Specialized mechanoreceptor systems in rodent glabrous skin

Abstract

Key points: An ex vivo preparation was developed to record from single sensory fibres innervating the glabrous skin of the mouse forepaw. The density of mechanoreceptor innervation of the forepaw glabrous skin was found to be three times higher than that of hindpaw glabrous skin. Rapidly adapting mechanoreceptors that innervate Meissner's corpuscles were severalfold more responsive to slowly moving stimuli in the forepaw compared to those innervating hindpaw skin. We found a distinct group of small hairs in the centre of the mouse hindpaw glabrous skin that were exclusively innervated by directionally sensitive D-hair receptors. The directional sensitivity, but not the end-organ anatomy, were the opposite to D-hair receptors in the hairy skin. Glabrous skin hairs in the hindpaw are not ubiquitous in rodents, but occur in African and North American species that diverged more than 65 million years ago.

Abstract: Rodents use their forepaws to actively interact with their tactile environment. Studies on the physiology and anatomy of glabrous skin that makes up the majority of the forepaw are almost non-existent in the mouse. Here we developed a preparation to record from single sensory fibres of the forepaw and compared anatomical and physiological receptor properties to those of the hindpaw glabrous and hairy skin. We found that the mouse forepaw skin is equipped with a very high density of mechanoreceptors; >3 times more than hindpaw glabrous skin. In addition, rapidly adapting mechanoreceptors that innervate Meissner's corpuscles of the forepaw were severalfold more sensitive to slowly moving mechanical stimuli compared to their counterparts in the hindpaw glabrous skin. All other mechanoreceptor types as well as myelinated nociceptors had physiological properties that were invariant regardless of which skin area they occupied. We discovered a novel D-hair receptor innervating a small group of hairs in the middle of the hindpaw glabrous skin in mice. These glabrous skin D-hair receptors were direction sensitive albeit with an orientation sensitivity opposite to that described for hairy skin D-hair receptors. Glabrous skin hairs do not occur in all rodents, but are present in North American and African rodent species that diverged more than 65 million years ago. The function of these specialized hairs is unknown, but they are nevertheless evolutionarily very ancient. Our study reveals novel physiological specializations of mechanoreceptors in the glabrous skin that likely evolved to facilitate tactile exploration.

Keywords: evolution; mechanoreceptor; touch sensation.

Figure 1. Innervation areas of the saphenous, tibial and median/ulnar nerves

Left: inside out configuration of the hindpaw hairy skin (A), the hindpaw glabrous skin (D) and the forepaw glabrous skin (G) preparation. Nerves indicated in red were not used and the location of the cut end is marked ‘X’. Middle: receptor locations of single units recorded from the saphenous nerve (B), the medial and lateral plantar nerve (E), which are two branches of the tibial nerve, and the median (red area) and ulnar nerves (green area) (H) are indicated as well as the skin territory with an overlapping innervation (brown). Blue circles indicate single‐unit receptive field centres (compiled data from the current study, data recorded earlier (Milenkovic et al. 2014) and unpublished experiments). D, digits; H, hypothenar pads; I, interdigital pads; T, thenar pads. Right: outside‐out configuration used in electrophysiological experiments (illustrated as mirror images). Dotted lines indicate the receptive fields of the saphenous nerve (C), the lateral and medial plantar nerve (F) and the median and ulnar nerve (I).

Figure 2. Stimulation protocols

A, example image of the stimulation motor (piezo actuator, Physik Instrumente) and the force feedback system (force sensor, Kleindiek Nanotechnik). B, left, schematic illustration of the sigmoidal vibrating stimulus (20 Hz) with increasing intensity. Right, example trace of a receptor fibre responding to the vibrating stimulus. The force at the time of the first action potential was measured. C, left, schematic illustration of the ramp and hold stimulation; four different velocities were used. Right, example trace of a receptor fibre responding to a ramp and hold stimulation; only the spikes at the dynamic phase of the stimulation were measured. D, left, schematic illustration of the ramp and hold stimulation; four different intensities were used. Right, example trace of a receptor fibre responding to a ramp and hold stimulation; only the spikes during the static phase of the stimulation were quantified.

Figure 3. Response properties of RAMs

A, top, schematic drawing of the predominant RAM anatomy in saphenous nerve preparation (hair follicle receptor), tibial nerve preparation (Meissner's corpuscle) and median/ulnar nerve preparation (Meissner's corpuscle). Bottom, representative example traces of RAMs in response to a ramp and hold stimulation with a velocity of 0.45 mm s⁻¹. B, minimal stimulation force needed to evoke an action potential in response to increasing amplitude vibrating stimuli (20 Hz); ANOVA: P > 0.05; error bars represent SEM. C, average spike frequency in response to moving stimuli. Repeated measures ANOVA: P < 0.0001; Bonferroni post hoc tests are indicated; ^*** P < 0.001, ^** P < 0.01, ^* P < 0.05; error bars represent SEM.

Figure 4. Meissner's corpuscle anatomy

A and B, sagittal vibrating blade microtome sections of the interdigital pad IV of the hindpaw (A) and the interdigital pad III of the forepaw (B). Top, immunofluorescence image of a running pads with labelled Meissner's corpuscle (anti‐S100) and myelinated nerve fibres (anti‐NF200). Middle and bottom, magnified representation of single Meissner's corpuscles. C, cumulative frequency plot of the number of fibres innervating a single Meissner's corpuscle. D, size (volume) of Meissner's corpuscle in the hind‐ and forepaw running pads; error bars represent SEM. E, Meissner's corpuscle density in the interdigital running pads III (hindpaw) and IV (forepaw); paired t test: P = 0.0188; error bars represent SEM. F, number of myelinated axons counted in the tibial and the median plus the ulnar nerve; error bars represent SEM. G, representative semi‐thin microscopy images used to quantify myelinated axon number. All scale bars in panels A and B are 50 μm.

Figure 5. Receptor properties of SAMs

A, representative example trace of a SAM response (hindpaw glabrous skin) to a ramp and hold stimulation. B, average spike count (bin 0.1 s) to a suprathreshold mechanical stimulus over 2 s duration. C, average spike frequency of SAMs in response to moving stimuli; repeated measures ANOVA: P > 0.05; error bars represent SEM. D, average spike frequency of SAMs in response to ramp and hold stimulation; repeated measures ANOVA: P > 0.05; error bars represent SEM. E, minimal force needed to evoke an action potential of SAMs in response to 20 Hz vibrating stimuli; repeated measures ANOVA: P > 0.05; error bars represent SEM. F, representative example trace of a AM response (hindpaw glabrous skin) to a ramp and hold stimulation. G, average spike frequency of AMs in response to ramp and hold stimulations; repeated measures ANOVA: P > 0.05; error bars represent SEM. H, minimal force needed to evoke an action potential in AMs in response to a fast moving ramp stimulation; repeated measures ANOVA: P > 0.05; error bars represent SEM.

Figure 6. Functional and anatomical properties of glabrous skin D‐Hair receptors

A, representative example traces from D‐hair receptor recordings in response to a ramp and hold stimulation with a velocity of 0.45 mm s⁻¹. B, average spike frequency in response to moving stimuli. Repeated measures ANOVA: P > 0.05; error bars represent SEM. C, minimal stimulation force needed to evoke an action potential in response to increasing vibrating stimuli (20 Hz); ANOVA: P > 0.05; error bars represent SEM. D, hairs at the glabrous hindpaw skin. Right panel, magnification. E, experimental set‐up of the hair deflection experiment: the mounted hair can be deflected in the direction towards the toe, the heel or sideways. Right panel, magnification to display the hair partially inside the glass capillary. F, D‐hair receptors respond with more action potentials (average spikes) and respond more reliable (percentage of receptors responding) to small hair deflections in the direction towards the toe compared to the direction of the heel. Error bars represent SEM. G, CaV3.2 positive nerve fibres cluster at one side of the follicle (whole‐mount preparation). H, whole‐mount hair follicle receptor staining; terminal Schwann cells in green (anti‐S100), myelinated nerve fibres in red (anti‐NF200) and autofluorescence in blue.

Figure 7. Glabrous skin hair receptors are found in North American and African rodent species

A–F, representative pictures of the glabrous hindpaw skin from laboratory rodents (top row) and three wild caught rodent species from Africa and North America. A and B, images of these very fine hairs in two laboratory inbred mouse strains, C57BL/6J (A) and CBA/J (B) mice. C, we observed no hairs on the glabrous skin of laboratory rats. D and E, sparse fine hairs were observed on the hindpaw glabrous skin of the North American white footed mouse (D) and quite dense hairs were found on the same region of the Grasshopper mouse (E), which has its habitat in the Arizona desert. F, very distinctive glabrous hairs were also observed on the glabrous skin of the Damaraland mole‐rat.

Figure 8. Rodent species that lack glabrous skin hairs

A, one North American rodent species that lacks glabrous D‐hair receptors. B–K, 10 African rodent species that clearly lack hindpaw glabrous hairs.