Salivary biochemical assessments to determine smoking cessation under smoking cessation program

Background: The study aim was to investigate the potential of salivary biochemical assessments to determine smoking cessation in patients under smoking cessation programs. Methods: The study validated smoking behavior and degree of nicotine dependence using Fagerström test for nicotine dependence in smokers and non-smokers. Protein components were examined using Raman spectroscopy. Salivary MMP-8 levels were analyzed using Quantikine human total MMP-8 immunoassay-kits. Thiocyanate (SCN - ) concentrations, peroxidase and glutathione-peroxidase activities were determined using spectrophotometric methods. Human mononuclear cell line was examined for M1 polarization. The data were analyzed using Fisher’s exact test and two-tailed paired t-test at α=0.05. Results: Amide I peak in smoker group was signicantly reduced as compared to non-smokers (p<0.05). Sulfhydryl bond, dialkyl disuldes SS-stretch and corresponding disulphide bond at 500-545 cm -1 were signicantly increased in smokers (p<0.05). The CH 2 wag showed higher intensity amongst smoker group representing a recoil of carbon atoms. Salivary MMP-8 levels of participants after smoking cessation was signicantly lower than smoking salivary levels (p<0.05). The SCN - and peroxidase concentrations were signicantly higher in smoking group whereas an M1 polarization was seen in non-smokers. Conclusion: Salivary analysis verify patients


Background
Saliva is a multi-constituent uid which can be collected by non-invasive means and has considerable potential for surveillance of general health through biochemical pro ling [1]. It contains several proteins, peptides, matrix metalloproteinases (MMP), antibacterial enzymes, growth factors, and immunoglobulins that have speci c functions towards health and disease [2,3]. Saliva sampling has little ethical and consent constraints as opposed to other types of sample collections. The composition of saliva is not as complex as serum and thus, the biochemical parameters can be accurately and rapidly pro led to assess any suggestive disease state [4]. Therefore, salivary biochemical assessments can be a potential marker to discern risk factors that eventually mediates a pathogenic effect.
Smoking is a major risk factor for strokes, heart attacks, cancer and pulmonary diseases. An estimated 5.4 million deaths were reported by World Health Organization in 2004 and 100 million deaths during the 20th century, due to smoking [5]. The nicotine present within cigarettes is known to metabolize several biochemical components in saliva including proteins, mucins, and MMPs [6][7][8]. The nicotine-mediated metabolites can be a potential biochemical marker to identify the smoking habits extent, addiction and cessation. Thus, there is a need to identify the biochemical parameters that can distinguish between smoking patterns. Raman spectroscopy analyses biological tissues and uids at molecular level while ciphering detailed information for diagnosis [9,10]. The method is rapid while requires only a small quantity of biological uids for analysis. The approach is dependent on vibration transitions due to inelastic scattering of monochromatic light at infrared ranges. No studies have utilized the technique to identify smokingrelated changes. Similarly, salivary MMPs and signi cant metabolites like thiocyanate (SCN − ) and peroxidase have not been investigated to identify changes in patients attempting smoking cessation under a smoking cessation program. Thus, the biochemical parameters and the technique (Raman spectroscopy) can be further investigated to identify its potential as a smoking cessation marker.
Seventy percent of the smokers express a voluntary desire to quit smoking, of which 80% return to smoking within the rst month. A < 20% quit smoking for 6 months with 3% remaining abstinent for more than a year [11]. Several health universities have initiated smoking cessation programs to help people quit smoking while routinely monitor smoking status and outcomes of their service users. The program participants are made to understand oral and general health hazards of tobacco which motivates them to quit smoking [12]. The Russell standards for smoking cessation suggested that smoking cessation should be biochemically validated through studies to exhibit variables that can verify smoking cessation in program participants [13]. However, no studies in the literature could be identi ed that examine the potential of salivary biochemical parameters to determine smoking cessation in participants under smoking cessation program. Thus, the present study aimed at investigating the potential of salivary biochemical assessments to determine smoking cessation in patients under smoking cessation programs. The tested null hypothesis in the present study was that there is no difference in the salivary biochemical components of the smokers and non-smokers or smoker who have quit smoking.

Study design, setting, and sampling
Patients enrolled in smoking cessation program were included in the study with no signi cant past medical history. The treatment provided under the program included an assessment of patient's current commitment, readiness and ability to quit smoking throughout the quitting process, co-validated smoking behavior and patient's degree of nicotine dependence using validated tools such as the Fagerström test for Nicotine Dependence. The program included behavioral support to assist the patient to set a quit date and support the patient to exercise self-control. All patients were given the opportunity to provide feedback through a pre-designed patient satisfaction questionnaire. Regular audits were also performed to monitor the program performance.
Through non-probability purposive sampling ve patients per group (n = 5) were selected based on voluntary participation after explaining the study design and thoroughly obtaining consent from the patients.

Page 4/16
The inclusion and exclusion criteria for the study was as follows: Inclusion criteria: Patients using smoked tobacco products: including cigarettes, cigars, and pipes.
A standard de nition of tobacco use in adolescents is any use in the past 30 days. But the standard patients will include patients that have been smoking for ≥ 3 years at least 10 cigarettes/day. Humans, all races, ethnicities, cultural groups.
Adults with age range between 21-50 years.
Patients with apparently healthy medical history.
Exclusion criteria: Patients with compromised medical history.
Adults under the age of 21 years.
Patients having less than 10 cigarettes/day.
Patients not using tobacco, including cigarettes, cigars, and pipes.

Saliva collection
The saliva was collected in both stimulated and unstimulated conditions. For stimulated conditions, subjects chewed a 5 gms piece of para n wax for 5 minutes before saliva collection. While for unstimulated conditions, it was collected in the resting phase of the patient. After saliva collection, it was maintained on ice and immediately centrifuged within 1 h of collection at around 12000 × g for 30 min to remove debris. The samples were subjected to 70% ethanol. This was done to precipitate protein samples from the saliva.

Raman micro-spectroscopy
The saliva was transferred in a germproof medicine glass and later put into sterile test tube by a disposable syringe used separately for each specimen (5 ml). The tubes were placed in a portable refrigerator (ETS compressor portable freezer; Dataran Mestika, Taman Mestika, 56100 Kuala Lumpur, Malaysia) under 4 0 C and moved to the venue of Raman investigation. The period of transportation did not exceed 30 minutes, which is well below the time to guarantee preservation of biological properties of the specimens [14]. All Raman analysis were carried out using Raman microscope (LabRam Horiba JobinYvon) at Nanocat Research Center, UM, Kuala Lumpur having Rayleigh rejection lter and a cooled charged-couple device detector. The spot size of the excitation beam was 1.1 µm with a collimation made by 100 µm pinhole. The diode laser (Coherent, CUBE) had an excitation wavelength of 785 nm with incident laser power on the sample upto 10 mW at × 20 objective. The exposure time was 20 s with 5 accumulations focused on the salivary specimen (placed inside by help of portion of a wet pellet placed between two glass slides before drying as a thin lm in a container lled with N 2 gas) by using a computer-controlled x-y-z stage to induce the Raman scattering with effect spectral acquisition of 600 lines/mm grafting. The scans were acquired obtained over the spectral regions from 1700 cm − 1 to 3200 cm − 1 (intensities of the amide I peak at 1740-1550 cm − 1 , the S-S stretch at 600-470 cm − 1 , and SH at 2,480-2,620 cm − 1 will be measured). Centroid cluster measurements were examined via three measurements at the same regions (1740 − 1550 cm − 1 , and 1248-1273 cm − 1, the S-S stretch at 600-470 cm − 1 and SH at 2480-2620 cm − 1 ). The measurements were done two times for each specimen to con rm the reproducibility of the technique. The intensity peaks at 1667 cm − 1 and 1246 cm − 1 , representative of the organic content by Amide I and III respectively, were taken as standards, to compare the readings and analyze the changes between smokers and non-smokers. The raw spectra were imported into MATLAB version (Math Works, Inc., Natick, MA) and baseline corrected by polynomial order.
This procedure was repeated with each subset reaching the validation to ensure the model was t for interpretation.

MMP-8 analysis
Around, 2 ml of saliva was collected and placed in a sterile saliva collecting tube according to the method mentioned in a previous paper [15]. The samples were initially stored at -75 o C. Salivary MMP-8 levels were analyzed using Quantikine human total MMP-8 immunoassay kits (R and D Systems, Inc., USA) by ELISA (Quantikine Human Total MMP-8 Immunnoassay; Catalog Number DMP800). Using a correction wavelength set at 540-570 nm, the absorbance was read at OD 450nm using a spectrophotometer.

Antioxidant enzymes and thiocyanate
The thiocyanate (SCN − ) levels were determined using spectrophotometric method where 1 ml of tricholoroacetic acid (20% v/v.) solution was mixed with diluted saliva sample (2 ml) to further precipitate the proteins. After performing centrifugation (4000 × g at 10 min), 1 ml of ferric nitrate-nitric acid reagent was added to a 2 ml supernatant. The absorbance was measured at OD 460nm using a standard protein assay kit (BIO-RAD, Hercules, CA, USA). The standard curve was obtained using serum albumin.
The peroxidase (Px) activity was measured by mixing 1 ml phosphate buffer (pH 7.0), 1 ml guaiacol solution, and 0.1 ml of a saliva sample. Later a stock solution of 1 ml H 2 O 2 was added and absorbance measured at OD 470nm . Glutathione-peroxidase (GSH-Px) levels were measured using RADSOD and RANSEL (Randox Laboratory Ltd., Ardmore, UK). The well of microplates was covered with monoclonal antibodies speci c to GSH-Px. The amount of GSH-Px was quanti ed by the enzymatic hydrolysis of pnitrophenyl phosphate and measured at OD 405nm in a microplate reader using a spectrophotometer.

Statistical analysis
The data were analyzed using SPSS software v. 20 (SPSS IBM statistics, Chicago, USA). All Raman spectroscopy measurements were tested using Fisher's exact probability test. The MMP-8, antioxidant enzymes, and thiocyanate levels were analyzed using two-tailed paired t-test. For all statistical tests, the signi cance levels was set at α = 0.05. Figure 1 shows Raman spectra of S-S stretch and Amide I derived from disulphide bonds of saliva. S-S stretch bands and corresponding Amide I, showed signi cant changes between the salivary samples while smoking and after quitting (p < 0.05). The Amide I peak in smoker's saliva (blue line) was signi cantly reduced as compared to non-smokers. The sulfhydryl bond, SS-stretch and the corresponding disulphide bond at 500-545 cm − 1 were signi cantly increased in smoking saliva (p < 0.05). As it can be seen in Fig. 1, the spectra of smoker's saliva are different from the spectra of saliva after quitting. The two intense peaks at 2440 cm − 1 , and 500 cm − 1 show marked changes in terms of reduction in intensities and are used as a measure of oxidation. The disulphide bridges are signi cantly higher in the smokers' saliva. Moreover, the peak position at 1310 cm − 1 in the non-smokers saliva proteins seemingly had shifted to 1318 cm − 1 in smoker's saliva proteins having lower intensity.

Raman spectroscopy
For the measurements of glycopeptides via Raman spectroscopy, the Amide I region shown in Fig. 1 was centred around 1665-1680 cm − 1 . The major peak in the Amide I region of is centred and assigned to βsheet which may be due to backbone, side chain or carbohydrate structures also, but this is not yet known. This investigation by Raman spectroscopy reported enabled peak assignment within the study and may also provide information about the structural characteristics of mucins within the saliva of smokers and non-smokers. The changes are seen amongst the peaks in terms of their glycosylation and secondary structure (Table 1).  Figure 2 represents the Raman spectra recorded in the region between 1365 and 1715 cm − 1 . The peaks at 1667 cm − 1 and 1446 cm − 1 are associated with organic components and well represented. The Amide I shows a drop in smoker's saliva (blue line) as compared to the non-smokers saliva representing C = O (carbonyl) stretch of the amide group, coupled with the in-plane N-H bending and C-N stretching vibration from the polypeptide chain in the protein and is hardly affected in the non-smokers saliva. On the contrary, the CH 2 wag showed higher intensity amongst smoker's saliva representing a recoil of the C atoms. The peaks at 1653 for smoker's saliva indicate polypeptide backbone of a protein due to amide and aromatic ring breathing. The mucin matrices are known to be observed by Raman shifts of the CH stretching peak at 1444 cm − 1 . These peaks are inherent to Raman spectra of saliva. The peak at 2067 cm-1 is very unique to saliva and likely caused by thiocyanate. The peak intensities are more signi cant (p < 0.05) as compared to spectrum of saliva from patients who had quit smoking (Fig. 3).
Smoking produced more signi cant changes in the peaks and showed higher intensities (p < 0.05).

Salivary MMP-8
The salivary MMP-8 levels ( Table 2) of participants after quitting smoking was signi cantly lower when compared to the salivary levels while smoking (p < 0.05). The smoking factor had a signi cant effect on the MMP-8 levels and the interaction of the two groups was also signi cant (p < 0.05).

Antioxidant enzymes and thiocyanate
Antioxidant values are summarised in Table 3. The SCN − and peroxidase levels were signi cantly higher in smokers' group (p < 0.05); whereas GSH-Px levels were signi cantly higher in the non-smoking group (p < 0.05). No signi cant correlation was found between SCN − and GSH-x (µg/mg) activities amongst smokers and non-smokers groups (p > 0.05) with signi cant differences between SCN − and peroxidase activities between the groups (p < 0.05).

Macrophages
The non-smokers groups saliva when exposed for 24 h increased IL6 and IL12 response indicating an M1 type polarization of the macrophages. The levels were signi cantly different when compared to the smoker's saliva for the human peripheral blood mononuclear cell line as the respective activity was lost (Fig. 4). The data represented salivary activity of smokers which is sensitive to M1 phenotype.

Discussion
Tobacco addiction and smoking are a chronic, relapsing condition and it is di cult to achieve long term abstinence for smokers [16]. Smoking becomes a highly conditioned behaviour, that can trigger relapse while attempts are made for quitting. In our study, it was demonstrated that recruiting patients for tobacco cessation and participation in a research study on tobacco control is a successful strategy. This further paved way for encouraging smokers to participate and stay engaged in the tobacco preventive and cessation programme. This study in particular is considered novel in measurement and analysis of salivary samples in a time period as minimum as 3 months after baseline with biochemical veri cation measures. The ndings of the spectroscopic and MMP-8 changes to other cessation studies maybe limited. No other studies were found in the literature with these two ndings for comparison amongst patients and participants reporting to the tobacco cessation clinics.
By using Raman spectroscopy, we con rmed the increased S-S stretch band from protein disulphide bonds in saliva when subjects were smoking. Smokers saliva have altered multimolecular mucin complex as previously reported verifying the detection of disulphide residues in our samples [17,18]. There is an oxidant/antioxidant imbalance of due to a signi cant decrease of GSH-Px activity in smoker's saliva [19]. From our results, the increase in protein oxidation in smoker's saliva, and a signi cant reduction seen amongst saliva of non-smoking status, we believe that enzymatic modi cations were minimal, and the tobacco cessation protocols had brought about a signi cant change in the patient's biochemical pro le.
There was a signi cant increase in the intensity of disulphide bond at 500-545 cm − 1 suggestive of change. Some differences between the two groups still existed, and the difference spectrum made them apparent. The characteristic vibrational modes are mainly dominated by the protein constituents of the salivary samples. The vibrational band analysis and assignments are done keeping in mind the group frequencies of the various analytes present in the sample [20]. Figure 1 shows the corresponding difference spectra of the different salivary groups revealing the signi cant spectral changes such as peak intensities, shoulder bands and speci c spectral ranges. The changes can be related certain vibrational bonds and clearly suggest that the Amide I peak in smoker's saliva (blue line) was signi cantly reduced as compared to non-smokers. These results suggest that there is a remarkable increase in the percentage of biochemical constituents among the saliva protein groups, indicating the diagnostic potential of Raman for saliva protein identi cation of the smoker's saliva. The reduction in Amide I band is a staggering difference when compared to the saliva in non-smoking status where the Amide I can gradually increase. These changes along with other spectra differences seen provide an understanding that there is less multimolecular mucin complex in the saliva of smokers. Although Amide I band decreases, the levels of disulphide bonds are different between smoker's saliva and non-smokers one. Further, in order to quantify the variations between smokers and non-smokers saliva, there was an increase in the CH 2 deformation vibration in smoker's saliva. In addition, in our parallel study, X-ray uorescence analysis showed different chemical elements in the saliva of smokers and non-smokers with biggest discrepancies found at Cr. Mn and Ca components in smoker's saliva (data not shown).
MMP-8 has been considered a key biomarker and major collagenolytic metalloproteinase in oral uids and its elevated levels have been reported to have been seen amongst periodontitis patients. In this particular study, the quantitative direct assesment of MMP-8 showed that the levels were signi cantly increased and lowered in smoking and non-smoking salivary specimens respectively (p < 0.05). This is primarily due to the fact that there is impairment of granulocyte function which is receptive to microbial changes too delivering excessive elastase, proteases and MMP-8 [21]. Therefore, the second null hypothesis that smoking may not have any effect on MMP-8 levels has to be rejected. It is well known that reactive oxygen species may signi cantly lower proteolytic activity and MMP-8 within the salivary specimens of smokers by inactivating latent-pro MMPs and MMP-8 fragment [22]. The strong association between smoking and periodontitis is also well established and might be reinforced by this point. Therefore, smoking may interfere with the levels of MMP-8 and should be considered when using saliva for diagnosis.
The spectral components found as a result of Raman are complex and the assignments are based on known salivary composition and literature [23]. Although the concertation of GSH-Px and Px levels were determined, it does not always relate to enzymatic activity within the saliva. The SCN − levels were signi cantly high (p < 0.05) and a signi cant correlation between SCN-and Px (U/mg) was found. This is a clear indication that the oxidative stress as a result of smoking increased the levels of oxidases. Px levels were remarkably lower in smokers is supported by the fact that Px serves as a protective defense mechanism where H 2 O 2 is converted into molecular oxygen and gets detoxi ed by the cigarette smoke [24]. It is also possible that GSH-Px levels were affected by smoking levels. The macrophage result was a clear indication that non-smokers saliva provoked the polarization toward the proin ammatory M1 phenotype. These ndings are consistent with the fact that saliva also invokes a proin ammatory response in oral broblasts [25]. CD14 is present in human saliva mainly on the surface of monocytes and macrophages [26]. However, M1 polarization by saliva likely occurs independently of soluble CD14. It is also known that the lipopolysaccharide that can induce M1 polarization [27]. It is reasonably safe to speculate that the salivary endotoxins can increase the expression of IL12 and IL6 by macrophages and causing them to develop a proin ammatory M1 phenotype.
A comprehensive effort to reduce tobacco-related disparities includes clinical cessation interventions, accessibility, social acceptability, as well as population-level measures to reduce the appeal of tobacco [28]. In our study, one of the limitations was a relatively short follow-up period of three months. It is known that smoking habits tend to come back within this time period [29]. There was no dropping out of participants and were quite motivated. Further efforts are being made to expand the program and improve its effectiveness.

Conclusion
Based on the present study results, it can be concluded that the salivary biochemical analysis can aid clinicians to verify smoking cessation in patients under smoking cessation programs.  Figure 1 Raman spectra of S-S stretch and amide I derived from disulphide bonds of saliva. The black line corresponds to saliva of patients who had quit smoking against the blue line, of same patients who were smoking. S-S stretch bands and corresponding, -SS-/protein ratio (red dotted line; arrow), was signi cantly higher in the smokers' saliva than in all others (p<0.05). The Amide I peak in smoker's saliva (blue line) was signi cantly reduced as compared to non-smokers. The selected range of Raman spectra of smokers and non-smokers saliva where the peaks at 2065 cm-1-2067 cm-1 is unique to saliva likely caused by thiocyanate.