The val66met polymorphism of brain-derived neurotrophic factor is associated with human esophageal hypersensitivity


  • D. H. Vasant,

    1. Gastrointestinal Centre, Institute of Inflammation and Repair, Faculty of Medical and Human Sciences, University of Manchester, Salford Royal NHS Foundation Trust, Salford, UK
    Search for more papers by this author
  • A. Payton,

    1. Centre for Integrated Genomic Medical Research, University of Manchester, Manchester, UK
    Search for more papers by this author
  • S. Mistry,

    1. Gastrointestinal Centre, Institute of Inflammation and Repair, Faculty of Medical and Human Sciences, University of Manchester, Salford Royal NHS Foundation Trust, Salford, UK
    Search for more papers by this author
  • D. G. Thompson,

    1. Gastrointestinal Centre, Institute of Inflammation and Repair, Faculty of Medical and Human Sciences, University of Manchester, Salford Royal NHS Foundation Trust, Salford, UK
    Search for more papers by this author
  • S. Hamdy

    1. Gastrointestinal Centre, Institute of Inflammation and Repair, Faculty of Medical and Human Sciences, University of Manchester, Salford Royal NHS Foundation Trust, Salford, UK
    Search for more papers by this author

Address for Correspondence
Professor Shaheen Hamdy, Gastrointestinal Centre, Institute of Inflammation and Repair, Faculty of Medical and Human Sciences, University of Manchester, Salford Royal Foundation Trust, Salford M6 8HD, UK.
Tel: +44 161 206 4363; fax: +44 161 206 4364;


Background  Recent evidence implicates brain-derived neurotrophic factor (BDNF) in visceral hypersensitivity and pain in functional gastrointestinal disorders. We hypothesized that presence of the val66met polymorphism in the BDNF gene would be linked to increased esophageal sensitivity to electrical stimulation.

Methods  A total of 39 healthy volunteers (20 males, mean age 30) compliant with inclusion criteria after screening procedures were genotyped for BDNF polymorphisms and completed an Hospital Anxiety and Depression Scale (HADS) questionnaire. Sensory (ST) and pain (PT) thresholds in the proximal (PE) and distal (DE) esophagus were determined using electrical stimuli to a swallowed intraluminal catheter with bipolar electrodes by an investigator blinded to the subjects’ genotype. For comparison, somatic ST and PT (hand and foot) were also tested. HADS scores together with esophageal and somatic thresholds were then correlated with BDNF polymorphism status.

Key Results  Eleven of 39 (28%) volunteers had at least one Met allele (Met carriers). When compared with Val/Val, Met carriers had lower esophageal PT (Median PT [mA]: Val/Val vs Met carriers, PE; 49.4 vs 44.3, = 0.033, DE: 63.8 vs 55.4, = 0.045) with higher proportion of Val/Val subjects in the upper quartile for PT in both PE (= 0.021) and DE (= 0.033), yet similar somatic PT (Median PT [mA] Hand; 33.6 vs 38.0, = 0.22, Foot; 44.7 vs 44.0, = 0.48). Sensitivity results were independent of anxiety (= 0.66) and depression (= 0.33) scores.

Conclusions & Inferences  val66met BDNF polymorphisms are associated with increased esophageal sensitivity to experimental electrical stimulation. Thus, BDNF genotype may be a useful biomarker for electrical sensitivity in the healthy human esophagus.


brain-derived neurotrophic factor


distal esophagus




functional chest pain


functional gastrointestinal disorders


gastro-esophageal reflux questionnaire


Hospital Anxiety And Depression Scale questionnaire


irritable bowel syndrome




McGill short-form pain questionnaire


proximal esophagus


pain threshold


sensory threshold


tropomyosin-related kinase B


upper esophageal sphincter




Functional chest pain (FCP) is a highly prevalent functional gastrointestinal disorder (FGID) previously estimated to occur in 13.6% of the general population.1 By definition these patients are distressed by chronic ‘non-burning’ midline chest discomfort, without evidence of gastro-esophageal reflux and exclusion of histopathologically confirmed esophageal disorders.2 Functional chest pain patients have consistently been shown to have increased sensitivity to experimentally induced esophageal pain, termed esophageal visceral hypersensitivity.3–5 Experimental evidence suggests that central sensitization is an important mechanism of esophageal visceral hypersensitivity.6 Despite these mechanistic advances, current treatment options for these debilitating visceral pain disorders remain limited7 and patient dissatisfaction is high.

Recent studies showing familial associations in FGIDs8,9 have triggered considerable speculation about environmental vs genetic factors. By formally identifying potential genetic factors, better approaches may be offered to personalize treatment strategies in the future for these patients moving away from the current symptom-based classification. However, candidate gene studies to date have proved inconclusive.10

Of interest, there is increasing evidence that brain-derived neurotrophic factor (BDNF), a neurotrophin involved in sensory nociceptive neurotransmission and a key player in central sensitization,11,12 has a role in visceral hypersensitivity. Several animal studies using experimental models of visceral pain13–16 and studies of both irritable bowel syndrome17 and chronic pancreatitis18 patients with visceral pain have shown associations between pain and increased uptake of BDNF. Moreover, animal studies have shown that BDNF deficiency and antagonism reduces visceral sensitivity.13–16 Indeed, a common single nucleotide polymorphism in the human BDNF gene located on the short arm of chromosome 11 that results in a valine to methionine (Met) substitution at codon66 (val66met) is known to affect intracellular packaging and regulated secretion of BDNF.19,20 Regulated secretion of BDNF in an activity-dependant fashion is fundamental to its role in controlling synaptic function and plasticity21 and hence may be an important driver in chronic pain syndromes. Moreover, the BDNF val66met polymorphism is highly prevalent, being present in approximately one third of the Caucasian populations.22,23

In view of the mounting evidence for a role of BDNF in visceral pain modulation coupled with the frequency and the functional significance of the polymorphism, we hypothesized that presence of the common val66met polymorphism in the BDNF gene (single nucleotide polymorphism database reference number: rs6265) would increase sensitivity to esophageal electrical stimulation in healthy subjects.

Materials and Methods


Forty-one previously ungenotyped healthy subjects were initially recruited. All subjects were in good health, gave written consent and complied with exclusion criteria (no history of cardiac, neurological, or gastrointestinal diseases and not taking any medication currently). To screen for clinically significant gastro-esophageal reflux, all participants completed the validated Mayo clinic gastro-esophageal reflux questionnaire (GERQ),24 where 2/41 subjects reported positive symptoms of gastro-esophageal reflux. Both had required acid suppressing medication within the previous year and were therefore excluded from the study. The 39 remaining subjects completed the experimental procedures which conformed to the World Medical Association declaration of Helsinki. Approval was granted by the Greater Manchester South Research Ethics Committee (LREC Reference: 10/1003/42). Studies were conducted in the clinical laboratories of the Gastrointestinal Sciences department of Salford Royal NHS Foundation Trust, part of the Manchester Academic Health Sciences Centre (MAHSC), United Kingdom.

Anxiety and depression scores

All subjects completed the validated Hospital Anxiety and Depression Scale (HADS) questionnaire on the day they were studied, prior to sensitivity testing and genotyping, to obtain current Anxiety (A) and Depression (D) symptom scores.25

Esophageal electromyography (EMG) and electrode placement

Subjects were intubated orally or nasally depending on their preference with a 3.2 mm diameter intraluminal catheter (Gaeltec Devices; Dunvegan, Isle of Skye, Scotland, UK) with two pairs of platinum bipolar ring electrodes (2 mm electrodes with interelectrode distance of 1 cm). The distal electrode pair was situated 5 cm from the tip of the catheter, whereas the proximal electrode pair was 20 cm from the catheter tip. Prior to intubation, a 20 cm distance above the proximal electrode pair was measured and marked on the catheter using tape. An earth lead was connected to a skin electrode sited over the upper portion of one of the sternocleidomastoid muscles of the neck. The catheter was then inserted beyond the proximal electrode pair (up to the 20 cm marker) before being connected via a preamplifier (Headstage, 1902 and 1401; Cambridge Electronic Design, Cambridge, UK) to a personal computer that allowed real-time visualization of EMG traces from both proximal and distal electrodes using the Signal Application Program (Cambridge Electronic Design). This had filters set at 200 Hz–2 kHz and allowed a sampling rate of 4–8 kHz. Response signals were processed through a 50/60 Hz noise eliminator (HumBug; Quest Scientific, North Vancouver, BC, Canada) to remove any unwanted electrical interference. While viewing the live EMG traces from both electrodes the catheter was slowly withdrawn 1 cm at a time, until the high tonic muscle activity zone of the upper esophageal sphincter (UES) was identified at the proximal electrode. The distance between the nares or the incisors to the 20 cm marker was then recorded and used to determine the distance to the UES. Once the UES position was determined for the proximal electrode, the catheter was inserted a further 5 cm and anchored. This ensured that the proximal electrode was positioned in the proximal esophagus (PE), 5 cm below UES, and the distal electrode was always positioned a further 15 cm below the proximal electrode into the distal esophagus (DE), 20 cm aboral to the UES.

Esophageal electrical stimulation

Esophageal electrical stimulation was selected as the experimental pain modality for this study as it has been shown to be a reliable and reproducible method of invoking experimental esophageal pain.26–28 Electrical stimulation of the esophagus is complementary to experimental balloon distension of the esophagus and both methods stimulate visceral afferent receptors, perhaps via different nociceptive pathways.

The intraluminal catheter, with one electrode stationed in the PE and one in the DE, was connected to a constant-current, high voltage stimulator (model DS7; Digitimer Ltd., Welwyn Garden City, UK) and externally triggered (model DSN; Digitimer Ltd.). The interelectrode impedance was monitored throughout to ensure mucosal contact. The stimuli were 200 μs square-wave pulses and were delivered at a frequency of 0.2 Hz. The maximum intensity of electrical stimulation used was 100 mA. Using an ascending method (2 mA increments) the intensity of stimuli was increased until the required threshold was achieved. To determine threshold values, the average of three recordings was calculated and used.

Measurement of esophageal sensory thresholds (STs)

Esophageal ST, defined as the stimulus intensity in mA when the subjects first become aware of a definite sensation within the esophagus, was determined using the electrical stimulation protocol in both PE and DE.

Measurement of esophageal pain thresholds (PTs)

A copy of the McGill short-form pain questionnaire (MPQ) was presented to all subjects.29 Pain thresholds (PTs) in the PE and DE were determined using the electrical stimulation protocol. Esophageal PT at each site was defined as the lowest intensity of electrical stimulation in mA required to first cause reports of ‘discomforting’ esophageal pain as per the present pain intensity index of the MPQ.

Measurement of somatic thresholds to electrical stimulation

Somatic sensitivity to electrical stimulation was tested using a pair of surface silver–silver gel electrodes (H69P; Tyco Healthcare, Gosport, UK) placed 1 cm apart over the first dorsal compartment of the right hand and on the dorsum of the right foot. The ST and PTs were then determined in the hand and foot using the same techniques and measurements as in the esophagus.

Subjective PT description

All subjects were asked to match their perception of the PT stimulus at each site from a list of descriptive terms on the MPQ.29

BDNF genotyping

Saliva specimens for genotyping were collected for DNA extraction using Oragene-250 self-contained DNA collection kits (DNA Genotek Inc, Ottawa, ON, Canada). All specimens were processed at the Centre for Integrated Genomic Medical Research (CIGMR), University of Manchester, using the same procedures described by Jayasekeran et al.30 Salivary DNA was extracted and purified using Applied Biosystems Assays-by-Demand kit (Applied Biosystems, Warrington, UK). A reaction mixture containing 2.5 μL of DNA Probe Master (Roche Diagnostics, West Sussex, UK), 20 ng genomic DNA in a 5 μL reaction volume, comprising 2 μL DNA, 0.125 μL Assay Mix (40X), and 0.375 μL nuclease-free water, was used for genotyping. The polymerase chain reaction (PCR) conditions were 95 °C for 10 min, then 50 cycles of 95 °C for 10 s and 60 °C for 30 s. The reaction was allowed to run and the PCR products were then electrophoresed. A Roche Lightcycler 480 genotyping real-time PCR platform (Roche Diagnostics) was then used to measure the sample reporter fluorescence. Researchers performing the sensory testing were blinded to genotyping data.

Data analyses

Mean PT and ST in mA for each subject at each of the four stimulation sites was calculated and HADS data were also inputted into the same spreadsheet. The subjects were then separated into two groups (Val/Val and Met carriers) based on their BDNF genotype. Met carriers had at least one Met allele (Fig. 1).

Figure 1.

 Summary of the experimental protocol.

Statistical analyses

Data were assessed for normality and were found to be non-normally distributed. Thereafter, group data from the two BDNF genotype cohorts were compared statistically using non-parametric tests using a standard software package (v2.7.8; StatsDirect Ltd., Cheshire, UK). All P values displayed are results from Mann–Whitney U-tests (unless stated otherwise). Data are displayed as median and inter quartile ranges.


Thirty-nine healthy volunteers (age range 19–57, mean age 30.2 years, 20 males) completed the study after screening.

BDNF polymorphism frequency

Brain-derived neurotrophic factor genotype frequency was 72% Val/Val and 28% Met carriers. Table 1 shows the genotype and gender distribution.

Table 1.   Gender and BDNF genotype distribution of healthy subjects (n = 39)
Number of subjects (n)Val/Val (28)Met carriers (11)
Val/Met (9)Met/Met (2)
  1. BDNF, brain-derived neurotrophic factor.

Male/Female (n)15/135/6

HADS questionnaire scores

Median anxiety (A) score (3) and depression (D) scores (1) were within the healthy normal range. Table 2 shows that median A and D scores between the two BDNF genotype groups were similar.

Table 2.   Median anxiety and depression scores
HADS questionnaireVal/ValMet carriersMann–Whitney
  1. HADS, Hospital Anxiety and Depression Scale.

Anxiety scores32 = 129.5, P = 0.66
Depression scores01 = 113, = 0.30

Esophageal electrode placement

In the subjects who selected oral catheter intubation (n = 22/39), the lower border of the UES was identified at a mean distance of 17.8 ± 0.6 cm from the incisors. The mean distance from the external nares to the lower border of the UES in subjects who passed the catheter nasally (n = 17/39) was 19.5 ± 0.5 cm.

Subjective PT symptoms

Overall, the word most commonly selected from MPQ to describe the PE PT was ‘throbbing’ (13/39), whereas DE PTs were commonly described as ‘burning’ (17/39). Hand PTs were most frequently perceived as ‘sharp’, ‘stabbing’ (both 13/39), whereas foot PTs were most frequently cited as ‘shooting’ (17/39). When divided into the two BDNF genotype groups the same descriptive terms were most commonly selected at each stimulation site.

Esophageal sensitivity to electrical stimulation

As reported in the literature, PE had lower thresholds to electrical stimulation than DE (Median STs; PE 15.3 mA, DE 23.6 mA; Mann–Whitney U = 326, n1 = n2 = 39, < 0.0001, Median PTs; PE 48.3 mA, DE 55.5 mA; = 525.5, n1 = n2 = 39, = 0.019).

Comparing esophageal sensitivity by BDNF genotype

Figure 2 shows that esophageal STs were similar between the two genotype groups. By contrast, Fig. 3 shows that Met carriers had lower PTs in both PE (= 0.033) and DE (= 0.045) compared with Val/Val subjects, with a higher proportion of Val/Val subjects being in the upper quartile for PT at both PE (= 0.021) and DE (= 0.033).

Figure 2.

 Comparison of esophageal sensory threshold (ST) data between brain-derived neurotrophic factor (BDNF) genotype groups. (A) Spread of esophageal STs. The distribution of STs is similar between the two groups at both esophageal levels. (B) Box and Whisker plots of median esophageal STs, showing inter quartile range. When compared between BDNF genotype groups, there was no significant difference in proximal esophagus ST (Medians [mA]; Met carriers 12.0 vs Val/Val 16.3, Mann–Whitney U = 179.5, n1 = 11, n2 = 28, = 0.22) and DE ST (Medians [mA] Met carriers 21.4 vs Val/Val 23.8, U = 181, n1 = 11, n2 = 28, = 0.20).

Figure 3.

 Comparison of esophageal pain threshold (PT) data between brain-derived neurotrophic factor (BDNF) genotype groups. (A) The spread plot demonstrates that at both esophageal sites Val/Val subjects had much higher esophageal PTs (>65 mA). Val/Val genotype frequency distribution in the upper quartile for both proximal esophagus (PE) PT (upper quartile value 56.4 mA, Fisher’s exact test = 0.021) and distal esophagus (DE) PT (upper quartile value 79.9 mA, Fisher’s exact test = 0.033) was significantly higher than in Met carriers. (B) A Box and Whisker plot showing median esophageal PTs and inter quartile range. When compared between BDNF genotype groups, Met carrier PTs were lower in both PE (Medians [mA]; Met carriers 44.3 vs Val/Val 49.4, Mann–Whitney U = 213, n1 = 11, n2 = 28, = 0.033) and in the DE (Medians [mA]; Met carriers 55.4 vs Val/Val 63.8, U = 213, n1 = 11, n2 = 28, = 0.045).

Comparing somatic sensitivity by BDNF genotype

In contrast to the esophageal data, there was no difference in hand (median STs [mA] Val/Val vs Met carriers: 2.5 vs 2.3, Mann–Whitney U = 120, n1 = 11, n2 = 28, = 0.15, median PTs [mA]; 33.6 vs 38.0, = 0.22) and foot (median STs [mA] Val/Val vs Met carriers: 4.3 vs 4.6, Mann–Whitney U = 158, n1 = 11 n2 = 28, = 0.45, median PTs [mA]; 44.7 vs 44.0, U = 152, n1 = 11, n2 = 28, = 0.48) sensitivity to electrical stimulation across the two BDNF genotype groups.


We studied a common polymorphism, BDNF val66met, given the plausible research evidence for a role in visceral pain processing, using biological end points (esophageal PTs to electrical stimulation). The results suggest that val66met polymorphisms of BDNF in healthy subjects may be associated with lower pain (but not sensory perception) thresholds to experimental electrical stimulation of the esophagus but not in the other somatic body regions tested. Our data also show that these findings were independent of the anxiety and depression scores measured using HADS questionnaire.

The precise reasons for a difference in esophageal sensitivity between the two genotype groups are unclear. However, one feasible explanation for these findings could be increases in activity-dependant secretion of BDNF in Met carriers. Coregulation between BDNF and glutamate is now thought to be vital in promoting neurogenesis and new synaptic connections (plasticity). There is evidence that glutamate (which is released by central terminals of primary nociceptor afferents in response to visceral pain) increases BDNF transcription and release. Brain-derived neurotrophic factor tropomyosin-related kinase B receptor activation via intracellular signaling pathways trigger glutamate release and potentiate NMDA receptor activation.11,31 Regulation of activity dependent BDNF secretion is known to be affected by the val66met BDNF polymorphism which may therefore disrupt the above mechanisms and promote synaptic plasticity and facilitate sensory neurotransmission in the brain and spinal cord dorsal horn neurons (central sensitization). Ultimately, this may lead to increased esophageal uptake of BDNF. A recent meta-analysis has concluded that there is no evidence for a correlation between serum BDNF levels and val66met polymorphisms in the BDNF gene,32 so simply testing ‘resting’ serum BDNF levels in FCP patients would not be a useful approach. Hence, to experimentally test our hypothesis would require the induction of an ‘activity dependant’ hyperalgesic state for BDNF quantification. This could be achieved using an esophageal model of central sensitization which has previously been described6 and would involve combining distal esophageal acid perfusion and painful esophageal stimuli (to confirm induction of visceral hyperalgesia) with genotyping for BDNF. A difference in serial serum BDNF levels after acid perfusion (activity-dependant levels) compared with baseline serum BDNF levels between the two genotype groups would be strong evidence to support our contention for a role of BDNF status being relevant to modulating esophageal sensation/hypersensitivity.

The lack of correlation with somatic PTs in our study is intriguing. Supportive evidence for this differential effect comes from a recent study where no association was found between BDNF val66met polymorphism prevalence in fibromyalgia patients when compared with healthy controls. Similar to our findings, the authors also demonstrated no difference in somatic mean PTs between 65 Val/Val and 28 Val/Met patients.33 We did not test other viscera in this study to corroborate our findings that this may be a specific non-somatic phenomenon. It would have been interesting to assess other gastrointestinal regions such as the rectum or colon to see if a similar relationship to BDNF polymorphisms also exist. Such a study would be an important component of any future studies looking at pan-gut sensitivity and genetic associations.

With regard to any putative psychogenetic interactions, we found that our sensitivity findings were independent of anxiety and depression; most likely a consequence of a psychologically healthy phenotype confirmed by HADS, which was similar in the two genotype groups. There is, however, well-documented evidence for an association between psychiatric symptoms and esophageal hypersensitivity.34 Furthermore, the incidence of anxiety and depression is reported to be high in patients with functional upper gastrointestinal symptoms,35,36 with higher symptom frequency in depressed patients.37 Several studies have reported associations between BDNF38–40 and the val66 polymorphism in anxiety41 and depression.42,43 This leads one to speculate whether presence of common genetic polymorphisms including BDNF val66met could partially explain susceptibility to coexisting visceral pain hypersensitivity anxiety and depression.

One of the limitations of our study was the relatively small sample size and the unequal distribution of subjects in the two groups, which reflects the frequency of the polymorphism in a previously ungenotyped population. Another limitation was that our subjects did not have formal esophageal manometry or endoscopy as part of the study prior to esophageal sensitivity testing. Without these additional procedures the presence of hiatus hernia in our healthy subjects cannot be excluded; however, the use of the GERQ to exclude subjects with clinically significant reflux symptoms at the screening stage makes large (asymptomic) herniae unlikely. Our esophageal EMG guided measurements of the distance to the lower border of the UES in both oral and nasal intubation were consistent with anatomical data from previous studies, confirming accuracy of our technique in esophageal electrode placement. In addition, in terms of somatic sensitivity testing, the use of electrical stimulation alone, while providing us with reliable and reproducible results to measure PTs, did not allow us to discriminate between modalities such as temperature, pin prick, and pressure; ideal testing would likely require multimodal approaches using several additional techniques including the cold pressor test. Indeed, there is now evidence that each modality represents a different specific dimension.44 Finally, we applied a single gene analysis which is subject to methodologic criticisms, given that there are likely to be many other genetic polymorphisms relevant to visceral hypersensitivity and FCP. Several other genetic polymorphisms, including Serotonin Transporter (SERT-SL),45 G protein β3 subunit (GNB3 825),46–50 Transient Receptor Potential Vanilloid 1 (TRPV1 G315C),51 have all recently been linked with upper FGIDs and may also be of interest in terms of esophageal sensitivity. Nonetheless, having an a priori hypothesis does allow single gene study methodology to be applied in a controlled setting.

In conclusion, our current study suggests that the presence of the val66met BDNF polymorphism is associated with visceral sensitivity of the esophagus (not somatic sensitivity) in response to experimental electrical stimulation in healthy individuals. Further studies are needed to validate the presence of the val66met BDNF polymorphism in patients with esophageal disorders such as functional esophageal pain. If this polymorphism is shown to be present in FCP, then, BDNF receptors may have therapeutic potential in select patients with esophageal pain.


The authors thank Drs. Tarig Algladi, Emilia Michou, and Vanoo Jayasekeran (University of Manchester) who helped with technical aspects of study design; staff at the Centre for Integrated Genomic Medical Research; and Ms Annie Herbert (Medical statistician). The GERQ was used by permission of Mayo Foundation for Medical Education and Research. All rights reserved. The HADS questionnaire was used with permission and a license agreement with GL assessment.


Dr. Dipesh Vasant was an NIHR Academic Clinical Fellow in Gastroenterology. The study was sponsored by the University of Manchester, UK, which did not have a role in the study design, in the collection, analysis or interpretation of data.


There were no financial, professional, or personal disclosures from any of the authors.

Author Contribution

DHV designed and performed the studies, analyzed the data, and wrote the manuscript; AP provided genetics advice and supervised the analysis of genetic specimens; SM helped with study design, provided technical assistance, and helped write the manuscript; DGT helped write the manuscript; SH conceptualized the study, helped with data interpretation and writing the manuscript.