Background
Serotonin (5-HT) is a monoamine neurotransmitter synthesized in both the central nervous system (CNS) and in the periphery. Serotonin is synthesized from the amino acid tryptophan by the enzyme tryptophan hydroxylase (TPH) and it is metabolized by the enzyme monoamine oxidase (MAO) into 5-hydroxyindoleacetic acid (5HIAA). In the CNS, serotonin is synthesized in the neurons of the midbrain raphe nuclei (centrally produced serotonin) and in the periphery in the enterochromaffin cells (ECCs) of the gut. The ECC produces almost 95% of the body’s total serotonin, which overflows into the gastrointestinal lumen and blood. Very little free serotonin can be found in plasma, however, as platelets concentrate and store the vast majority of peripheral serotonin.Reference Berger, Gray and Roth1,Reference Muck-Seler and Pivac2
The rate-limiting step in serotonin synthesis is catalyzed by two genetically distinct isoforms of TPH. TPH2 is preferentially expressed in the brain, while TPH1 is expressed peripherally and in the pineal gland, where serotonin serves as precursor for melatonin production. Thus, serotonin biosynthesis is spatially and biochemically segregated between the periphery and the CNS by the blood–brain barrier.Reference Walther, Jens-Uve and Saleh3,Reference Zhang, Beaulieu and Sotnikova4 The gold standard measurement of central serotonin is its metabolite, 5-HIAA in cerebrospinal fluid (CSF) via lumbar puncture (LP).Reference Wester, Bergstrom and Eriksson5
Serotonin is stored intracellularly in vesicles and granules and is released by exocytosis. Reuptake of neuronal serotonin across presynaptic plasma membrane is mediated by the serotonin transporter (SERT), a member of the SLC6 sodium-dependent solute transporter family (SLC6A4).Reference Blakely, DeFelice and Galli6 SERT is highly expressed in platelets that do not synthesize serotonin, but store serotonin produced by the ECC.
Serotonin plays a diverse role in the regulation of cardiovascular function, breathing, circadian rhythms, sleep, appetite, mood, and learning.Reference Berger, Gray and Roth1 It influences a wide range of psychiatric disorders, and serotonin has been targeted for many pharmacologic approaches to managing depression and anxiety.Reference Baldwin and Rudge7 Serotonin and its numerous receptor subtypes are involved with behavioral and neuroendocrine processes that provide life-preserving responses to stress. Serotonin plays an important role in vascular resistance and blood pressure control, and a functional SERT polymorphism has been associated with migraine.Reference Schurks, Rist and Kurth8 Serotonin also has been the topic for research in neurodegenerative disorders, spinal cord injury, and sudden unexpected death in epilepsy.Reference Azmitia and Nixon9–Reference Richerson11 In addition, prior to assuming its role as a neurotransmitter, serotonin acts as a trophic signal in the developing brain by directing neural ontogeny of the 5-HT and other neurochemical systems. As a developmental signal, serotonin regulates cell migration, axonal outgrowth, and synaptogenesis, and is a coordinating mediator of intracellular communication. It is therefore conceivable that changing levels of serotonin during prenatal development may alter subsequent structure and function of the serotonergic system.Reference Gaspar, Cases and Maroteaux12
The aim of the current study was to find an easy, noninvasive peripheral biomarker that correlates with central serotonin turnover. The clinical significance of this aim would be to avoid invasive LPs for diagnosis and therapeutic monitoring of patients with primary neurotransmitter deficiencies, such as biopterin defects, for example, GTP cyclohydrolase deficiency, 6-pyruvoyltetrahydropterin synthase deficiency, and sepiapterin reductase deficiency.Reference Hyland13 Patients with primary neurotransmitter deficiencies are diagnosed by measuring CSF serotonin and dopamine metabolites, but most importantly need serial LPs to monitor the effects of treatment, such as serotonin replacement therapy with 5-Hydroxytryptophan.
Methods
University Research Ethics Board approval was obtained for enrolling patients and performing the study (IRB#H14-01822). All children between 6 months and 17 years, evaluated in Pediatric Neurology and Biochemical Genetics clinics, for a variety of neurological presentations, including developmental delay and seizures were recruited in a nonconsecutive manner. Only children whose routine clinical care included an LP for CSF collection and neurotransmitter metabolite analysis were included. An additional salivary and blood sample was obtained at the time of LP. Only patients whose LP was unsuccessful were excluded. No LP was performed on children for the sole purpose of this study.
LP was performed in a fasted state, in the morning. Fractions were collected in separate tubes according to protocol, immediately frozen and analyzed according to established methods in a clinical laboratory using previously reported methods.Reference Hyland13,Reference Hyland, Surtees and Heales14
Blood sample was collected in Na citrate tubes, then submitted to a soft spin at 140 g × 10 minutes. The supernatant was removed to obtain platelet-rich plasma and frozen at −80°C until further analysis. All experiments were conducted in clear plastic tubes and at room temperature. Approximately 0.2–0.5 ml saliva samples were collected using a clean pipette from the mouth, frozen immediately and kept in −80°C until further analysis. Platelet and salivary serotonin were measured using a highly sensitive Enzyme Immunoassay kit (DLD Diagnostika GMBH, Catalog# 630/96).
Analysis: Pearson’s coefficient of correlation (PCC) was calculated for pairwise comparisons of saliva, plasma, and CSF serotonin. Significance was set at α = 0.05, using a two-tailed test.
Results
Twenty-six patients were recruited, of whom 18 had suitable samples and were included for analysis. The age of patients ranged from 6 months to 15 years with an average age of 9.1 years (standard deviation 4.9), and were 57% male (Table 1).
Table 1: Results of platelet, saliva serotonin, and CSF 5HIAA levels
Reference ranges for CSF 5HIAA age related: 6–9 years: 66–338 nmol/L; 10–14 years: 67–189 nmol/L; 15 years: 67–140 nmol/L.
Twelve patients had CSF 5HIAA level in the normal range for their age. Five patients had values below normal, and one patient had a supranormal value (Table 1).
There was no correlation between platelet serotonin and CSF 5HIAA levels (PCC:0.010, p = 0.969) or between salivary serotonin and CSF 5HIAA (PCC: 0.258, p = 0.318). There was a strong negative correlation between salivary and platelet serotonin (PCC: −0.679, p = 0.002) (Figure 1).
Figure 1: Scatter plots of CSF, saliva, and serum serotonin. Regression line of fitted values. (A) CSF serotonin vs. platelet serotonin, PCC = 0.01, p = 0.97. (B) CSF serotonin vs. saliva serotonin, PCC = 0.258, p = 0.32. (C) Saliva vs. platelet serotonin, PCC = −0.68, p = 0.002.
Discussion
We know the source of serotonin in the brain and in blood. There is no clear data on the origins of salivary serotonin, although we know that many of the catecholamines and hormones are readily detectable in saliva, but other than steroid hormone levels, their quantification doesn’t correlate very well with plasma levels. Most steroids can cross readily the phospholipid membrane of epithelial cells lining the blood vessels by simple diffusion, but this has not been proven for serotonin.Reference Kennedy, Dillon, Mills and Ziegler15 Serotonin has opposing effect on salivary glands and on secretion of saliva. Serotonin treatment decreases salivary flow rate and protein content. This is a receptor-mediated action of 5-HT, by increasing intracellular cAMP accumulation.Reference Vaibhav and Marshall16 This is the most likely explanation for the negative correlation between platelet and salivary serotonin. Increase in blood serotonin will decrease salivary flow rate and protein content.
Saliva is a clinically informative biofluid that is used for novel approaches for diagnosis, monitoring, and management of patients with systemic disease.Reference Malamud and Rodriguez-Chavez17 Such salivary biomarkers are cortisol, nitrite, uric acid, sodium, chloride, pH, amylase, and lactoferrin used in end-stage renal disease; cancer antigens in malignancies; and specific antibody reactivity to pathogens in infectious diseases.Reference Malamud and Rodriguez-Chavez17 It is easily collected and stored and would be ideal for use as a peripheral biomarker.
Although we don’t know the exact mechanism by which serotonin enters the oral cavity, it appears that it is mostly influenced by the blood serotonin levels and doesn’t appear to be a good peripheral biomarker for central serotonin turnover (Figure 2).
Figure 2: Illustration of sources of central and peripheral serotonin and their correlation.
Previous experiments have sought a noninvasive surrogate marker of central serotonin, including serotonin in whole blood, plasma, platelets, saliva, and urine. An alternative, noninvasive method of neurotransmitter assessment could eliminate the need for painful and resource-intensive LP and increase the possibility of quick, routine testing and monitoring of central serotonin levels.Reference Sarrias, Cabre, Martinez and Artigas18–Reference Audhya, Adams and Johansen21 Differences in study design, however, yield conflicting results. The mechanism responsible for any observed correlations between centrally and peripherally measured serotonin remains to be elucidated.
CSF 5HIAA has been used in this study instead of CSF 5-HT (serotonin) as it has been shown to reflect central serotonin turnover.Reference Wester, Bergstrom and Eriksson5 The distribution of 5-HT and its metabolite 5HIAA was similar and correlating in a study comparing ventricular CSF monoamine transmitter and metabolite concentrations in 14 autopsy cases.Reference Wester, Bergstrom and Eriksson5 It is used as a standard measurement for diagnosing all primary neurotransmitter disorders.Reference Hyland13 On the other hand, plasma and salivary 5HIAA levels are more variable and highly fluctuating in response to high dietary intake of biogenic amine foods (bananas, pineapple, plantain, and kiwi), in contrast with platelet serotonin which does not show significant change.Reference Kema, Schellings and Meiborg22
There have been conflicting reports comparing CSF serotonin with plasma or platelet serotonin, and our study corroborates that there is no correlation among these measures. We have measured CSF 5HIAA, as marker for central serotonin turnover, which closely correlates with CSF 5-HT. We think that this did not influence the results of the study. One of the limitations of the study is that the cohort was too small for meaningful statistical analysis. However, when we plotted the data (figure not shown), the correlation made sense.
Some of the pitfalls in previously reported 5-HT measurements that may have not been accounted for and some of our precautions to mitigate for these pitfalls were:
1. A diet rich in tryptophan (bananas, chocolate, pineapples, plums, and nuts) will increase the levels of plasma and to lesser degree that of platelet serotonin. We collected all samples after a 12-hour overnight fast, to avoid even the small influence of diet on platelet serotonin.
2. Time of the day for collection is important, as 5-HT levels are highest in the morning in plasma. We collected all samples between 9 and 11 AM.
3. Certain medications (i.e. SSRIs, MAO inhibitors) will increase central serotonin levels.Reference Anderson23 None of our patients were taking these medications.
4. Technical challenges (using glassware will cause platelets to stick to the wall). We used plastic tubes for centrifuge and storage of samples.
5. Plasma has very low levels of 5-HT, and platelets are difficult to isolate and work with. We have developed a platelet isolation method previously used in SERT activity measurement.Reference Horvath, Selby and Poskitt24
6. CSF 5HIAA level has a gradient, and collection of the sample has to be performed according to a strict protocol (i.e. collection of the first 0.5 ml of CSF after the first drop). We used the strict CSF sample collection protocol from the Hyland Lab.
Conclusions
We found no correlation between platelet serotonin and CSF 5HIAA levels or between salivary serotonin and CSF 5HIAA, suggesting that salivary serotonin should not be used as peripheral biomarker for central serotonin turnover in the setting of a Pediatric Neurology or Metabolic clinics for diagnosis or monitoring of treatment of patients with primary neurotransmitter deficiencies.
Acknowledgments
We are grateful to the patients and families for their participation in the study, our medical colleagues from the Divisions of Pediatric Neurology and Biochemical Genetics for their management and referral, our Research Assistant Ms. Delia Apatean for recruiting patients and collecting saliva samples and Ms. Grace Yang for performing the Enzyme Immunoassay experiments in the lab.
Funding
This work was supported by the Rare Diseases Foundation grant #F14-02307
Conflict of Interest
The authors have no conflicting interests to declare.
Statement of Authorship
GAH worked on conception and design, acquisition of data and drafted the manuscript; CE worked on conception and design, acquisition and analysis of the data and edited a significant portion of the manuscript; MD worked on acquisition and analysis of the data and edited a significant portion of the manuscript.
