The Relationship between Plasma and Salivary NO_x

Abstract

Several studies have shown that fasting plasma nitrite (NO₂⁻) is an indicator of endothelial nitric oxide synthase (NOS) activity while plasma nitrate (NO₃⁻) or the sum of NO₂⁻ and NO₃⁻ (NO_x) do not reflect NOS function. Plasma NO₂⁻ can also be elevated through dietary NO₃⁻ where the NO₃⁻ is partially reduced to NO₂⁻ by oral bacteria and enters the plasma through the digestive system. NO₃⁻ is taken up from plasma by salivary glands and the cycle repeats itself. Thus, one may propose that salivary NO₂⁻ is an indicator of plasma NO₂⁻ and consequently of NO production. Many brands of nitric oxide (NO) saliva test strips have been developed that suggest that their product is indicative of circulatory NO availability. However, data supporting a relationship between salivary and plasma NO₂⁻ or NO bioavailability is lacking. Here we have measured basal salivary and plasma NO₂⁻ and NO₃⁻ to determine if any correlation exists between these in 13 adult volunteers. We found no significant correlation between basal salivary and plasma NO₂⁻. Also no correlation exists between salivary NO₃⁻ and plasma NO₂⁻. However, we did see a correlation between salivary NO₃⁻ and plasma NO₃⁻, and between salivary NO₂⁻ and plasma NO₃⁻. In a separate study, we compared the efficiency of salivary NO₃⁻ reduction with the efficacy of increasing plasma NO₃⁻ and NO₂⁻ after drinking beet juice, a high NO₃⁻-containing beverage, in 10 adult volunteers. No significant correlation was observed between the ex vivo salivary reduction of NO₃⁻ to NO₂⁻ and plasma increases in NO₃⁻ or NO₂⁻. These results suggest that measures of salivary NO₃⁻, NO₂⁻ or NO_x are not good indicators of endothelial function. In addition, the efficiency of saliva to reduce NO₃⁻ to NO₂⁻ ex-vivo does not demonstrate one’s ability to increase plasma NO₂⁻ following consumption of dietary NO₃⁻.

1. Introduction

Determining a diagnostic marker of endogenous NO bioavailability has been a major topic of research, one which would have clinical implications for monitoring cardiovascular disease, metabolic syndrome, and other conditions [1; 2; 3; 4]. Endothelial dysfunction has been noted as a key event in early atherosclerosis. Due to defective synthesis, decreased levels of endothelium-derived NO characterize endothelial dysfunction [5]. In addition, increased scavenging of NO by oxygen radicals results in low NO bioavailability associated with endothelial dysfunction [6; 7]. Individuals with endothelial dysfunction show a decrease in flow-mediated dilation (FMD) and an increase of intima media thickness (IMT), both representative of atherosclerosis [8; 9; 10; 11]. Given the cost of procedures to measure FMD and IMT, it would be useful to establish a simple blood test to diagnose endothelial dysfunction or low NO bioavailability due to other conditions. The Kelm and Moncada labs have demonstrated that the majority of plasma NO₂⁻ is derived from constitutive NOS-activity [12; 13]. Observations have shown that upon regional nitric oxide synthase (NOS) inhibition in forearm circulation, vascular resistance increases linearly as plasma NO₂⁻ levels decrease, thereby establishing plasma NO₂⁻ as a potential measure of endothelial function [12].

In some of the same studies, plasma NO₃⁻ was ruled out as an indicator of NOS function. The Moncada lab showed that only 16% of isotopic L-arginine infused into circulation was represented in plasma NO₃⁻ levels versus 90% of plasma NO₂⁻ [13]. The Kelm laboratory demonstrated no significant change in plasma NO₃⁻ in mammals with inhibition of NOS activity [12]. These data are somewhat expected since plasma NO₃⁻ has many NOS-independent factors which can drastically change the basal levels such as dietary NO₃⁻ intake, denitrifying liver enzymes, and renal function [14; 15]. Interestingly, Hibbs and colleagues showed that inducible-nitric oxide synthase (iNOS) is one of the main contributors to circulating NO₃⁻ due to the increase of NO production after the addition of the cytokine-IL2 [16]. However, the general consensus is that, under most conditions, plasma NO₃⁻ does not reflect NOS function or NO bioavailability. In addition, there are even some investigators who suggest that plasma NO₂⁻ does not accurately reflect eNOS function and NO bioavailability [17].

In recent years, dietary NO₃⁻ has become a known contributor to the pool of bioavailable NO [18]. It is known that dietary NO₃⁻ is reduced in the oral cavity by tongue flora, specifically by Actinomyces and Veilonella species [19; 20]. Once ingested, NO₂⁻ is non-enzymatically reduced to NO in the gastric acidic milieu [21]. NO₃⁻ and remaining NO₂⁻ are rapidly absorbed in the small intestine. Plasma NO₂⁻ can then be reduced to NO by various mechanisms [18; 22]. Though most of the circulating NO₃⁻ is excreted in urine, approximately 25% is extracted by the salivary glands and recycled through the enterosalivary circulation [23]. Complementary to endogenous NO production, this cycle of dietary NO₃⁻ being converted to NO in physiology is referred to as the nitrate-nitrite-nitric oxide pathway [18].

Through this physiological pathway, it has been shown that dietary NO₃⁻ will increase plasma NO₃⁻ and NO₂⁻. In addition, dietary NO₃⁻ has been shown to lower blood pressure with short and long term effects, be vasoprotective and reduce platelet aggregation, along with having acute effects on cerebral blood flow and an increase in exercise tolerance and performance [18; 24; 25; 26; 27; 28; 29; 30]. Daily dietary NO₃⁻ ingestion also improves endothelial function and vascular stiffness in hypercholesterolemia [31; 32; 33; 34].

As evidence suggests that basal plasma NO₂⁻ levels reflect NOS function and bioavailable NO, these measurements may have clinical utility. However, based on the nitrate-nitrite-NO cycle, one may also suggest that salivary NO₂⁻ could have the same utility, as recently pointed out [1]. Indeed, commercially available products exist that measure salivary NO₂⁻ and claim to report NO bioavailability. However, until now, no published studies have shown a positive correlation between basal plasma NO₂⁻ and salivary NO₂⁻ levels. Thus, in this work, we sought to investigate the basal levels of plasma and salivary NO₂⁻ and NO₃⁻.

When studying increased plasma NO₂⁻ after a high NO₃⁻ load, Lundberg et al. observed attenuation after using an antibacterial mouthwash [35], suggesting saliva’s conversion from NO₃⁻ to NO₂⁻ greatly affects plasma NO₂⁻. Consumption of high NO₃⁻-containing food or drinks increases plasma NO₃⁻, NO₂⁻- and thus NO_x. However, many studies have observed a significant variation with the increase in plasma levels among individuals [13; 24; 25; 28; 35]. It appears that some individuals are poor or non-responders with respect to dietary NO₃⁻ interventions as measured by increases in plasma NO₂⁻[24; 27; 36; 37]. It would be useful to easily determine individuals’ efficacy at converting oral NO₃⁻ to plasma NO₂⁻. In this study we hypothesized that saliva would reflect ability to convert NO₃⁻ to NO₂⁻ in the oral cavity and that this ability would correlate with an individual’s dietary NO₃⁻ to plasma NO₂⁻ conversion efficacy. Thus we conducted a second study where we examined both ex-vivo conversion of NO₃⁻ to NO₂⁻ in saliva as well as in vivo conversion of dietary NO₃⁻ to plasma NO₂⁻.

2. Methods

2.1 Study Design

All human subjects use was approved by an internal review board following federal and institutional guidelines. For the basal levels study, 13 volunteers (8 male and 5 female) participated between the age of 18 and 80 years old. Volunteers reported to the lab at 9:00 am on the day of their participation. Individuals did not eat or drink within two hours of their participation. If the volunteers had eaten any food the morning of the sampling, they were told to avoid any high NO₃⁻ foods (ex. spinach, beets, lettuce, and other green leafy vegetables). In addition, volunteers did not use mouthwash but were permitted to brush their teeth. Upon arrival, blood was drawn from each volunteer from an antecubital vein and collected in a 4 mL lithium heparin vial. Simultaneously, volunteers expectorated 5 mL of saliva which was collected in a sterilized 50 mL Corning tube.

The beet juice study was ancillary to a larger study aimed at investigating potential additional benefits of beet juice combined with supervised exercise compared to supervised exercise alone. The larger study provided a great opportunity for the ancillary one discussed here to compare in vivo conversion of oral nitrate to plasma nitrite and ex vivo salivary conversion efficiency. 10 participants (5 male and 5 female) above the age of 55 were recruited. All recruits consumed a bottle of concentrated beet juice (Beet it Sport shot, 500 mg NO₃⁻) a day for 6 weeks. On the first day of weeks 1, 3 and 6 participation, each recruit came in for sampling. As described for the study above, volunteers did not use mouthwash or eat any high NO₃⁻ foods but were permitted to brush their teeth. Blood was drawn before and 1 hour after beet juice consumption. Immediately before the blood draw, participants expectorated a 5 mL saliva sample into a 50 mL sterilized Corning tube. One plasma sample was excluded due to hemolysis during sample preparation. In addition, two anaerobic saliva samples were excluded due to their having dried out during the deoxygenation procedure.

2.2 Measurements

Since salivary NO strips claim to be indicative of physiological NO, we sought to determine what these strips actually test in saliva. Two brands of NO test strips, Nitric Oxide Test Strips (Berkeley Test) and Nitric Oxide Indicator Strips (Neogenis; Austin, TX), were placed in solutions of NaNO₃⁻ and NaNO₂⁻. Since the strips have a colorimetric indicating tip, concentrations were varied in order to darken the strip with more reactant.

Similar measuring techniques were used for both basal and beet juice studies. Blood was taken from an antecubital vein and collected in a 4 mL Lithium heparin vial. The tubes were immediately centrifuged at 11,500 g for 2 min. The supernatant plasma was removed and immediately frozen on dry ice in aliquots of ~0.4 mL of plasma and stored in a −80 °C freezer. Plasma NO₃⁻ and NO₂⁻ were determined and labeled “basal levels.” For the beet juice studies, plasma NO₂⁻ levels were determined in the before- and 1 hr after-beet juice consumption timepoints. The difference was reported as the Δ plasma NO₂⁻.

Plasma NO₃⁻ and NO₂⁻ levels were measured using an HPLC-based Eicom NOx Analyzer, model ENO-20 according to instructions of the manufacturer. For all measurements, standard curves were obtained and used for quantitative measurements.

For the basal levels study, each tube of saliva was centrifuged at 11,500 g for 5 minutes to exclude a large pellet of bacteria. The supernatant salivary matrix was removed and immediately frozen on dry ice in aliquots of ~1.0 mL of saliva until time of analysis. Freezing saliva is a method tested and employed by our lab which maintains the integrity of the sample after exclusion of the bacterial pellet[38; 39]. Before instrumental analysis, saliva was thawed and mixed 1:1 by volume with methanol for deproteination (a method tested which maintains NO₃⁻ and NO₂⁻ but allows for a cleaner sample to eliminate possible syringe clogging of the HPLC system of the NO_x analyzer). Salivary NO₃⁻ and NO₂⁻ were determined and labeled “basal levels.”

For the beet juice study, each tube of saliva (5 mL) was immediately split into two samples, one designated for aerobic testing and the other for anaerobic testing. Both samples were placed in a round bottom flask and left to incubate in a 37 °C water bath. The anaerobic sample was capped and flushed with argon for 1 hr to deoxygenate and induce anaerobic activity. After 1 hr of incubation, 10 mM NaNO₃⁻ was added to each sample to mimic a high NO₃⁻ beverage. Starting with a 0 min time point, the NO₂⁻ produced was detected at 10 minute increments up to 90 minutes. The salivary rate (k) of NO₂⁻ production was calculated and reported in μM/min.

Salivary NO₃⁻ and NO₂⁻ levels were measured using chemiluminescence-based Nitric Oxide Analyzers (Sievers, Inc.) according to instructions of the manufacturer. For all measurements, standard curves were obtained and used for quantitative measurements.

2.3 Statistical Analysis

For the basal levels study, salivary NO₃⁻ and NO₂⁻ levels are reported as the mean value of three injections. Plasma NO₃⁻ and NO₂⁻ are reported as the sole value measured. For the beet juice study, the aerobic and anaerobic salivary rate of NO₂⁻ production was plotted against the change in plasma NO₂⁻. A linear trendline was added to each plot to show relative correlation. Spearman correlation coefficients (r) were calculated to measure the linear correlation between the two variables plotted. An r value less than −0.5 or greater than 0.5 was considered significant. In addition, two-tailed p-values were calculated. A p-value less than 0.05 was considered significant.

3. Results

We first investigated sensitivity of oral commercial test strips to NO₂⁻ and NO₃⁻. The salivary NO test strips reacted with NaNO₂⁻ in solution, but not NaNO₃⁻ (Fig 1). There was no colorimetric change when the strips were submerged in NaNO₃⁻ solution of 1 and 10 mM concentrations. When the strips were submerged in NaNO₂⁻, a very apparent colorimetric change was seen. No detectable reaction was observed using the low concentration of 10 μM NaNO₂. A light pink color was apparent at a concentration of 50 μM and the color darkened with increasing concentrations up to 10 mM NaNO₂⁻.

Fig. 1.

Sensitivity of test strips to NO₂⁻ and NO₃⁻. Nitric Oxide Saliva Test Strips, Neogenis (A) and Berkeley (B), were dipped in a NaNO₃⁻ or NaNO₂⁻ solution. New unused test strips have a white tip where saliva is intended to be placed. Reaction with the strip caused a light pink to deep red colorimetric change.

As these test strips appear to measure salivary NO₂⁻, then, assuming plasma NO₂⁻ reflects NO bioavailability, the strips would measure NO bioavailability as long as salivary NO₂⁻ correlates with plasma NO₂⁻. Thus, we compared basal salivary NO₂⁻ to plasma NO₂⁻. We found that salivary NO₂⁻ in volunteers ranged from 51 to 257 μM while salivary NO₃⁻ ranged from 9 to 681 μM. Plasma NO₂⁻ in volunteers ranged from 60 to 184 nM while plasma NO₃⁻ ranged from 19 to 61 μM. Importantly, plasma NO₂⁻ did not significantly correlate with salivary NO₂⁻ (Fig 2A, p = 0.87). Plasma NO₃⁻ significantly correlated with salivary NO₃⁻ (Fig 2B, p = 0.05). Plasma NO₃⁻ significantly correlated with salivary NO₂⁻ (Fig 2C, p = 0.02). Basal plasma NO₂⁻ did not correlate with basal salivary NO₃⁻ (Fig 2D, p = 0.41).

Fig. 2.

Examination of basal salivary vs plasma levels of NO₂⁻ and NO₃⁻ (A) Basal plasma NO₂⁻ and salivary NO₂⁻ levels, p=0.87. (B) Plasma NO₃⁻ and salivary NO₃⁻, p=0.048. (C) Plasma NO₃⁻ and salivary NO₂⁻, p=0.020. (D) Plasma NO₂⁻ and salivary NO₃⁻ levels, p=0.41.

For the second study we investigated whether ex-vivo conversion of NO₃⁻ to NO₂⁻ in saliva correlates with in vivo conversion of dietary NO₃⁻ to plasma NO₂⁻. Changes in plasma NO₂⁻ ranged from −270 to 740 nM. The average increase in plasma NO₂⁻ was 150 nM. The rates of salivary NO₂⁻ production ranged from 0 to 201 μM/min in aerobic saliva and 0 to 402 μM/min in anaerobic saliva. There was no correlation between the rate of aerobic salivary NO₂⁻ production and the change in plasma NO₂⁻ (Fig 3A). The trendline R² value was 0.0240. The calculated r and p-values were 0.02 and 0.23, respectively. In addition, there was no correlation for the rate of anaerobic salivary NO₂⁻ production versus the change in plasma NO₂⁻ (Fig 3B). The trendline R² value was 0.0003. The calculated r and p-values were 0.08 and 0.76, respectively. Each plot includes all of the week 1, 3 and 6 time points for each individual. It is worth noting that there was no correlation when comparing only week 1, 3 and 6 time points individually, aerobic or anaerobic salivary reduction of NO₃⁻ (data not shown).

Fig. 3.

Comparison of ex vivo salivary conversion of NO₃⁻ to NO₂⁻ to in vivo conversion of oral NO₃⁻ to plasma NO₂⁻. Patients (n = 10) were given 2.35 oz of concentrated beet juice (Beet-It) each day for six weeks. Blood was drawn from each patient before and 1 hr after (Δ Plasma NO₂⁻) beet juice consumption at weeks 1, 3, and 6 (all 3 timepoints plotted for each individual). Saliva samples were expectorated early each morning (before breakfast or tooth brushing) at weeks 1, 3, and 6. Δ Plasma NO₂⁻ was plotted versus the rate of salivary NO₂⁻ production after 10 mM inorganic NO₃⁻ was added to (A) aerobic (p-value=0.23, corr=0.02) and (B) anaerobic saliva (p-value=0.76, corr=0.08). Neither condition shows a correlation.

4. Discussion

We have investigated potential methods to assess individuals’ basal NO bioavailability and a potential salivary-based method to assess the individuals’ ability to convert dietary NO₃⁻ to plasma NO₂⁻. Our major findings were (i) Nitric Oxide Test Strips indicate a relative level of NO₂⁻ in saliva, (ii) basal plasma and salivary NO₂⁻ do not correlate, and (iii) ex vivo conversion of salivary NO₃⁻ to NO₂⁻ is not an indicator of in vivo dietary NO₃⁻ conversion to plasma NO₂⁻. We also found that plasma NO₃⁻ correlated with both salivary NO₃⁻ and NO₂⁻. As expected, there was no correlation between plasma NO₂⁻ and salivary NO₃⁻.

As the Kelm lab has demonstrated, plasma NO₂⁻ may be indicative of endothelial function [8]. If salivary NO₂⁻ reflected plasma NO₂⁻, it could be an easy way to measure endothelial function, as suggested recently [1]. However, we found no relation between basal and salivary NO₂⁻ (Figure 2A). These data suggest that salivary NO₂⁻ would not be a valid way to measure endothelial function and NO bioavailability. In addition, as we found that commercially available test strips measure salivary NO₂⁻ (Figure 1), those strips are not likely to accurately assess NO bioavailability. It should be noted that we used the test strips in a setting where liquid was applied to them. Future work may consider whether other employments, such as putting the strip directly on the tongue, give different results.

Our study for basal levels was designed primarily to determine if physiological levels of salivary NO₂⁻ and plasma NO₂⁻ correlated. Without recently using mouthwash, consuming a high NO₃⁻ food or beverage, or medications, salivary and plasma NO₂⁻ in volunteers should reflect normal physiological levels. We observed no correlation between basal salivary and plasma nitrite levels. Interestingly, plasma NO₃⁻ correlated with both salivary NO₃⁻ and NO₂⁻ (Fig 2B & 2C). As noted by the Kelm laboratory, plasma NO₃⁻ is influenced by many NOS-independent factors such as dietary NO₃⁻ intake, saliva formation, and bacterial synthesis in the bowel [12]. The positive correlations between plasma NO₃⁻ and salivary NO₂⁻ and NO₃⁻ could be understood in terms of the nitrate-nitrite-NO cycle [18]. Salivary NO₃⁻ is converted to salivary NO₂⁻ by oral bacteria and thus, these levels correlate with each other [18; 19; 20; 22; 23]. NO₃⁻ enters the plasma and is taken back up into saliva [18; 22; 23]. Thus salivary and plasma levels of NO₃⁻ are linked. However, under basal conditions, plasma NO₂⁻ levels are largely due to eNOS function [40] and thus do not correlate with salivary NO₃⁻ or NO₂⁻. Indeed, we did not see a correlation between plasma and salivary NO₂⁻ or NO₃⁻.

To further rationalize the lack of a correlation between plasma and salivary NO₂⁻ levels, it is important to consider nitrogen oxide chemistry occurring during consumption and in the gut. As suggested by work looking at nitrogen oxide production in the oral cavity, it is likely that salivary NO₂⁻ is converted to further reduced nitrogen oxides such as NO, nitrous oxide, ammonia, and/or nitrogen gas [41; 42] and also reactive nitrogen oxide species [43] by bacteria which make up the human oral microbiome [44]. Additionally, the acidic milieu of the gastric lumen will likely affect the levels of plasma NO₂⁻ as evident in research showing increased NO levels after swallowing NO₂⁻ containing saliva [21; 22]. Thus, variations in factors that influence gastric pH and nitrogen oxide absorption as well as those that influence chemistry in the oral cavity could lead to additional variations in plasma NO₂⁻ that are not reflected in salivary NO₂⁻.

The beet juice study was designed for exploring a correlation between in-vivo conversion of dietary NO₃⁻ to plasma NO₂⁻ and the efficacy of ex-vivo salivary NO₃⁻ conversion. Several labs have shown that dietary NO₃⁻ increases plasma NO₂⁻ substantially [23; 24; 25; 26]. However, the efficacy of this conversion varies dramatically among individuals, perhaps due to their microflora [13; 24; 25; 28; 35]. It would be useful to be able to easily assess individuals’ ability to convert dietary NO₃⁻ to plasma NO₂⁻. We hypothesized that the rate of ex-vivo salivary NO₃⁻ to NO₂⁻ conversion would serve this purpose. However, we saw no correlation between the in vivo efficacy of converting dietary NO₃⁻ to plasma NO₂⁻ and the ex-vivo conversion rate. This lack of correlation may be due to the bacteria in the saliva samples we collected not being representative of the oral bacteria that are mainly responsible for NO₃⁻ to NO₂⁻ conversion in vivo. These bacteria are thought to form biofilms[19; 20] so that saliva may not necessarily reflect their number, distribution, or reflect optimum conditions for NO₃⁻ reduction.

Possible limitations to this study should be acknowledged. For the basal and beet juice studies, the sample sizes were 13 and 10 respectively. A larger sample size might show stronger correlations. For both studies we also used ex vivo salivary samples. As mentioned we used expectorated saliva to assess salivary NO₃⁻ to NO₂⁻ conversion. Tongue scraping or swabbing or NO₃⁻ to NO₂⁻ conversion in the oral cavity itself (gargling with NO₃⁻ and then expectorating) might show better correlations with in vivo dietary NO₃⁻ to plasma NO₂⁻ conversion. Plasma samples in the beet juice study were taken before and 1-hr after beet juice consumption. We have previously seen that plasma NO₂⁻ levels are close to maximum one hour after consumption [24]. However, others [25; 26] have found that plasma NO₂⁻ does not reach a semi-steady maximum value until three hours after consumption so that may have been a better choice for sampling. It’s possible that several time points (such as at 1, 2 and 3 hours after consumption) are needed to obtain the maximum plasma NO_x levels after a high NO₃⁻ food or beverage for all individuals.

5. Conclusions

Our results argue against use of salivary NO₂⁻ as a marker for NO bioavailability. In addition, the efficacy of ex-vivo NO₃⁻ to NO₂⁻ conversion cannot be used to assess individuals’ ability to convert dietary NO₃⁻ to plasma NO₂⁻.

Highlights.

• The correlation of basal plasma and salivary NO_x are investigated

• Ex-vivo salivary NO₃⁻ reduction is compared to in-vivo dietary NO₃⁻ to plasma NO₂⁻

• We report that basal salivary and plasma NO₂⁻ levels show no correlation.

• Salivary NO₃⁻ to NO₂⁻ does not correlate with in vivo dietary NO₃⁻ to plasma NO₂⁻

Abbreviations

NOS

nitric oxide synthase

FMD

flow-mediated dilation

iNOS

inducible nitric oxide synthase

IMT

intima media thickness