If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Denver Health Medical Center, 777 Bannock Street, Denver, CO 80204, United StatesDepartment of Neurology, University of Colorado, 1635 Aurora Court, Aurora, CO 80045, United States
Menthone has been identified as the dominant constituent of seizure-scented sweat.
•
Canines were unable to distinguish between fear-scented and seizure-scented sweat.
•
Menthone may be an important pre-ictal biomarker of pending seizure.
•
We believe we are the first to identify menthone as a human alarm pheromone.
Abstract
Objective
In our canine scent detection research involving a specific volatile organic compound (VOC) associated with human epileptic seizure, we began to suspect involvement of the primitive neural networks associated with production of a previously undescribed human alarm pheromone as the origin of our seizure scent. We hypothesized that if we presented fear-scented sweat to our canine seizure scent detection team, and they identified the fear scent as their seizure scent, then that would suggest that they are identical compounds.
Methods
Following consent and approval, sweat samples taken from volunteers associated with the Brooke Gordon Comprehensive Epilepsy Center at Denver Health were processed by the Canine Assistants (CA) service dog team that had been imprinted to recognize the unique seizure scent from our previous study. In part one, sweat samples were collected from subjects, who had no prior history of epilepsy or seizures, under two different testing environments: watching a scary movie (It) and a neutral/comedy movie (Airplane!). In part two, a larger follow-up study utilizing fear sweat, exercise sweat, epilepsy sweat, and other distractor scents were provided in a multiple choice paradigm to better understand the inter-rater reliability of the canine responses.
Results
In part one, our canine seizure scent detection team identified fear-scented sweat samples as their seizure scent in 4 of 5 study participants. There was almost perfect agreement of seizure scent detection during fear scent trials between the canine seizure scent detectors with a kappa value of 0.814 (95% CI: 0.668–0.960). In part two, (utilizing eleven different subjects) our canine scent detection team identified samples of either fear or seizure sweat with a sensitivity of 82% and a specificity of 100% (no false positives) from among the multiple choices offered. Additionally, there was 92% agreement between the members of the canine scent detection team.
Significance
While this hypothesis testing study is small and deserves replication, it confirms that the Canine Assistants seizure scent detection team consistently and accurately identified fear-scented sweat as their seizure scent, implying that the VOC, menthone, is common to both conditions. This further implies that human seizure propagation and fear network circuitry may share a common anatomy, and that menthone may not only be an early seizure biomarker, but a newly described human alarm pheromone.
“There are no neural circuits unique to seizures, therefore propagation of seizure activity must be conducted along existing circuits of normal physiologic processes” [
In 2016 a collaboration between Canine Assistants (CA) and Florida International University (FIU) set out to identify possible scent constituents that seizure alert dogs might be recognizing [
Davis P. The investigation of human scent from epileptic patients for the identification of a biomarker for epileptic seizures [PhD dissertation on the internet]. Florida International University; 2017. Available from: https://search.proquest.com/docview/2103114969.
In part one of the CA/FIU study, investigators first examined the sweat of a group of persons with epilepsy (PWE) compared to a non-PWE CONTROL cohort using head space gas-chromatography to evaluate the volatile organic compounds (VOCs) present [
Davis P. The investigation of human scent from epileptic patients for the identification of a biomarker for epileptic seizures [PhD dissertation on the internet]. Florida International University; 2017. Available from: https://search.proquest.com/docview/2103114969.
]. Subtracting out the baseline VOCs present in the CONTROL group compared with the PWE group, the PWE cohort then underwent a secondary analysis in which seizure (postictal) sweat was compared to their baseline inter-ictal sweat profiles. Subtracting out these inter-ictal baseline compounds yielded nine unique VOCs that were present only in patients with epilepsy, only following a seizure, and not at other times.
In part two of the CA/FIU study, pure laboratory standards of the three most prominent VOCs identified in part one were combined to form a putative standardized “seizure scent” and mailed to Canine Assistants for canine imprinting and identity confirmation. First, the standard “seizure scent” was imprinted in four CA dogs. Second, these four dogs were asked to identify the standard scent from five multiple choice options containing the target scent as well as four other distractors including the interictal sweat sample from a person with epilepsy. Third, once the dogs successfully identified the standardized seizure scent from multiple choices, they were presented a seizure (postictal) sweat sample from the same patient with epilepsy in the second phase of testing. The dogs confirmed this postictal scent as identical to the laboratory standard scent. Fourth, three new dogs were imprinted on the postictal scent from the same patient with epilepsy in tests two and three. Finally, the three new dogs were presented the original laboratory standardized “seizure scent”, and all three confirmed it was identical to the human postictal scent that they had just learned [
Davis P. The investigation of human scent from epileptic patients for the identification of a biomarker for epileptic seizures [PhD dissertation on the internet]. Florida International University; 2017. Available from: https://search.proquest.com/docview/2103114969.
In “Canine Detection of Volatile Organic Compounds Unique to Human Seizure” Maa et al. confirmed the presence of the putative unique “seizure scent” in a prospective manner.
Investigators collected sweat samples from 60 patients undergoing phase I or phase II epilepsy monitoring and demonstrated that the Canine Assistants service dog team (previously imprinted with the laboratory “seizure scent”) was able to positively identify their seizure scent an average of 68 min prior to clinical onset of seizure and for an average of 81 min after seizure offset. Therefore the postictal seizure scent identified by the CA/FIU collaboration was confirmed to emerge in the pre-ictal time period as well. Additionally, this scent was not associated with sweat samples collected during nonepileptic seizures (NES), the scent was present in both focal and genetic generalized convulsive seizures, and finally the presence of a positive scent-detection carried an 82.2% probability of a clinical seizure to follow [
Maa EH, Arnold J, Ninedorf K, Olsen HA. Canine detection of volatile organic compounds unique to human epileptic seizure. Epilepsy Behav 2021;115:107690, https://doi.org/10.1016/j.yebeh.2020.107690.
While the implications of an early-warning biomarker for the treatment and prevention of an epileptic seizure are staggering, the results also offer an exciting opportunity to opine on the origin of the seizure scent and how it informs our understanding of the earliest stages of, and networks involved with, seizure evolution.
From the initial cohort of patients with epilepsy in the CA/FIU collaboration, the nine unique VOCs were: menthone, menthyl acetate, 3-ethoxy-3,7-dimethyl-1,6-octadiene, camphor, pentadecanal, valencene, (−)-β-bourbene, β-cubene, and 4-tert-butylcyclohexyl acetate [
Davis P. The investigation of human scent from epileptic patients for the identification of a biomarker for epileptic seizures [PhD dissertation on the internet]. Florida International University; 2017. Available from: https://search.proquest.com/docview/2103114969.
]. Of these compounds, menthone, menthyl acetate, and 3-ethoxy-3,7-dimethyl-1,6-octadiene were of highest concentration and therefore were chosen for the original CA/FIU laboratory standard. During mixing studies, using laboratory derived standards, it become evident that menthone specifically appeared to be critical for consistent canine scent detection of seizure samples [J. Arnold, personal communication, American Epilepsy Society Meeting 2017].
A cursory review of menthone in the literature, however, suggests it is primarily a component of the essential oil of peppermint. The primary pathway for monoterpene biosynthesis in peppermint (adapted in Fig. 1) shows the progression from the monoterpene precursor, geranyl diphosphate, with intermediate steps through menthone and finally to menthol [
But there is no known human source of menthone biosynthesis. A plant source of menthone, to later be released on-demand during human epileptic seizure, could be rationalized by exogenous, dietary intake. However, between the eleven subjects enrolled in Florida from the original trial [
Davis P. The investigation of human scent from epileptic patients for the identification of a biomarker for epileptic seizures [PhD dissertation on the internet]. Florida International University; 2017. Available from: https://search.proquest.com/docview/2103114969.
Maa EH, Arnold J, Ninedorf K, Olsen HA. Canine detection of volatile organic compounds unique to human epileptic seizure. Epilepsy Behav 2021;115:107690, https://doi.org/10.1016/j.yebeh.2020.107690.
], it is highly improbable that all subjects would have ingested sources of peppermint in temporal proximity to an epileptic seizure.
In attempting to resolve this conflict a compelling link arose, between the plant and animal origins of this compound, from a completely unlikely literature [
]. When crushed, the acarid mite, Tyrophagus similus, emits a characteristic smell that triggers escape behavior in neighboring conspecific mites. By placing crushed mites in hexane, researchers extracted an alarm pheromone that was able to reproduce the acarid mite escape behavior in concentrations as low as 100 ppm. Using mass spectrometric analysis they were able to first identify, then chemically synthesize, and finally demonstrate escape behavior in healthy acarid mites with their synthetic pheromone. The pheromone was characterized as isopiperitenone, and one of its synthetic intermediates was menthone (seen in the plant monoterpene biosynthetic pathway, Fig. 1).
Another economically important insect pest, the aphid, produces an alarm pheromone using the sesquiterpene pathway, specifically a compound called (E)-β-farnesene [
Verheggen FJ, Haubruge E, Mescher MC. Alarm pheromones-chemical signaling in response to danger. In Gerald Litwack, ed: Vitamins and hormones, Vol. 83, Burlington: Academic Press, 2010, pp. 215-240.
]. This warning of impending danger signals conspecific aphids to stop feeding and move away from the signaler or to drop off the host plant. Interestingly, Eβf is also a common component of plant volatile emissions, induced by herbivore feeding or mechanical damage, and is a constituent of essential oils found in several plant families including Asteraceae, or the daisy family [
Verheggen FJ, Haubruge E, Mescher MC. Alarm pheromones-chemical signaling in response to danger. In Gerald Litwack, ed: Vitamins and hormones, Vol. 83, Burlington: Academic Press, 2010, pp. 215-240.
]. In a compelling example, insect infestation by the aphid causes mechanical damage to the wild potato plant that in turn uses a plant-derived Eβf to trick the aphids into halting their attack [
Verheggen FJ, Haubruge E, Mescher MC. Alarm pheromones-chemical signaling in response to danger. In Gerald Litwack, ed: Vitamins and hormones, Vol. 83, Burlington: Academic Press, 2010, pp. 215-240.
The complex relationship between plant and insect using the same chemical signal suggests these primitive biosynthetic pathways may be evolutionarily conserved. So to answer the original question “is the menthone that Canine Assistants identified 68 min prior to seizure onset endogenously or exogenously produced?” a series of questions follow: Is the menthone produced by the peppermint plant only produced by plants? or like Eβf, does peppermint produce menthone to counter a specific insect infestation as an alarm pheromone? If so, does this mean that an insect actually produces menthone, and if an insect can produce menthone, can other phyla within kingdom Animalia also produce it (See Fig. 2)? If this too is the case, and more specific to our first canine study, is it possible that the menthone in seizure sweat is endogenously produced by humans? And if humans can produce menthone endogenously must it be a product of an existing circuit of a “normal physiologic process”? Finally, if an existing circuit of a normal physiologic process is responsible for endogenous production of menthone, is this circuit related to human alarm pheromone production?
Fig. 2Pathway for human cholesterol biosynthesis; with overlay of plant monoterpene pathway identifying possible location of missing cyclization enzyme.
To quickly test this idea, we hypothesized that the Canine Assistants dog team, already scent-imprinted to identify seizure sweat samples (menthone ≫>≫ menthyl acetate, and 3-ethoxy-3,7-dimethyl-1,6-octadiene) would be unable to distinguish between the seizure-scent and sweat samples collected from non-PWE subjects provoked into fear.
2. Methods
2.1 Study population
The study population was derived from volunteers affiliated with the Denver Health Comprehensive Epilepsy Center, but without a diagnosis or suspicion of epilepsy. Subjects aged between 18 and 89 years old were included. Study consent was obtained under the protocol approved by our institutional review board (COMIRB #18-0399).
2.2 Canine teaching and sample testing method
The canine teaching and testing method were previously described [
Maa EH, Arnold J, Ninedorf K, Olsen HA. Canine detection of volatile organic compounds unique to human epileptic seizure. Epilepsy Behav 2021;115:107690, https://doi.org/10.1016/j.yebeh.2020.107690.
A single trainer (JA) was responsible for teaching a team of 13 dogs at the Canine Assistants headquarters in Alpharetta, Georgia. The FIU laboratory produced the 3 VOC mixture (ratio unknown) captured on a sterile cotton pad (100% cotton, 2 × 2, 8 ply, gauze pads; DUKAL Corporation, Syosset, NY, USA), placed in airtight, ten-ml glass, clear, screw top headspace vials with PTFE/Silicone septa (SUPELCO, Bellefonte, PA, USA) and mailed to Canine Assistants.
At Canine Assistants, from their earliest rearing, all of their service dogs are taught to communicate with their human handlers by responding to yes/no questions. Initially a trainer extends her LEFT hand that has been rubbed with a treat and repeats the word “yes”; followed by extending the unscented RIGHT hand and repeating the word “no”. When the dog goes to smell the LEFT hand or touches its nose to the LEFT hand, the dog is given the treat. Within a few repetitions the dog is able to associate “yes” with touching the LEFT hand, and “no” with touching the RIGHT hand, and over time this becomes generalizable by the dog to a variety of questions presented in this manner [
For the original epilepsy monitoring unit study, the standard seizure scent was imprinted by the following protocol: after opening the FIU vial, each dog was sequentially presented with the laboratory standard, and reinforced with the phrase “this is your seizure” and “this is your smell”. Testing began several minutes later by asking each dog “is this your seizure?” and “is this your smell?” while presenting the seizure scent or a volunteer’s non-seizure sweat sample. After one or two trials, each dog correctly signaled “yes” (touching of the trainer’s LEFT hand) when the standard seizure scent was presented or “no” (touching of the trainer’s RIGHT hand) when a volunteer’s sweat sample was presented [
Davis P. The investigation of human scent from epileptic patients for the identification of a biomarker for epileptic seizures [PhD dissertation on the internet]. Florida International University; 2017. Available from: https://search.proquest.com/docview/2103114969.
Maa EH, Arnold J, Ninedorf K, Olsen HA. Canine detection of volatile organic compounds unique to human epileptic seizure. Epilepsy Behav 2021;115:107690, https://doi.org/10.1016/j.yebeh.2020.107690.
As a proof of concept and in order to provoke fear sweat, two testing conditions were explored in the same small cohort of initial subjects: (1) the movie IT was viewed as the fear-provoking stimulus [
Subjects had sweat samples collected identically to the protocol previously described in Maa et al., but timing of collection was modified due to the novel stimuli. Subjects had sweat samples taken 15 min prior to movie onset, between 30 and 50 min after movie onset, and 15 min after movie offset. Movies were watched together as a cohort, in the same room, at the same ambient temperature, separated by one day.
Sweat sample collection protocol:
A Puritan Sterile Foam Tipped Applicator (6″ plastic handle, blue cap, 1 swab per tube, Guilford, ME, USA) was suspended by study personnel so that subjects could rub the foam tip between the palms for 10 s. The entire foam tip and applicator were then carefully placed back into the transport tube and labeled. After entire sweat collection was completed, each individual swab was removed from the transport tube, the foam tip was separated from the applicator, and touchlessly dropped into a Mylar Smell Proof Barrier Bag (Premium Vials; 1 oz; 3 1/8″ × 5 1/8″ airtight; zipper stand up pouch; heat seal-able Zip Lok; matte black), and then heat-sealed. For each subject an Excel database table with all samples labeled by stimulus and date/time was created for processing by the above canine protocol.
Part 2
To confirm the CA canine seizure scent team results from part one, a multiple-choice paradigm was created for a larger cohort of subjects. One fear-provoked sweat sample was collected 30–50 min after movie onset, in a similar manner as Part 1, for the second cohort of subjects. A neutral stimulus was not collected in the second protocol, instead prior to movie start, all subjects jogged in place for 5 min to collect an exercise sweat sample. A baseline sample, pre-exercise, was also collected from each subject. Ictal and interictal sweat samples from a single subject undergoing intracranial EEG monitoring, similar to our prior study, were also collected. Finally, a blank foam-tipped applicator was also included to make six possible choices per subject.
2.4 Sweat sample testing protocols
Part 1
Blinding of the trainer was achieved by separating the data entry/database handling from the canine response phase. A research assistant was responsible for handing the individual unlabeled pouch to the trainer (JA), awaiting the canine response from the trainer, and then updating the result database. The trainer received each pouch, opened the heat seal and zip loc, allowed the dog to briefly smell the pouch contents without touching the pouch, quickly resealed the zip loc pouch, then asked the dog “is this your seizure?… is this your smell?” and then reported the dog response “yes” with a touch of the trainer’s left hand or “no” with a touch of the trainer’s right hand. Each dog was tested for all samples in isolation and out of view of the other dogs, to avoid inter-rater influence. Similar to the original study, only four dogs were used at one time out of concern that the VOC dissipation would affect the ability of the latter dogs to detect the seizure/fear scent.
Part 2
After test samples were collected by the above protocols, each subject was associated with the following labeled test samples: Fear, Baseline, Exercise, Ictal, Interictal, and Blank with initials and age. Next a key with consecutive numbers was generated next to all samples (N = 1–90). Then all samples were individually transferred to small glass jars, sealed, key number labeled, placed into a duffle bag, and shaken by study personnel. Dogs were individually presented all 90 jars in random order, as handed to trainer (JA), before the next dog was tested. Study personnel recorded results Y/N for each key labeled response.
2.5 Statistical method
In the first and second analyses, inter-rater agreement between dogs for detection of fear scent or detection of both fear and seizure scent, respectively, was assessed by calculating Fleiss’s kappa statistic for multiple raters using the DescTools library, freely available through R v. 4.0.4 open source statistical software package. Kappa statistics estimate the strength of agreement between raters of events above that which would be expected due to chance alone. The strength of agreement beyond that which is due to chance alone was labeled as follows: kappa value <0.4, poor to fair agreement; 0.41–0.60, moderate agreement; 0.61–0.80, substantial agreement; and 0.81–1.00, almost perfect agreement. 95% confidence intervals and p-values were calculated using normal approximations as per Fleiss’s statistical method for calculating kappa with multiple raters. In the second analysis, sensitivity and specificity were also calculated to determine the accuracy of the dogs in detecting the combined outcome of both fear and seizure scents.
3. Results
3.1 Subject characteristics
Subjects provided consent according to our local review board protocol (COMIRB #18-0399), see Table 2 for part one and part two subjects.
In the first analysis, the canine seizure scent detection team identified samples of fear scent as their seizure scent in 4 of 5 study participants, with no identifications of seizure scent during non-fear scent exposure trials (see Table 3). There was almost perfect agreement of seizure scent detection during fear scent trials between the canine seizure scent detectors with a kappa value of 0.814 (95% CI: 0.668–0.960, p < 0.001), representing 81% agreement above that expected due to chance alone.
Table 3PART 1 Fear sweat sample responses from canine seizure scent detection team.
To further clarify whether the canine team was just identifying fear sweat as different from non-fear sweat or actually believed the fear sweat scent was the seizure scent they were trained to identify, the larger follow-up study was conducted. From the multiple choice paradigm, the canine seizure scent detection team identified samples of either fear or seizure scent as their seizure scent with a sensitivity of 0.82 and a specificity of 1.0, with no false-positive detections of seizure or fear scent when the sample was neither seizure nor collected during the fear provocation. Additionally, there was almost perfect agreement of seizure scent detection during fear or seizure scent trials between the canine seizure scent detector team members with a kappa value of 0.916 (95% CI: 0.851–0.981, p < 0.001), representing 92% agreement above that expected due to chance alone.
]. The characteristic feature of pheromones is that they exert their clinical effects while excluding conscious perception and are emitted at super low concentrations in ppb (10−9) and much lower [
]. In addition to arthropods, invertebrates (mollusks, flatworms, and annelids) release signals when injured that provoke anti-predator behavior in local conspecifics [
Verheggen FJ, Haubruge E, Mescher MC. Alarm pheromones-chemical signaling in response to danger. In Gerald Litwack, ed: Vitamins and hormones, Vol. 83, Burlington: Academic Press, 2010, pp. 215-240.
]. In vertebrates, fish contain specialized epidermal cells that release alarm pheromones when traumatized that signal conspecifics to exhibit increased vigilance and escape behaviors [
Verheggen FJ, Haubruge E, Mescher MC. Alarm pheromones-chemical signaling in response to danger. In Gerald Litwack, ed: Vitamins and hormones, Vol. 83, Burlington: Academic Press, 2010, pp. 215-240.
]. The primary substance responsible for fish fright behavior is hypoxanthine-3-N-oxide which produces a reaction that can be visually detected as a sudden reduction in movement that rapidly propagates to nearby individuals throughout the entire group [
Verheggen FJ, Haubruge E, Mescher MC. Alarm pheromones-chemical signaling in response to danger. In Gerald Litwack, ed: Vitamins and hormones, Vol. 83, Burlington: Academic Press, 2010, pp. 215-240.
While mammals are widely studied in their use of pheromones to attract mates, mark territories, and coordinate group behavior (synchronizing estrus), relatively little is known about mammalian alarm pheromones [
Verheggen FJ, Haubruge E, Mescher MC. Alarm pheromones-chemical signaling in response to danger. In Gerald Litwack, ed: Vitamins and hormones, Vol. 83, Burlington: Academic Press, 2010, pp. 215-240.
]. It is believed that mammalian alarm pheromones must be highly volatile, low molecular weight compounds such as alcohols, aldehydes, ketones, fatty acids, steroids, or esters with a propensity to spread and penetrate and have a short duration of action [
Verheggen FJ, Haubruge E, Mescher MC. Alarm pheromones-chemical signaling in response to danger. In Gerald Litwack, ed: Vitamins and hormones, Vol. 83, Burlington: Academic Press, 2010, pp. 215-240.
], at least in rodents. Human evidence of VNO has historically been controversial, but a recent review reaffirms its existence particularly in infants and children, but whether or not it functions still remains controversial [
In humans, the viscero-sensory descriptions associated with fear have for so long been suspected that its use has become axiomatic in our literature, history, and art.“and shall make him of quick understanding in the fear of the Lord: and he shall not judge after the sight of his eyes, neither reprove after the hearing of his ears:” – King James Version (KJV) Isaiah 11:3“I have almost forgotten the taste of fears: The time has been, my senses would have cool’d to hear a night-shriek; and my fell of hair would at a dismal treatise rouse and stir as life were in’t: I have supt full with horrors; Direness, familiar to my slaughterous thoughts, cannot once start me. – Shakespeare’s 'Macbeth' (1606) act 5, sc. 5, l. 9“Be scared. You can't help that. But don't be afraid. Ain't nothing in the woods going to hurt you if you don't corner it or it don't smell that you are afraid. A bear or a deer has got to be scared of a coward the same as a brave man has got to be.” – William Faulkner“In my memory, Stamps [Arkansas] is a place of light, shadow, sounds and enchanting odors. The earth smell was pungent, spiced with the odor of cattle manure, the yellowish acid of the ponds and rivers, the deep pots of greens and beans cooking for hours with smoked or cured pork. Flowers added their heavy aroma and, above all, the atmosphere was pressed down with the smell of old fears and hates and guilt.” – Maya Angelou
Recently, investigators have begun to examine human alarm pheromone effects and have shown a surprising array of clinical and radiographic changes. In a series of studies using scary movies to provoke fear-scented sweat, investigators have demonstrated the human ability to differentiate a presumed alarm pheromone from a neutral condition odor, elucidated several behavioral impacts induced by a presumed alarm pheromone, and also showed real-time radiographic impacts by fMRI.
Two studies demonstrated the ability to differentiate between fear-evoked versus a neutral sweat odor. The first group demonstrated the ability of female subjects to distinguish between the odor of sweat collected from other female subjects while watching a scary film from that collected from a neutral film [
]. The second group used a mixed-sex cohort to collect axillary odors while watching a scary film or a happy film and exposed the odors to a separate mixed-sex cohort one week later. Men correctly chose the “happy” scent of women (but not men) who viewed the comedy film. Both men and women correctly chose the “fearful” odor of men (but not of women) who viewed the scary film [
Separate psychology departments have examined the chemosensory signaling effects of fear (Rice University) and anxiety (Christian-Albrechts-University) on visual processing and performance. In the first study, axillary sweat was taken from 12 males while waiting for an academic examination (anxiety) and later while exercising. Sixteen females were presented faces with different expressions, and when tested in the presence of the previously collected exercise scent, tended to rate faces with neutral expression more favorably, and when presented stimuli in the presence of the anxiety scent, tended to rate neutral faces more negatively [
]. In the second study, testing in the presence of fearful sweat tended to bias women toward interpreting ambiguous facial expressions as more fearful. This effect was lost when the test expression was more discernable [
]. Using a similar anxiety scent collection paradigm, the Albrechts laboratory tested the startle reflex of a mixed-sex group of subjects and demonstrated increased amplitudes of electromyographic eye-blink muscle response in the presence of the anxiety scent as compared with the exercise scent [
]. The Rice laboratory then examined fear chemosignaling’s effect on cognitive performance. In a double-blind trial, female participants performed a word association task under three separate conditions: in the presence of fear sweat, neutral sweat, and a control odor. Participants performed more accurately and as quickly in the fear condition as compared with the other testing environments [
The conclusion of these studies suggests that a human alarm pheromone exists, and its function promotes increased vigilance and threat appraisal in the setting of uncertain stimuli. These physiological and neurocognitive changes would naturally offer a survival advantage, but the fact that subjects were unable to discern objective differences in the odors of fear/anxiety sweat versus exercise sweat raises interesting questions about the anatomical networks involved in the subconscious processing of the fear.
Perhaps not surprisingly then, radiographic evidence of these effects has also been demonstrated. Using sweat samples collected from subjects undergoing an emotional stressor (first-time tandem skydive), investigators conducted fMRI studies exposing these sweat samples to an unrelated cohort. Functional MRI revealed significant activation of the LEFT amygdala when subjects were exposed to the odor of stress sweat versus exercise sweat, and this finding was independent of same-sex or opposite-sex donor-detector pairs [
]. In fact, a “major fear circuit” in the brain is described within the lateral and central parts of the amygdala in the lobus temporalis, with connections to the periaquaeductal grey (PAG) of the diencephalon and mesencephalon, and then “output generating parts of the brainstem and medulla” [
]. Behavioral reactions include either a “freeze”, “flight”, or “fight” response. Physiologic reactions include increase in heart rate, muscular tension, sweating, and the adrenal mediated release of cortisol. [
]. This “fear-network” circuit involving the dorsal anterior cingulate cortex (dACC), bilateral anterior insular cortex (AIC), amygdala, orbitofrontal cortex (OFC), anterior thalamus, ventral putamen and pallidum, and midbrain substantia nigra/ventral tegmentum, was confirmed by the largest fMRI meta-analysis to date [
]. The “central-autonomic-interoceptive network” primary input is through the anterior insula, which is responsible for generating an integrated awareness of one’s cognitive affective and physical state. The network’s primary output is the dorsal anterior cingulate which facilitates homeostatic autonomic and behavioral responses through secondary involvement of viscerosensory and visceromotor sites including the dorsal midbrain (periaqueductal gray), ventromedial thalamus, hypothalamus, and the nucleus of the solitary tract within the pontomedulary junction [
Maa EH, Arnold J, Ninedorf K, Olsen HA. Canine detection of volatile organic compounds unique to human epileptic seizure. Epilepsy Behav 2021;115:107690, https://doi.org/10.1016/j.yebeh.2020.107690.
], this scent is identifiable an average of 68 min before the onset of a clinical-electrographic seizure as confirmed by intracranial EEG. Given our understanding of the anatomy of seizure propagation along limbic circuits and other eloquent and non-eloquent cortical and subcortical networks, does the generation of a potential alarm pheromone point to nodes of the fear circuit lying an hour upstream of our seizure networks? Or are they the same? And why is it common to both focal and generalized seizures?
The genesis of our canine seizure scent work comes from a long history of interest in seizure detection. A detection system must be able to first identify a quantitative value (a seizure biomarker) and secondly classify a threshold quality of the biomarker to maximize sensitivity and specificity [
]. Many strategies have been employed in the search for a clinically useful and economically viable seizure detection method utilizing biomarkers including scalp EEG, intracranial EEG, EMG (electromyography), EKG (electrocardiography), accelerometry, video/audio, and canine olfaction [
]. The time window range of seizure detection varies by biomarker and technology. In a series of patients with medication-resistant temporal lobe epilepsy, heart rate acceleration was noted to precede clinical seizure in 92% of cases and up to five seconds before seizure onset [
]. Quantitative analysis and individualized machine learning using EEG to generate a ten-minute prediction horizon was demonstrated to work in one subject and not in others in a small cohort [
Looking at a different EEG biomarker, Litt et al. measured the energy of the EEG signal (defined as the summation of the square of the voltage at each point over a designated time period) and showed that it begins to rise seven hours prior to seizure onset. However because the long-term cumulative energy was elevated for four to six hours after a seizure, EEG samples used in their analysis only qualified for inclusion if separated by more than six hours. This led to their hypothesis that the mechanism giving rise to seizure clusters may be different than those leading to isolated seizures, and that seizure generation likely begins well before an hour prior to clinical seizure [
]. Further, they noted that visual inspection of the pre-seizure EEG recordings did not demonstrate obvious findings that corresponded to changes in accumulated energy. The actual number of epileptiform discharges did not differ statistically from baseline periods [
Maa EH, Arnold J, Ninedorf K, Olsen HA. Canine detection of volatile organic compounds unique to human epileptic seizure. Epilepsy Behav 2021;115:107690, https://doi.org/10.1016/j.yebeh.2020.107690.
], during which canine scent detection transitioned from clearly negative to equivocal three hours before a series of strongly positive scent detections occurred (see Table 4). What made this subject unusual was that despite a typical scent profile for a true-positive seizure identification, no intracranial electrographic or clinical seizure was ever identified during this window of time [
Maa EH, Arnold J, Ninedorf K, Olsen HA. Canine detection of volatile organic compounds unique to human epileptic seizure. Epilepsy Behav 2021;115:107690, https://doi.org/10.1016/j.yebeh.2020.107690.
]. The changes that generated the menthone and other volatile organics unique to seizure, appeared hours before a suspected seizure, but in this case, did not actually require a seizure be manifest. Our VOC findings appear to best mirror the Litt et al.’s prediction window and supports that whatever drives the increased energetics of a patient’s EEG, may be related to both phenomena. Given the time course of onset of increased EEG cumulative energy (and menthone detection), numerous mechanisms were hypothesized including intrinsic voltage-dependent currents with longer time constraints, sleep cycle and diurnal hormone levels, slow accumulation of extracellular K+, changes in intracellular osmolarity, changes in redox state, and even transcriptional changes [
At the long end of the seizure forecast window Proix et al., using chronic intracranial EEG data from patients with implanted responsive neurostimulators, described a multidien pattern to seizure frequency that was specific to the patient and allowed for a prediction window of 24–72 h [
]. Clinically speaking, this risk window may lose some specific practical application for patients, for instance, driving restrictions may be overly burdensome compared with having only one to two hours of warning.
4.3 Mammalian menthone?
“Absence of evidence is not evidence of absence.”
Isoprenoids, (also known as terpenes), are the largest class of natural products with over 55,000 constituents with properties and uses ranging from cell wall and membrane biosynthesis, electron transport, photosynthetic pigments, plant and animal pheromones, pharmaceuticals, neutraceuticals, and even biofuels; all of this chemical and biological diversity is accomplished from the surprising feature that all known isoprenoids are derived from one of two C5 precursors: isopentyl diphosphate (IPP), or dimethylallyl diphosphate (DMAPP) which can be produced by either of two routes: the mevalonic acid pathway or the 2C-methyl-d-erythritol 4-phosphate (or MEP) pathway [
]. As a rule of thumb the mevalonate pathway is prevalent in eukaryotes and archaeabacteria, and the non-mevalonate pathway is widespread in eubacteria [
]. As mentioned before, menthone is well described in the plant chemistry literature for its industrial and financial benefits. Following condensation of IPP and DMAPP by geranyl diphosphate synthase the cyclic monoterpene precursor geranyl diphosphate (GPP) is produced (See Fig. 1). In the human cholesterol synthetic pathway, this precursor is further processed into squalene and other intermediates (See Fig. 2); however, as far as we are aware, the presence of a GPP cyclization enzyme that could take the monoterpene precursor down the menthone pathway has not been characterized in humans; despite what our study is suggesting. This area of investigation remains fertile.
5. Conclusion
Our simple hypothesis tests confirm with high sensitivity (82%), specificity (100%), and inter-rater agreement (92%) that fear-scented sweat provoked from non-PWE smelled like the seizure scent of PWE to the canine detection team imprinted to recognize the scent (predominantly of menthone) from our prior study [
Maa EH, Arnold J, Ninedorf K, Olsen HA. Canine detection of volatile organic compounds unique to human epileptic seizure. Epilepsy Behav 2021;115:107690, https://doi.org/10.1016/j.yebeh.2020.107690.
To answer the original question whether menthone is produced from endogenous or exogenous sources, it is important to first remember the CA/FIU study identified what was thought to be “unique” VOCs to epileptic seizure [
Davis P. The investigation of human scent from epileptic patients for the identification of a biomarker for epileptic seizures [PhD dissertation on the internet]. Florida International University; 2017. Available from: https://search.proquest.com/docview/2103114969.
]. Menthone and the other eight compounds reported in their paper were not part of the non-PWE or interictal PWE sweat profiles. It was only part of the ictal PWE sweat profile; however, in our current study we believe we have demonstrated this scent can be provoked and is identifiable in non-PWE subjects as well.
While this does not confirm the endogenous or exogenous source of menthone, it appears to confirm the idea that the circuit that produces this scent in seizure is native to humans in general and not specific to PWE, and that the network is likely involved with fear perception and fear response. Additionally, this circuit is common to both focal and genetic generalized seizures, and may not actually require the clinical or electrographic evolution of seizure to be completed (Table 4).
Identification, communication, and response to fearful stimuli are some of the most primitive functions of self and conspecific preservation of living systems and are present throughout the plant and animal kingdom. In humans, fear signaling has been documented in our earliest literature, recently demonstrated with the scientific method [
]. Serendipitously, our canine scent detection studies in epilepsy have revealed a specific compound we believe to not only represent an early biomarker in the evolution of an epileptic seizure, but because there are “no neural circuits unique to seizures” and “seizure activity must be conducted along existing circuits of normal physiologic processes” we believe we may have identified menthone as a potential human alarm pheromone. Further study by independent groups replicating our research on a much larger scale will hopefully confirm these findings and expand our knowledge in the fields of epilepsy and fear psychology.
Acknowledgements
I wish to thank Jennifer Arnold and Canine Assistants for their ongoing enthusiasm in exploring the extent of the canine scent detection field; the Brooke Gordon Family for their continued support of the Denver Health epilepsy program; Katie Ninedorf who continues to help with all of our scent collections, and Dr. Mark Spitz my mentor in all things epilepsy. I would especially like to thank my son, Sebastian, who after weeks of cajoling, finally talked me into watching the movie “It” with him. It was the hair-raising sensation while watching the movie that resulted in formulating the hypothesis of this research paper.
Disclosures
None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
References
Gale K.
Focal trigger zones and pathways of propagation in seizure generation.
(Chapter 2)in: Schwartzkroin Epilepsy: Models, mechanisms, and concepts. Cambridge University Press,
2007: 48-90
Davis P. The investigation of human scent from epileptic patients for the identification of a biomarker for epileptic seizures [PhD dissertation on the internet]. Florida International University; 2017. Available from: https://search.proquest.com/docview/2103114969.
Maa EH, Arnold J, Ninedorf K, Olsen HA. Canine detection of volatile organic compounds unique to human epileptic seizure. Epilepsy Behav 2021;115:107690, https://doi.org/10.1016/j.yebeh.2020.107690.
Verheggen FJ, Haubruge E, Mescher MC. Alarm pheromones-chemical signaling in response to danger. In Gerald Litwack, ed: Vitamins and hormones, Vol. 83, Burlington: Academic Press, 2010, pp. 215-240.
00312-7&id=gr1.jpg){kind=link}
00312-7&id=gr2.jpg){kind=link}