Initial study of three different pathogenic microorganisms by gas chromatography-mass spectrometry

Background: Diagnoses of respiratory tract infections usually happen in the late phase of the disease and usually result in reduction of the pathogen load after broad-spectrum antibiotic therapy, but not in eradication of the pathogen. The development of a non-invasive, fast, and accurate method to detect pathogens has always been of interest to researchers and clinicians alike. Previous studies have shown that bacteria produce organic gases. The current study aimed to identify the volatile organic compounds (VOCs) produced by three respiratory tract pathogens, including Staphylococcus aureus, Escherichia coli and Candida albicans. Methods: The VOCs produced were identified by gas chromatography–mass spectrometry (GC-MS), with prior collection of microbial volatile compounds using solid phase microextraction (SPME) fiber. The volatile compounds were collected by obtaining bacterial headspace samples. Results: Results showed that these three organisms have various VOCs, which were analyzed under different conditions. By ignoring common VOCs, some species-specific VOCs could be detected. The most important VOC of E. coli was indole, also some important VOCs produced by S. aureus were 2,3-pentandione, cis-dihydro-α-terpinyl acetate, 1-decyne, 1,3-heptadiene, 2,5-dimethyl pyrazine, ethyl butanoate and cyclohexene,4-ethenyl. Furthermore, most of the identified compounds by C. albicans are alcohols. Conclusions: The detection of VOCs produced by infectious agents maybe the key to make a rapid and precise diagnosis of infection, but more comprehensive studies must be conducted in this regard.


Introduction
Infectious diseases are the main reason for morbidity and mortality in developing countries, especially among children 1 . Staphylococcus aureus is a common inhabitant of the upper respiratory tract in children, and the causative agent for many infections. It is believed that people under 20 are more likely to have these bacteria. There is a greater possibility that S. aureus exists in the respiratory tract of infants aged 3 months or younger than in people of other ages 2 . Moreover, S. aureus is colonized in the nasopharynx in 10-35% of children, and in almost 35% of the adult population 3 .
Escherichia coli is one of the most significant pathogens affecting preterm infants 4 . Some studies in developing countries have suggested that gram-negative rods (such as E. coli) are the major causes of infection in premature infants (0-6 days) 5-7 . Furthermore, infections caused by E. coli are one of the most important causes of death in the early neonatal period 5 . Candida albicans is an opportunistic pathogen and an agent of nosocomial infection 8 .
Generally, the causative agents of respiratory tract infections are diagnosed in late phases of the disease 7 . Such infections need broad-spectrum antibiotic therapy, the consequences of which are a reduction in the pathogen load, but not eradication. Moreover, such therapies increase the probability of drug-resistant infections spreading 9 . Accurate and rapid detection of pathogens is a critical step for adequate treatment of infection 10 . and a noninvasive diagnostic method that has a high degree of accuracy needs to be developed 11 .
It has been shown that bacteria produce organic gases. Different types of microorganisms have a distinct metabolism, and they produce various types of volatile organic compounds (VOCs) 12-14 .
Attempts have been made to identify the VOCs of pathogenic organisms [15][16][17][18][19][20] . There are several sophisticated methods available that have been used for recognizing VOCs; these include gas chromatography-mass spectrometry (GC-MS) 21 , selected ion flow tube mass spectrometry (SIFT-MS) 22 , electronic noses (eNoses) 23 , and ion-molecule reaction mass spectrometry (IMRMS) 24 . Previous studies suggest that GC-MS is the most appropriate and reliable technique for the isolation and identification of VOCs 25-27 .
The current study aimed to identify the volatile organic compounds (VOCs) produced by three respiratory tract pathogens, including Monocultures of all strains were cultured 24 hours in nutrient agar, and then sub-cultured aerobically at 37°C in 30 ml of two different types of broth medium, Mueller Hinton broth (MB) and tryptic soy broth (TSB), in 100 ml sterilized glass bottles. For a more careful assessment of VOCs produced by each microorganism, the headspace was extracted from both media at three different time points: 2, 4 and 24 hours. To increase the possibility of VOC production, bottles containing cultured microorganism were shaken at 150 rpm during incubation time 28 . A suspension of microorganisms with approximately OD 600 ~0.5 in culture media was used during the headspace extraction 10 , and the corresponding sterile broth mediums were used as the blank samples 29 .

Headspace extraction
A solid phase microextraction (SPME) fiber holder (57330-U, Sigma-Aldrich) containing fiber coated with divinyl benzene/ carboxen/poly dimethyl siloxane 50/30 µm (DVB/CAR/PDMS) (57328-U, Sigma-Aldrich) was used for absorption of volatile compounds from the headspace of pathogens. To provide conditions that increase the rate of VOC absorption, after incubation time, 2ml of NaCl 36% was added to each culture. Then the DVB/CAR/ PDMS fiber was suspended from the top of the bottle containing the culture and placed on a magnetic stirrer hotplate at 70°C for 30 minutes 30 . After that, the fiber was placed at the injection site of GC-MS and all the absorbed VOCs entered the device. Eventually each VOC is represented as a chromatogram peak in the monitor that is connected to the GC-MS. For thermal desorption, the SPME fiber remained in the injector for 2 minutes before it was exposed to the headspace of the pathogen samples 31 . To avoid possible false discoveries each state was tested at least three times.

GC-MS
To study the bacterial VOCs, a Thermo-Finnigan Trace GC-MS system (Thermo Quest-Finnigan Co) equipped with a DB-5 column (60 m length, 0.25 mm inner diameter, and 0.25 µm film thickness) with helium carrier gas at a flow rate of 1.1 ml/min was used. The starting temperature was 50°C, increasing at a rate of 10°C/minute up to 250°C. The GC-MS was set in splitless mode and a quadrupole ion trap with ionization energy of 70 eV was used in the filament.
VOCs were identified using the National Institute of Standards and Technology (NIST) reference library. To analyze the GC-MS

Amendments from Version 2
In response to the referees, some spelling mistakes in chemical compounds were modified throughout the paper.

See referee reports
REVISED data, Xcalibur 3.0 with Foundation 3.0 SP2 software (Thermo Fisher Scientific) was used, and the kovats retention index (RI) was calculated for each chromatographic peak.
When calculating the RI, a series of standards were used: n-alkanes were injected into the GC-MS the day before starting experiments, using the same temperature profile that would be used for the analysis of VOCs. The NIST17 Mass Spectral Library (NIST7/2017/EPA/NIH) was used to identify each compound according to its RI. Since there may be several types of volatile compounds have similar RI, to validate the final results extensive studies were also performed by a phytochemist to determine if the compounds were organic. The common VOCs released from the sterile environment (Blank samples) and tests were not considered.

Results
The VOCs produced by S. aureus, E. coli and C. albicans were assessed under six different conditions (using two types of media and taking measurements at three time points). The Xcalibur raw files for these three pathogens are available at https://doi. org/10.6084/m9.figshare.5178004.v1 32 .
One chromatogram of the six chromatograms obtained is displayed in Figure 1, showing the chromatogram obtained 4 hours after culture in TSB medium, for each pathogen. The five The processed GC-MS data obtained in the current study is available in a total of 18 tables as supplementary GC-MS data. It shows the details of the VOCs detected for each of the three pathogens, each analyzed under different conditions (using two types of media and taking measurements at three time points, as explained above).
For a better overview the detected VOCs are shown in three tables (at the 2 hour time point in Table 1, at the 4 hour time point in Table 2 and at the 24 hour time point in Table 3), alongside the      percentage of the total area that the average peak of the detected VOC covered. In other words it is proportional to amount of the compound that is present.

C. albicans in TSB
Some VOCs were common among organisms and were generated by two or three organisms at an approximately equal rate, including 1,2-benzenedicarboxylic acid, 1,9-decadiene, 2,5-(1,1-dimethylethyl)-phenol, 2,6-bis(1,1-dimethylethyl)-4-methyl-phenol, 3-propionyl oxy pentadecane and anisol (Table 1). Some common VOCs were produced at a greater rate between one organism and another. It can be concluded that these VOCs could also be more important in the organism that produces greater quantities. 1-penten-3-ol was produced from E. coli in TSB medium after 2 hours (0.02%); under identical conditions, more of it was produced by S. aureus (5.14%) than by E. coli. Furthermore, indole was produced from E. coli after 2 hours of culture in two types of medium (82.61% for MB and 90.97% for TSB) and was also produced by S. aureus after 2 hours in TSB medium, although at a much lower rate (0.48%) ( Table 1).

Discussion
As previous studies have shown, organisms are able to produce either common or specific VOCs 33-35 . In the current study, GC-MS was used to detect VOCs generated by three pathogenic organisms in the human respiratory tract. The VOCs of E. coli, S. aureus and C. albicans were analyzed at three different time points, using two different types of media (Figure 1).
Results of the current study suggest that VOCs exclusively produced by E. coli are 1-(1,5-dimethyl)-4-hexyl-4-methyl-benzene, 2,6-dibutyl-2,5-cyclohexadiene-1,4-dione, benzophenone, bisabolene, copaene, decanol, dodecanol, indole, limonene, muurola-4,5-diene, nerylacetate, phenyl ethyl pyrrole, sesquiphellandrene, tetradecane, 2-acetyl-1-pyrroline and 2-methyl tetradecane. The most important compound among these is Indole, because it is generated at the three time points and also it was the most produced VOC by E. coli (at least 82%). Other studies have confirmed this finding 28,29,35 . E. coli produced tryptophanase and this enzyme degrades tryptophan to indole and the other compounds 36 . In future studies, it is advisable to measure the amount of indole in the exhaled air of infected patients with E. coli and compare it with the current results. This is because in the patient's lungs the level of tryptophan is not the same as culture medium. It is also suggested that the amount of released indole from this bacterium should be evaluated under at in-vitro conditions and with using the simplest culture medium (relative to TSB and MB). In this way, we will have a more detailed thought of the importance of the Indole production by E. coli.
It is suggested that the findings of future studies on the exhaust air of respiratory infections patients with these three pathogens should be compared with the identified VOCs in this study. Although there may be some differences between the results of in-vitro and in-vivo studies there seems to be significant similarities over the dominant detected VOCs.
Finding a non-invasive and rapid method for diagnosis of infectious agents is a subject of interest, so it has been investigated in several studies 33,42-45 . The current study showed that using SPME fiber and GC-MS for extraction and detection of VOCs allowed detection of more specific VOCs for the three pathogenic respiratory tract organisms, E. coli, S. aureus and C. albicans, which could be used as biomarkers for their identification. It is essential that more comprehensive studies be conducted to create a more complete profile of VOCs for these organisms, and so that the methods can be developed further.

Norman Ratcliffe
Institute of Biosensor Technology, University of the West of England, Bristol, UK The paper shows some interesting preliminary results. The authors should be complimented for their good English. The works cries out for more repeats and the authors appreciate this. Within the abstract, and in the text throughout, chemical names should not have capital letters, more importantly some of the compound names need checking, particularly 1,3-heptadiene-3-yne which is in the abstract, it cannot be right. Also check in tables, spelling and also are you sure: 1,2-Benzene dicarboxulic acid, this should be 1,2-benzenedicarboxylic acid 1,3-Butadiyene, do you mean 1,3-butadiene? Likewise 1,2-Butadiyene ? 1,3-Heptadiene-3-yne? Can't be Dibutyl phatalate should be phthalate, also its other name is 1,2-Benzenedicarboxylic acid, dibutyl Best to be consistent with names, I only draw attention to this case as an example, because ester. 1,2-benzene dicarboxylic acid has been stated to be found, and when they have the same naming, its brings out the thought that the acid may have come from the ester (I say this, rather than the other way around as the butyl ester is a common plasticizer, and can be a contaminant). Longifolrne spelling! And it appears just below, they are the same? Do they have the same rts? Longifolene Phatalic acid, butyl ester, spelling Sesquiphellandrene and beta-Sesquiphellandrene, are both listed are they the same? The researchers tackle an area of significant interest, rapid determination of bacterial species, especially associated with life threatening illness, using a relatively new approach, that of VOC analyses. In general the paper reads well.
The analytical method for VOC analyses is ok. Could the authors confirm whether the cfu/ml count is approx. for the different species at the same time points to enable a good comparison of VOCs to be The analytical method for VOC analyses is ok. Could the authors confirm whether the cfu/ml count is approx. for the different species at the same time points to enable a good comparison of VOCs to be made. One major weakness to the work is that only one analysis for each species at one time point and media was undertaken, the literature gives examples of several analyses being undertaken for similar studies. The title is too ambitious, rather than diagnoses being purported, maybe an" initial study of…" would be more accurate.
The authors could comment on how their research would ultimately fit into a clinical test, if the VOCs were to be analysed in breath, would the same volatile profile be expected?
In the text it was stated that "Extensive studies were also performed by a phytochemist to determine if the compounds were organic" , some explanation would be good as to what this means.
Some other matters, VOC abbreviation used twice in the abstract, also in quite a few places the chemical names have capital letters and are misspelt and merged with another word.
In Results, what is uncommon determined by, some more discussion on stats…

If applicable, is the statistical analysis and its interpretation appropriate? Partly
Are all the source data underlying the results available to ensure full reproducibility? Yes

Are the conclusions drawn adequately supported by the results? Partly
No competing interests were disclosed.

Competing Interests:
I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.