Board of AMRI Hospital, Salt Lake, Kolkata, India (Study no.: AMRI/ETHICS/2013/1). The current protocol has also been approved by the institutional administrative of S. N. Bose Centre, Kolkata, India (Ref. no.: SNB/PER-2-6001/13-14/1769) and the methods were carried out in accordance with the approved guidelines. Informed written consents were taken from all patients participating in this study.

Breath samples collection and measurements

All the human subjects enrolled for the study completed their endoscopic examinations and ¹³C-UBTs, 1-2 days prior to glucose breath test (GBT). On the study day before GBT, all the patients were instructed for their mouth-washing to prevent any kind of contact of ingested test meal with the oral cavity bacteria. After an overnight fasting (10-12 hours), an initial baseline breath sample was collected in a 750 ml breath collection bag (QUINTRON, USA, SL No.QT00892) from each subject. After that a test meal of 75 mg U-¹³C₆ labelled D-glucose (CIL-CLM-1396-CTM, Cambridge Isotope Laboratories, Inc. USA) or 75 mg unlabeled glucose dissolved in 50 ml water was orally administered to the patient and then subsequent breath samples were collected at 15 minute intervals up to 120 minute. The physical activities of the subjects were restricted inside a room during the test. For the measurements of ¹⁸O/¹⁶O and ¹³C/¹²C isotope ratios of exhaled breath CO₂, a laser-based high-precision ICOS system was employed and the detailed description of the ICOS was given in the following section.

Integrated cavity output spectrometer (ICOS) for breath analysis

For high precision isotopic measurements of breath CO₂, a high-resolution carbon dioxide analyzer, based on off-axis integrated cavity output spectroscopy (ICOS) method, has been utilized in this study. The detailed description and the measurement accuracy of ICOS method in comparison to the conventional isotope ratio mass spectrometry (IRMS) have been previously demonstrated elsewhere^22,23. In brief, the laser-based ICOS spectrometer (CCIA 36-EP, Los Gatos research, USA) exploits a high-finesse optical cavity (~59 cm) with two high reflectivity mirrors (R ~ 99.98%) at the both ends of the cavity. This arrangement provides an effective optical path-length of around 3 km through the measuring gas sample, thus offering a high-precision measurement. A continuous wave distributed feedback diode laser operating at ~2.05 μm is repeatedly tuned over 20 GHz to scan the absorption features of ¹²C¹⁶O¹⁶O, ¹²C¹⁸O¹⁶O and ¹³C¹⁶O¹⁶O at the wavenumbers of 4874.448 cm⁻¹, 4874.178 cm⁻¹ and 4874.086 cm⁻¹ respectively. The absorption features of ¹²C¹⁶O¹⁶O, ¹²C¹⁸O¹⁶O and ¹³C¹⁶O¹⁶O, corresponding to the R (27), P (36) and P (16) ro-virational lines respectively, in the (2,0⁰,1) ← (0,0⁰,0) vibrational combinational band of CO₂, have been utilized to measure the ¹³C/¹²C and ¹⁸O/¹⁶O isotope ratios simultaneously. The isotopic enrichments of ¹³CO₂ and ¹²C¹⁸O¹⁶O have been expressed as the conventional notations i.e., δ¹³C (‰) and δ¹⁸O (‰) respectively, relative to the international standard Pee Dee Belemnite (PDB).

where the standard PDB values for (¹³C/¹²C)_Standard and (¹⁸O/¹⁶O)_Standard are 0.0112372 and 0.0020672, respectively. The accuracy and precision of ICOS method for the δ¹³C‰ measurements were determined by measuring three calibration standards, containing 5% CO₂ in air analyzed by IRMS (Cambridge Isotope Laboratory, USA), with δ¹³C values ranging from baseline-level (−22.8‰) to high-level (−7.33‰) including the mid-level (−13.22‰) whereas a standard NOAA air tank was used for the calibration of δ¹⁸O (‰) measurements (Supplementary Table 2 and Supplementary Table 3). A 25 mL breath sample was injected into the ICOS cell with a syringe/stopcock for the measurements. High-purity dry nitrogen (HPNG10-1, F-DGSi SAS, France, purity >99.99%), as the carrier gas, was used to purge the cavity and dilute the breath samples.

Statistical method

All the data were presented as mean ± SE (Standard Error). For statistical analyses, we performed non-parametric Mann-Whitney test and one way ANNOVA test. A two sided p value <0.05 was taken account as statistically significant of data. Box-Whiskers plots were utilized to demonstrate the statistical distribution of isotopic enrichments of exhaled breath CO₂. To obtain the optimal diagnostic cut-off values for δ_DOB¹⁸O‰ and δ_DOB¹³C‰ associated with ¹³C-labelled (¹³C-G) and unlabelled glucose (¹²C-G) metabolism, we performed receiver operating characteristic curve (ROC) analysis (Supplementary Table 4, Supplementary Table 5, Supplementary Table 6 and Supplementary Table 7). All the data were analysed using Origin Pro 8.0 (Origin Lab Corporation, USA) and Analyse-it Method Evaluation software (Analyse-it Software Ltd, UK, version 2.30).