Hypergravity within a critical period impacts on the maturation of somatosensory cortical maps and their potential for use-dependent plasticity in the adult

Abstract

We investigated experience-dependent plasticity of somatosensory maps in rat S1 cortex during early development. We analyzed both short- and long-term effects of exposure to 2G hypergravity (HG) during the first 3 postnatal weeks on forepaw representations. We also examined the potential of adult somatosensory maps for experience-dependent plasticity after early HG rearing. At postnatal day 22, HG was found to induce an enlargement of cortical zones driven by nail displacements and a contraction of skin sectors of the forepaw map. In these remaining zones serving the skin, neurons displayed expanded glabrous skin receptive fields (RFs). HG also induced a bias in the directional sensitivity of neuronal responses to nail displacement. HG-induced map changes were still found after 16 wk of housing in normogravity (NG). However, the glabrous skin RFs recorded in HG rats decreased to values similar to that of NG rats, as early as the end of the first week of housing in NG. Moreover, the expansion of the glabrous skin area and decrease in RF size normally induced in adults by an enriched environment (EE) did not occur in the HG rats, even after 16 wk of EE housing in NG. Our findings reveal that early postnatal experience critically and durably shapes S1 forepaw maps and limits their potential to be modified by novel experience in adulthood.

experience-dependent shaping of morphological and functional brain circuits leads to appropriately organized neural networks that subserve adult brain functions. The effect of the so-called epigenetic factors is most prominent within postnatal “critical periods” (Hensch 2004; Knudsen 2004), during which the central nervous system shows a powerful structural and functional plasticity, reflecting a process of adaptation of genetically specified neural properties to environmental constraints. The barrel cortex has been extensively investigated as a model of the experience-dependent refinement of the primary somatosensory (S1) cortex during development. Damage to whisker follicles during the first few days after birth prevents the formation of barrels within layer 4 of the S1 cortex (O'Leary et al. 1994; Woolsey and Wann 1976; for a review see Erzurumlu 2010; Erzurumlu and Gaspar 2012). By contrast, whisker trimming does not affect barrel-pattern formation (Fox 1992; Henderson et al. 1992; Rema et al. 2003; Simons and Land 1987) but impairs the neurophysiological properties of cortical neurons (Fox 1992; Rema et al. 2003; Shoykhet et al. 2005; Simons and Land 1987; Stern et al. 2001) as well as whisker discrimination (Carvell and Simons 1996). Surprisingly, few studies have investigated the maturation of body somatotopic maps in the S1 cortex (Armstrong-James 1975; McCandlish et al. 1993; Seelke et al. 2012), and little is known about the experience-dependent plasticity of neural circuits underlying the postnatal development of these somatosensory maps.

Somatosensory cortical maps reflect a dynamic competition between proprioceptive and cutaneous inputs converging onto cortical networks (Coq and Xerri 1998; Jenkins et al. 1990; Recanzone et al. 1992; Xerri et al. 1994). Gravity is a fundamental element in the external environment, stimulating vestibular receptors and exerting mechanical constraints on the musculoskeletal system while providing a reference frame for posture and movements. We reasoned that changes in gravity modify the relative weight of proprioceptive and cutaneous inputs from the limbs converging on the S1 cortex and thus provide a unique opportunity to investigate the developmental emergence of somatosensory maps in an environment modifying inputs from the limbs, without sensory deprivation.

Decades of research have shown that environmental conditions and sensorimotor training shape the morphological and functional organization of sensory and motor cortical networks throughout life, though to a lesser extent than during developmental stages (for review see Barnes and Finnerty 2010; Buonomano and Merzenich 1998; Xerri 2008). An important conceptual challenge is to decipher the interplay between developmental and adult experiential plasticity.

Our first aim was to examine both short- and long-term effects of early exposure to hypergravity (HG) on rat forepaw representation. HG has been shown to affect the development of the vestibular (Ross and Tomko 1998) and motor systems (Brocard et al. 2003) and their functions such as air righting during free fall, locomotion, postural control, and limb movement when the HG animals were subsequently exposed to normogravity (Bojados et al. 2013; Bouët et al. 2003, 2004). The enduring motor changes found in rats that grew up in altered gravity stress the key role of gravity during the motor development and suggest the existence of critical periods (Walton et al. 1992). However, it is still unknown whether HG has an impact on the emergence of somatosensory maps. During the developmental period, major morphological and biochemical changes take place as behavior progresses through different stages of locomotion. The rat motor system matures slowly over the first postnatal weeks; the pups do not walk spontaneously before the end of the second week of life (Altman and Sudarshan 1975), and clear topographical organization of forelimbs begins to appear by postnatal day 15 (P15) in S1 (Seelke et al. 2012). Changes in gravity induce alteration in joint and muscle proprioception (McCall et al. 2003). Consistently, HG during early development has been shown to modify postural control and locomotion in rodents (Bojados et al. 2013). Therefore, we hypothesized that HG, by modifying the pressure exerted on the ventral surfaces of the paws and muscle spindle stimulation, thereby changing the pattern of cutaneous and proprioceptive inputs to the S1 cortex compared with normogravity, would shape the topographical organization of the forepaw maps in a distinctive way.

Our second aim was to determine whether early sensory experience constrains experience-dependent plasticity in the mature cortex. We reasoned that gravity-induced changes in somatosensory inputs offer insight into how early developmental experience impacts cortical map remodeling resulting from novel sensory experience at maturity. There has been no equivalent earlier study on the effect of neonatal exposure to HG on later use-dependent plasticity of the mature somatosensory cortex. Only one related study examined the impact of neonatal deprivation on use-dependent plasticity of the adult rat barrel cortex (Rema et al. 2003). These authors have shown that trimming two adjacent whiskers from birth for 21 days reduced responses of cortical layer II/III neurons at maturity. This early deprivation was also shown to impede the adult sensory plasticity induced by paired use of the neonatally trimmed whiskers in supragranular neurons. We investigated the effect of early HG rearing on experience-dependent cortical plasticity in adult animals housed in an enriched environment known to induce structural and biochemical changes in the mature cortex (for reviews see Rosenzweig and Bennett 1996; van Praag et al. 2000) as well as to remodel the forepaw representation in the S1 cortex (Coq and Xerri 1998). Our hypothesis was that the potential of cortical maps for plastic changes in response to enriched environment at maturity would be constrained by early somatosensory experience in HG during a critical period.

MATERIALS AND METHODS

Animals

Principles of laboratory animal care were respected and experiments carried out in accordance with Directive 2010/63/EU of the European Parliament and of the Council of 22 September 2010 on the protection of animals used for scientific purposes. The authors have been granted a license from the Ministère de l'Enseignement Supérieur et de la Recherche to conduct animal research. In addition, the research protocol of the present study has been approved by a national review committee (01677.01). For the sake of clarity, the labels of the experimental groups are specified in Table 1.

Table 1.

Labels for experimental groups

Rats Reared in NG (Control Cages) Until P21
NG	Rats reared in NG until P21
NG1wS	Rats reared in NG until P21 and then housed in standard environment for 1 wk
NG8wS	Rats reared in NG until P21 and then housed in standard environment for 8 wk
NG16wS	Rats reared in NG until P21 and then housed in standard environment for 16 wk
NG16wE	Rats reared in NG until P21 and then housed in enriched environment for 16 wk
NGp	Rats reared in NG until P21, used for forepaw print measurements in HG or NG

Rats Reared in HG (2G) Until P21
HG	Rats reared in HG until P21
HGn-	Rats reared in HG until P21 with trimmed forepaw nails
HG1wS	Rats reared in HG until P21 and then housed in standard environment for 1 wk
HG8wS	Rats reared in HG until P21 and then housed in standard environment for 8 wk
HG16wS	Rats reared in HG until P21 and then housed in standard environment for 16 wk
HG16wE	Rats reared in HG until P21 and then housed in enriched environment for 16 wk
HGp	Rats reared in HG until P21, used for forepaw print measurements in HG or NG

HG, hypergravity; NG, normogravity; P21, postnatal day 21.

To investigate the effects of a hypergravity exposure during the developmental period on forepaw representation in the S1 cortex, three experimental groups of male Long-Evans rats were used in which cortical mapping was carried out at postnatal day 22 (P22): rats reared in normogravity (NG group), rats fertilized, born, and housed in a hypergravity (HG; 2G) environment generated by a centrifuge (see below) until the age of 3 wk (P21; HG group), and rats reared in HG with trimmed forepaw nails (HGn⁻ group) (Fig. 1A). To constitute HG and HGn⁻ groups, couples of rats were placed in the centrifuge generating the HG environment. After 1 wk of HG exposure, adult males were taken out. After birth, the newborn rats remained with their mother, and the pups were nursed normally until P22. The centrifuge was stopped once a week during a half hour for food renewal, water supply, and home cage cleaning. In the HGn⁻ group, the centrifuge was stopped every 2 days, from P5 until P22, and the animals were removed from the cages for about 10 min for trimming of forepaw nails. The NG animals were reared in gondolas similar to those equipping the centrifuge, with identical bedding (sawdust), in the same room, and under the same conditions of room temperature, dark-light cycle, and ambient noise, but without centrifugation. A fan was used to generate air flow through openings in the front door of these static gondolas to replicate the rotation-induced airflow experienced by the HG rats.

Fig. 1.

Experimental protocols and procedure. Synoptic representations show the experimental protocols used to evaluate short-term and enduring effects of early exposure to hypergravity on the development of forepaw cortical maps in S1 (A) and its influence on adult map use-dependent plasticity (B). Rats were born and housed in hypergravity (HG; 2G) generated by a centrifuge (C) until the age of 3 wk. Electrophysiological mapping of the forepaw representation was performed at the 22nd postnatal day (P22) in HG rats and HGn⁻ rats (with trimmed forepaw nails) and after 1 wk (HG1wS rats), 8 wk (HG8wS rats), or 16 wk (HG16wS rats) spent in normogravity (NG). The maps were compared with those obtained in age-matched groups of rats reared in NG from birth (NG, NG1wS, NG8wS, and NG16wS groups). Paw print measurements were performed in 2 additional groups of rats reared in HG or NG and tested under both NG and HG conditions (D). Mapping data were also recorded in rats raised in either HG or NG during the first 21 postnatal days and then maintained in an enriched environment (E) for 16 wk (HG16wE and NG16wE rats, respectively). The results were compared with those obtained in rats raised in either HG or NG during the first 21 postnatal days and then housed in a standard environment (HG16wS and NG16wS rats, respectively).

The enduring effects of HG were studied in rats born and reared in HG until P22, and then rats were removed from the centrifuge and placed in NG under standard laboratory housing conditions for 1 wk (HG1wS group), 8 wk (HG8wS group), or 16 wk (HG16wS group) before electrophysiological mapping was performed. These groups were compared with age-matched groups of rats reared under NG conditions from birth, placed for 3 wk in static gondolas and then for 1 wk (NG1wS group), 8 wk (NG8wS group), or 16 wk (NG16wS group) in standard cages (Fig. 1A).

The influence of early HG rearing on the use-dependent plasticity of the adult forepaw S1 map was analyzed by comparing the effect of an enriched environment (EE) in rats raised in NG vs. HG during the first 3 postnatal weeks. In the EE, rats were housed in a group of 12 in two large cages (76 cm wide × 76 cm long × 40 cm high) connected by two tunnels and containing objects of different shapes, sizes, and textures to promote tactile experience; these objects were renewed daily to stimulate exploratory behavior in the rats. Animals housed in standard cages were in groups of three in Plexiglas cages (26.5 × 42.5 × 18 cm) without objects. The rats had food and water ad libitum and were housed on a 12:12-h light-dark cycle (see Coq and Xerri 1998). Rats born and raised in either HG or NG during the first 21 postnatal days and then maintained for 16 wk in an EE (HG16wE and NG16wE, respectively) were compared with those born and raised in either HG or NG during the first 21 postnatal days and housed in a standard environment for 16 wk (HG16wS and NG16wS, respectively) (Fig. 1, B and E).

Electrophysiological mapping was performed in all these experimental groups (10 rats per group; total: 110 rats). To determine the effects of hypergravity on the paw skin surfaces in contact with the floor during stance and locomotion, forepaw print measurements were performed at P22 in the NGp and HGp rats. These measurements were performed in additional animals reared in HG (HGp group) or NG (NGp group) during the first 21 postnatal days (10 rats per group; total: 20 rats) and tested under HG or NG conditions at P22 and under NG conditions after 1 and 2 wk spent in NG. Rats attributed to the same experimental groups were taken from different litters.

Hypergravity Environment

The apparatus generating the HG environment (see Bouët et al. 2003, 2004) consisted of a centrifuge made of four free-swinging gondolas fixed to the four extremities of two horizontal cross-arms driven at a constant rotation speed by a DC motor (3.5 kW) located in the vertical axis of the centrifuge. The gondolas (0.55/0.38/0.30 m³), located 76.5 cm from the axis of rotation, were equipped with standard home cages for rats, including an aeration and lighting system providing a 12:12-h light-dark cycle. Gondolas were equipped with a video system, composed of a wide-angle camera connected to a monitor, that was used to establish the birth date of the litters. The gravito-inertial vector of 2G inside the gondola was produced with an angular velocity of 3.81 rad/s, which resulted in a constant tilt angle of the gondolas of 60° from the horizontal axis. The resultant force was measured using a force sensor located in the center of the floor of the gondola. During centrifugation, rats were therefore subjected to a vector perpendicular to the gondola's floor, i.e., similar to that experienced in normal gravity (corresponding to the dorsoventral axis of the animal) (Fig. 1C). The rats were totally free-moving in the home cage. Accordingly, they were additionally exposed to variable Coriolis forces depending on the speed and direction of the animal's motion within the home cage. Video monitoring of the nursing rats and their pups in both NG and HG groups was performed 3 times per week during 15-min observation periods. In the HG rats exposed to NG, postural and locomotor behavior was observed twice a day for 15 min during the first week.

Electrophysiological Mapping

Surgical procedure.

Anesthesia was induced with halothane and an initial injection of pentobarbital sodium (30 mg/kg ip). Rats were maintained at an areflexive level of anesthesia throughout the experiment by supplemental doses of pentobarbital sodium (3 mg/kg ip). The core temperature was continuously monitored by a rectal thermistor probe and maintained between 37°C and 38°C by a heating pad. The head was placed in a stereotaxic frame. To prevent cerebral edema, cerebrospinal fluid was first drained through an opening in the dura covering the foramen magnum after resection of posterior neck muscles. A craniotomy (about 16 mm²) was then made to expose the S1 cortex to be mapped. The dura protecting the exposed part of the cortex was incised and resected. The cortical surface was bathed in a thin layer of warm silicone fluid to prevent drying. At the end of the mapping session, rats received a lethal dose of pentobarbital sodium (150 mg/kg ip) and the brain was prepared for histological processing.

Electrophysiological mapping procedure.

The mapping procedure used in the present study has been described in detail previously (Xerri and Zennou-Azogui 2003). Magnified images of the exposed parietal cortex, and the ventral and dorsal surfaces of the forepaw contralateral to the cerebral hemisphere to be mapped, were digitized using a high-resolution camera mounted on an operating microscope. Cutaneous receptive fields (RFs) were drawn on the digitized images of the forepaw, and placements of microelectrode penetrations were recorded on the digitized image of the cortex using Map 0.925 software (Peterson and Merzenich 1995). Conventional multiunit recording and RF mapping techniques were used to reconstruct the forepaw representation. Sites of electrode penetration were identified relative to the vasculature of the cortical surface. Unit clusters were recorded with parylene-coated tungsten microelectrodes (about 1 MΩ at 1 kHz; WPI, Hitchin, UK). The electrode was moved perpendicular to the cortical surface in Cartesian coordinates by a three-dimensional (3-D) stepping micromanipulator (Märzhäuser; FST, North Vancouver, BC, Canada). Using the recording artifact generated by the microelectrode contact with the cortex surface as a zero level, we advanced the electrode to a depth of about 650–700 μm to record responses from clusters of two to four neurons in layer IV (Coq and Xerri 1998; Waters et al. 1995). The inter-electrode penetration distance was close to 70–80 μm in all groups of rats. The amplitude of the background noise usually ranged from 15 to 20 μV with a signal-to-noise ratio ranging from 4 to 6. The multiunit signal was amplified, filtered (bandwidth: 0.5–5 kHz), and displayed on an oscilloscope. This signal was also rectified and passed through a discriminator with an output signal proportional to the part of the input signal that was higher than an adjustable threshold set just above the background noise. The output of the discriminator was then delivered to an audio monitor. At each recording site, bursts of activity elicited by natural stimulation allowed us to classify neuronal responses as cutaneous or noncutaneous. The cutaneous RFs of small clusters of neurons were defined at each recording site as the areas of skin where just-visible skin indentation (about 100–150 μm) elicited reliable changes in multiple-unit discharge. This stimulation, which is within the dynamic range of cutaneous receptors (Gardner and Palmer 1989), was produced with a fine-tipped, hand-held glass probe and monitored using magnifying glasses (×4). The curvature of fingers and pads was taken into account while the RFs were drawn. The ridges running along the glabrous skin of the digits and palm were used as reliable landmarks to delineate the RFs. The RF areas were transferred to the digital image of the forepaw and were subsequently measured offline using Map 0.925 software. Responses elicited by upward and downward nail displacements were examined while the tips of the digits were firmly maintained to minimize joint movement and skin deformation. As a general rule, light pressure on the nails elicited a burst of activity facilitating the assessment of neurons' directional sensitivity, i.e., the direction of nail displacement (upward, U; downward, D; or both, U+D) eliciting the most prominent neuronal responses. These responses were categorized as cutaneous and referred to as “nail” responses. In our classification, high-threshold responses identified by taps and pressure on tendons, intrinsic muscles, or joint manipulations, while no cutaneous response was found, were classified as proprioceptive (Coq and Xerri. 2000; Jenkins et al. 1990; Recanzone et al. 1992; Xerri et al. 1998).

The mechanical thresholds of neuronal responses to cutaneous glabrous stimulation were determined using von Frey monofilaments (Semmes-Weinstein aesthesiometer; Stoelting, Wood Dale, IL) that apply indenting stimuli at a relatively constant, predetermined force. The most commonly used filaments were 2.83 (diameter, 0.127 mm; bending force, 0.068 g; bending pressure, 5.37 g/mm²), 3.22 (0.152 mm; 0.166 g; 9.15 g/mm²), 3.61 (0.178 mm; 0.407 g; 16.36 g/mm²), and 3.84 (0.203 mm; 0.692 g; 21.39 g/mm2) (see Fig. 2B). Neuronal responses elicited by von Frey filaments above 3.84, starting from 4.08 (0.229 mm; 1.202 g; 29.20 g/mm²), were classified as noncutaneous, presumably proprioceptive. For reproducible measurements, the filaments were used at a relatively constant room temperature (about 24°C). The stimulation consisted in pressing a filament gently against the skin, perpendicular to its surface and at the center of the RF, until the filament began to bend. This procedure was done 5–10 times for each filament. We used stimulus series of increasing and decreasing strengths to determine the mechanical threshold eliciting noticeable changes in neuronal discharge.

Fig. 2.

Electrophysiological map of cutaneous surfaces obtained in a 2-wk-old rat. Map elaboration is based on the response characteristics (somatosensory submodality, location, and size) of the receptive fields (RFs) of neurons recorded within layer IV of the S1 cortex. A: magnified image of the forepaw area of the cortex with a superimposed drawing of the boundaries encompassing cortical sites, the RFs of which were restricted to a common forepaw subdivision. Within these boundaries, colored circles indicate cortical sites where cutaneous hairy and glabrous skin responses were elicited, whereas colored squares show cortical sites where neuronal responses were elicited by nail displacements. Green and red triangles mark cortical sites excited by lower lip and wrist stimulation, respectively. Violet squares refer to cortical sites classified as “proprioceptive” (responses to strong pressure on skin and tendons and/or joint movement). Black squares mark cortical sites with no stimulus-evoked discharges. B: photomicrograph (top) of Nissl-stained section from the forepaw area of cortex with an electrolytic lesion in layer IV (arrow) and recording trace (bottom) illustrating a typical multiunit discharge elicited by a von Frey filament (3.22: 0.166 g, 9.15 g/mm²) applied at the center of an RF on digit 3. C: colored sectors of the cutaneous map with black dots indicating cutaneous (skin or nail sensitive) cortical sites. Maps recorded at P15 appeared to be adultlike in their topographical organization. D: samples of cutaneous RFs on glabrous and hairy skin surfaces corresponding to the cortical sites illustrated in A and C.

After the recording session, we used Canvas software (ACD Systems) to produce maps of the forepaw representation by drawing boundaries encompassing cortical sites where corresponding RFs were restricted to a common forepaw subdivision, i.e., finger, palmar, pad. Borders were drawn midway between adjacent recording sites where RFs were located on distinct and separate skin subdivisions. The same principle was used to draw boundaries encompassing cortical recording sites where noncutaneous responses were obtained. A boundary line crossed cortical sites at which a single RF included different but adjoining skin subdivisions of the forepaw. Borders encompassing cortical sites where fingernail responses were obtained were also drawn. A boundary line crossed cortical sites at which both cutaneous and nail responses were recorded (Fig. 2, A and C). The cutaneous forepaw map boundary was delimited by cortical sites exhibiting cutaneous responses to stimulation of the lower lip or wrist or noncutaneous responses, and by no stimulus-evoked sites. Canvas software was used to calculate the surface area of each cutaneous and nail region of the cortical map.

Experimental measurements.

Cortical areas devoted to forepaw representation, (i.e., excited by stimulation of glabrous or hairy skin areas or by nail displacements) were calculated for each rat and described by their absolute areas (mm²). To assess relative representation of glabrous, hairy skin surface, and nails as parts of the forepaw representation, the corresponding parts of the map were expressed as a percentage of the whole cutaneous map of the forepaw. Average values were computed for each group of rats. The absolute sizes of glabrous and hairy RFs were measured (in mm²; Map 0.925 software), normalized relative to the ventral and dorsal skin forepaw areas, respectively, and expressed as percentages. The relative RF areas measured in each rat were averaged, and the mean RF size was calculated for each group of rats. The stimulation thresholds obtained using the von Frey filaments were processed to calculate an average force threshold for each animal (Xerri and Zennou-Azogui 2003, 2014). Every threshold value (0.068, 0.168, 0.407, and 0.692 g) was multiplied by the number of recording sites displaying that threshold, and the sum of these weighted values was divided by the total number of sites tested. A mean threshold was then calculated for each experimental group. To assess directional sensitivity of fingernails, their displacement was induced using a hand-held electronic von Frey device, designed in our laboratory, to apply single stimuli through a probe (0.8 mm in diameter; Blanc and Coq 2007). The force envelope randomly applied (mN) by hand to induce nail extension or flexion was converted by the electronic von Frey device into a potential variation (0.1 mN corresponded to 10 mV), which was recorded at 1 kHz with a multichannel acquisition system (Plexon, Dallas, TX). Stimuli and neuronal responses were synchronized using a TTL signal emitted by the von Frey at the onset of the nail contact. The force profile and duration varied across trials. Only relatively similar force patterns were retained to determine the units' sensitivity to nail displacement. Neuronal activity was recorded in a 600-ms time window that started from the trial onset. We elicited 15–20 stimulus trials for each of the cortical locations recorded. The percentage of neuronal responses exhibiting clear directional sensitivity to either extension or flexion of nails, or no directional sensitivity, was calculated for each rat and averaged for each experimental group.

Histology

To ascertain the location of recording responses in S1 layer IV, several electrode tracks in each experimental group were marked with electrophysiological lesions in the center of the forepaw map by passing cathodal current (10-μA DC, 10 s) through the recording electrode positioned at recording depth. After the mapping session, the rats were given a lethal dose of pentobarbital sodium and perfused transcardially with 0.9% physiological saline followed by a solution containing 4% paraformaldehyde in 0.1% sodium phosphate buffer (pH 7.4). The brain was removed and postfixed in a 4% paraformaldehyde solution containing 10% sucrose in phosphate buffer. Coronal sections 50 μm thick were cut on a freezing microtome and processed for Nissl staining. Histological data confirmed the location of the recorded neurons in layer IV (Fig. 2B).

Paw Print Measurements

The animals stepped on an inkpad placed for 3 min in their rearing cage. The animals were then immediately anesthetized with halothane, and pictures of the left and right forepaws were taken using a charge-coupled device camera for subsequent measurement of inked paw surfaces. In addition, after cleaning of the stained skin surfaces and 1 h after recovery from initial anesthesia, NGp rats were tested in hypergravity, whereas HGp rats were tested in normogravity. Possible long-term effects of hypergravity on forepaw prints were also evaluated under normogravity conditions, 1 and 2 wk after NG housing, in NGp and HGp animals. Inked surfaces of the right and left forepaws were measured (in mm²) using Canvas software (ACD Systems). A mean forepaw print value was calculated for each rat and then for each group of rats.

Statistical Analysis

Results are means ± SD. Statistical treatment was done with analysis of variance (ANOVA) supplemented with multiple comparisons (Newman-Keuls post hoc test; Statistica software; StatSoft). Repeated-measures ANOVA was used to compare ipsative data (i.e., a set of responses that always sums to the same total: 100%) consisting of the proportions of neuronal responses to nail displacement falling within the defined categories (cf. Greer and Dunlap 1997). Repeated-measures ANOVA was also used to compare forepaw prints collected over time or under different gravity conditions in the same animals. The proportions of neuronal responses to nail displacement are ipsative data, i.e., a given set of responses always sums to the same total (100%). We used repeated-measures ANOVA to compare the proportions of responses obtained in HG and NG rats, a procedure that takes into account the inter-individual variability within each experimental group (Greer and Dunlap 1997; Shaffer 1981).

RESULTS

Methological Consideration: Coriolis Force in HG

Because the centrifuge used to generate HG also induced a Coriolis force, we had to consider its possible influence on the animals' behavior and effects on the forepaw. The Coriolis acceleration was computed using the following formula:

$$\overrightarrow{CA}=2\overrightarrow{\omega }\times \overrightarrow{v},$$

where CA stands for the Coriolis acceleration, v is the rat's velocity, ω is the angular velocity of the centrifuge, and × is the vector product.

Knowing that in our experimental conditions, ω = 3.81 rad/s and the cage deviation from the horizontal was 60°, the maximum rat velocity in their cage being estimated at about 0.10 m/s, we have been able to compute several cases, depending on the orientation of the rat's displacement. Obviously, the highest value of the Coriolis acceleration was obtained when the rat moved in the lengthwise direction of the cage, thus perpendicular to the ω vector. The CA value is ± 0.838 m/s² (the signs indicate the direction of the Coriolis acceleration). Thus it appears that the value of the additional force on the rat's paws (and nails) corresponded to an acceleration of 0.085G (0.025 N for a rat weighing 30 g), i.e., a small fraction of the 2G “vertical” acceleration. In the case of transversal direction of displacement in the cage, knowing that the speed vector was at 30° from the ω vector, the CA value is lower (± 0.419 m/s²) and corresponds to an acceleration of 0.043G. On the basis of these calculations, in our experimental conditions, the Coriolis forces were of little significance in relation to the hypergravity force. In addition, considering that the rat moves in various directions within the cage, the Coriolis acceleration acting on the forepaw has various orientations. Therefore, the mechanical effects of Coriolis forces on the forepaw will tend to cancel one another out.

Behavioral Observations

Regular video monitoring of the rats' behavior during the early postnatal period in both NG and HG rats did not reveal any behavioral modification in HG rats with respect to the control NG rats housed in gondolas. After the HG rats were exposed to NG, we did not detect any abnormalities in postural or locomotor behavior, other than an intensification of the exploratory behavior that was found for the first 2 days in NG. Importantly, no behavioral sign of stress (diminished grooming, exploratory activity, lactation, or nursing) was observed in HG rats compared with controls.

Organization of Forepaw Somatosensory Maps in 3-Wk-Old NG Rats

This study represents the first description of high-density electrophysiological forepaw maps obtained in rat pups at P22. We also obtained a complete forepaw map for two P15 rats (Fig. 2). Interestingly, this map and those recorded at P22 appeared to be adultlike in their topographical organization and RF sizes (Coq and Xerri 1998; Xerri and Zennou-Azogui 2003). We found a complete representation of the glabrous and hairy forepaw skin surfaces. The cutaneous representation displayed a somatotopic order, with a clearly segregated digit representation sequentially ordered from the rostrolateral to caudomedial cortical sector, and a palmar pad representation sequentially ordered along the medial edge of the map. The dorsal skin surface of the palm was contained in either a continuous or fragmented region at the lateral border of the forepaw map. Topographically organized cortical sectors displaying specific responses to nail displacements were found to be embedded within the representational regions of the forepaw skin. The borders of forepaw representation consisted of cortical sites either responsive to cutaneous stimulation of the lip or the arm or to noncutaneous inputs, or classified as unresponsive. Small islets of noncutaneous zones encountered in the maps tended to disrupt cutaneous topographical neighborhood relationships. As generally observed in adult rats, cortical cells clusters sampled in the representational cutaneous zone exhibited skin RFs covering glabrous or hairy skin surfaces generally restricted to one main subdivision of the forepaw (single digit, palmar pad).

Short-Term Effects of Early Exposure to Hypergravity

Organization of S1 forepaw maps.

This study represents the first investigation of the experience-dependent developmental plasticity of forepaw representations in the S1 cortex using high-density electrophysiological mapping. A conspicuous feature of somatosensory maps derived from HG rats was a threefold enlargement of cortical sectors driven by nail displacements (0.58 ± 0.12 mm²) compared with those of NG rats (0.19 ± 0.05 mm²) (F_1,18 = 81.12; P < 0.001). In the former, nail representation was found to take over the cortical zones serving the skin surfaces (Fig. 3A). Skin representations in HG rats (0.76 ± 0.14 mm²) were significantly smaller than those of NG rats (1.20 ± 0.21 mm²) (F_1,18 = 35.45; P < 0.001). This HG-induced decrease was found for both the glabrous (mean decrease: 36%; HG: 0.65 ± 0.15 mm²; NG: 1.02 ± 0.22 mm²; F_1,18 = 20.05; P < 0.001) and hairy skin areas (39%; HG: 0.11 ± 0.04 mm²; NG: 0.18 ± 0.05 mm²; F_1,18 = 11.77; P < 0.01). No significant change in the overall area of the forepaw cutaneous map (including skin and nail representational regions) was observed in HG rats compared with NG rats [HG: 1.33 ± 0.14 mm²; NG: 1.43 ± 0.21 mm²; F_1,18 = 1.45; P = 0.24, not significant (ns)].

Fig. 3.

Effect of early exposure to hypergravity on the development of the somatotopic cutaneous maps of the forepaw. A: representative maps obtained on P22 in a rat born and housed in normogravity (NG) or hypergravity (HG) and in a rat reared in HG with forepaw nails trimmed until the age of 3 wk (HGn⁻). Glabrous (B) and hairy skin RFs (C) of which the locations on the forepaw were used to reconstruct the cutaneous maps illustrated in A. Note the HG-induced increase in the size of glabrous RFs recorded in HG and HGn⁻ rats. Note in HG and HGn⁻ rats the HG-induced decrease in the cortical area serving skin surfaces of the forepaw. Nail representation was shown to take over the skin zones in the HG rat, whereas in the HGn⁻ rat, proprioceptive zones colonized the cortical territory, compensating for the decrease in nail and skin areas. Same conventions as in Fig. 2.

This finding led us to hypothesize that hypergravity increased sensory inputs generated by nail displacements. Therefore, we assessed the forepaw map reorganization in animals whose nails were regularly clipped (HGn⁻ rats). The cortical area activated by nail displacement in HGn⁻ rats (0.07 ± 0.06 mm²) was strongly reduced compared with that of HG (0.58 ± 0.12 mm²; F_2,27 = 98.06; P < 0.001) and NG rats (0.19 ± 0.05 mm²; P < 0.01). Interestingly, the glabrous skin sectors were similarly reduced in the HG (0.65 ± 0.15 mm²) and HGn⁻ (0.74 ± 0.08 mm²) groups (F_2,27 = 14.87; P = 0.19, ns). This reduction was found to be only partially compensated for by an increased hairy skin representation in the HGn⁻ rats (0.27 ± 0.09 mm²), which was greater than that of either NG (0.18 ± 0.05 mm²; F_2,27 = 16.34; P < 0.01) or HG rats (0.11 ± 0.04 mm²; P < 0.001). Therefore, the total surface area of cutaneous map (skin + nail regions) was smaller in the HGn⁻ (1.08 ± 0.13 mm²) than in the HG (1.33 ± 0.14 mm²; F_2,27 = 12.06; post hoc, P < 0.01) or NG groups (1.43 ± 0.21 mm²; P < 0.001). Proprioceptive zones were found to colonize the cutaneous territory in these HGn⁻ rats (Fig. 3A). Overall, these findings show that early exposure to HG strongly modified the relative representation of competing cutaneous (glabrous or hairy skin, nail) and proprioceptive inputs converging onto layer IV cortical neurons.

Neuronal responsiveness.

We reasoned that if early rearing in HG strongly modulates sensory inputs generated by nail movement during locomotion and changes in posture, this environment may bias the directional sensitivity of cortical neurons to upward (U) vs. downward (D) nail displacement. Repeated-measures ANOVA showed a main effect of directional sensitivity, i.e., the percentage of responses to U, D, or both U and D displacement (U+D; F_2,36 = 22.21; P < 0.001), no main effect of rearing environment (F_1,18 = 0.00001; P = 0.99, ns), and an interaction between these two factors (F_2,36 = 12.79; P < 0.001). Post hoc tests indicated that NG rats exhibited a majority of neuronal responses elicited by both U and D nail displacement compared with those elicited by U or D movement (P < 0.01 or P < 0.001, respectively), which were found in similar proportions (P = 0.39, ns; Table 2). By contrast, HG rats displayed a larger proportion of neurons sensitive to U nail displacement than to D (P < 0.001) or U+D nail displacement (P < 0.03). In HG rats, the neuronal population sensitive to U nail displacement increased (P < 0.001), whereas that sensitive to U and D movements decreased (P < 0.03), compared with those populations recorded in NG rats. No difference in D nail response proportion was shown between NG and HG rats (P = 0.07, ns; Fig. 4). These results suggest that the HG increase in the population of U responses resulted from an effect on U+D neurons. To assess neuronal responsiveness to glabrous skin stimulation, a weighted mean was calculated on the basis of the mechanical thresholds obtained using the von Frey filaments (see materials and methods). The mechanical thresholds to cutaneous stimulation recorded within the reduced glabrous skin cortical sectors in the HG rats were similar to those recorded in the NG rats (F_1,18 = 1.87; P = 0.19, ns; Table 3).

Table 2.

Directional sensitivity of cortical neurons to nail displacement

Groups	NG	NG1wS	NG8wS	NG16wS
U	27.8 ± 9.6% (26)	20.6 ± 2.1% (22)	22.3 ± 13.3% (26)	19.4 ± 9.5% (29)
D	22.4 ± 12.1% (18)	13.3 ± 3.5% (14)	10.3 ± 6.3% (12)	9.5 ± 8.1% (13)
U+D	49.8 ± 18.3% (42)	66.1 ± 10.9% (79)	67.4 ± 10.9% (82)	71.1 ± 10.2% (108)

Groups	HG	HG1wS	HG8wS	HG16wS
U	53.1 ± 5.6% (146)	50.6 ± 16.2% (135)	22.7 ± 7.9% (70)	20.8 ± 10.8% (69)
D	11.0 ± 8.7% (29)	8.6 ± 7.3% (24)	6.5 ± 6.1% (19)	8.8 ± 7.6% (27)
U+D	35.9 ± 9.0% (115)	40.8 ± 21.2% (111)	70.8 ± 12.5% (216)	70.4 ± 14.6% (234)

Values are mean (±SD) percentages; numbers in parentheses are no. of neuronal responses to upward (U), downward (D), or both U+D nail displacement obtained in rats reared in normogravity from birth (NG, NG1wS, NG8wS, NG16wS) and in aged-matched rats born and reared in hypergravity during the first 3 postnatal weeks at P22 (HG) and after 1 (HG1wS), 8 (HG8wS), or 16 wk (HG16wS) of housing in normogravity. See text for statistics.

Fig. 4.

Influence of hypergravity on neurons' directional sensitivity to nail displacement. A: classification of neuronal responses according to their sensitivity to nail upward (a) or downward (b) displacement induced by an electronic von Frey device generating force profiles. B: percentages of neuronal responses elicited by upward (U), downward (D), or both upward and downward (U+D) movement of the nails obtained on P22 in rats born and housed in normogravity or hypergravity (see Table 2 for mean percentages of neuronal responses within experimental groups). Note that hypergravity resulted in a higher percentage of neurons responding to nail upward displacement.

Table 3.

Neuronal responsiveness to glabrous skin stimulation

Groups	Mechanical thresholds, g
NG	0.203 ± 0.027
HG	0.233 ± 0.064
NG1wS	0.247 ± 0.058
HG1wS	0.231 ± 0.062
NG8wS	0.411 ± 0.057
HG8wS	0.365 ± 0.067
NG16wS	0.393 ± 0.049
HG16wS	0.343 ± 0.025
NG16wE	0.459 ± 0.061
HG16wE	0.427 ± 0.068

Values are mean (±SD) mechanical thresholds (bending force) of neuronal responses obtained using von Frey monofilaments in rats reared in normogravity from birth (NG, NG1wS, NG8wS, NG16wS) and in age-matched rats born and reared in hypergravity during the first 3 postnatal weeks at P22 (HG) and after 1 (HG1wS), 8 (HG8wS), or 16 wk (HG16wS) of housing in normogravity. Mechanical thresholds recorded in NG and HG rats housed for 16 wk in enriched environment (NG16wE and HG16wE) from P22 are also shown. See text for statistics.

Size of cutaneous RFs.

In both HG and NG groups, most of the cortical sites sampled exhibited cutaneous RFs usually covering the glabrous or the hairy skin of one or two forepaw subdivisions (digit phalange, palmar pad). The size of the glabrous RFs normalized relative to the ventral skin area of the forepaw was on average larger in HG rats (5.14 ± 0.48%) than in NG rats (3.58 ± 0.57%; F_1,18 = 43.80; P < 0.001). Therefore, a lower spatial selectivity was found for the neurons within the reduced cortical regions serving the glabrous skin surfaces in the HG rats. The mean sizes of glabrous RFs were similar in HGn⁻ (5.35 ± 1.32%) and HG rats (F_2,27 = 12.31; P = 0.59, ns) but larger than those of NG rats (P < 0.001). By contrast, hairy RFs sizes were similar in all groups (NG: 6.46 ± 2.15%; HG: 5.98 ± 1.03%; HGn⁻: 6.77 ± 1.01%; F_2,27 = 0.70; P = 0.50, ns; Fig. 3, B and C).

Paw print evaluation.

Forepaw print measurements, taken as an estimation of the ventral skin areas in contact with the floor during stance and locomotion, were obtained in HGp and NGp rats of matching body weights (HGp: 44.49 ± 5.95 g; NGp: 43.01 ± 6.18 g; P = 0.59, ns) under both NG and HG conditions. Repeated-measures ANOVA yielded a main effect of gravity condition (F_1,18 = 76.28; P < 0.001), no effect of rearing environment when the NG and HG rats were tested under the same gravity conditions (F_1,18 = 0.085; P = 0.77, ns), and no significant interaction between these two factors (F_1,18 = 4.32, P = 0.06, ns). Mean HGp and NGp rat forepaw prints were similar when obtained under HG testing conditions (HGp: 40.67 ± 9.41 mm²; NGp: 43.52 ± 3.85 mm²; P = 0.06, ns) or under NG testing conditions (HGp: 33.87 ± 5.65 mm²; NGp: 32.47 ± 2.99 mm²; P = 0.34, ns). As expected, rats from either rearing environment exhibited larger forepaw prints when tested in HG than in NG (HGp rats: P < 0.0003; NGp rats: P < 0.001). Interestingly, the ratios of HG to NG print areas recorded in each rat (HGp: 1.19 ± 0.11; NGp: 1.26 ± 0.21) appeared to be similar to the ratio of HG to NG glabrous RF areas measured (in mm²; 1.27 ± 0.21; P = 0.35 and P = 0.33, respectively).

Enduring Effect of HG Rearing

Organization of S1 forepaw maps.

To properly describe possible enduring effects of early exposure to HG, we compared the map data recorded in two groups of age-matched animals (Figs. 5 and 6). In the first group, rats were reared in HG for 3 wk and then housed in standard environment in NG until the end of postnatal week 19 (HG, HG1wS, HG8wS, and HG16wS rats; see materials and methods). In the second group, rats were reared in standard environment in NG from birth until postnatal week 19 (NG, NG1wS, NG8wS, and NG16wS rats). For the total surface area of the forepaw map (Fig. 6A), the ANOVA disclosed main effects of both early rearing environment (F_1,72 = 86.63; P < 0.001) and subsequent NG housing duration conditions (F_3,72 = 12.55; P < 0.001), as well as an interaction between these two factors (F_3,72 = 11.93; P < 0.001). A gradual increase in the total area of the forepaw map was found over the 16 wk examined in the rats reared in NG from birth (NG: 1.43 ± 0.21 mm²; NG16wS: 2.30 ± 0.35 mm²; P < 0.001; Figs. 5 and 6). This increase was found for the cortical territories allocated to hairy skin (NG: 0.18 ± 0.05 mm²; NG16wS: 0.47 ± 0.12 mm²; P < 0.001) and nail (NG: 0.19 ± 0.05 mm²; NG16wS: 0.56 ± 0.18 mm²; P < 0.001) representations (Figs. 5 and 6, B and C), as well as for the glabrous skin representations, although to a lesser extent (NG: 1.02 ± 0.22 mm²; NG16wS: 1.27 ± 0.23 mm²; P < 0.01; Fig. 6D). By contrast, no expansion of the forepaw map was recorded in rats reared in HG for the first 3 postnatal weeks and housed in NG over the following weeks (HG: 1.33 ± 0.14 mm²; HG16wS: 1.39 ± 0.33 mm²; P = 0.84, ns; Figs. 5 and 6). We examined whether this apparent “freezing effect” of the HG rearing affected similarly the different sectors of the forepaw cutaneous map. This HG-induced effect was observed for glabrous skin areas (HG: 0.65 ± 0.15 mm²; HG16wS: 0.49 ± 0.06 mm²; P = 0.21, ns): ANOVA yielded no effect of duration of NG housing (F_3,72 = 0.74; P = 0.53 ns) but did show a main effect of early rearing environment (F_1,72 = 177.70; P < 0.001) and an interaction between these two factors (F_3,72 = 4.69; P < 0.01; Figs. 5 and 6D). As for the hairy skin representation (HG: 0.11 ± 0.04 mm²; HG16wS: 0.19 ± 0.07 mm²; P = 0.26, ns), ANOVA showed main effects of duration of NG housing (F_3,72 = 22.77; P < 0.001) and early rearing environment condition (F_1,72 = 90.35; P < 0.001), as well as an interaction between these two factors (F_3,72 = 7.17; P < 0.001; Figs. 5 and 6B). Similar effects were found for the nail representation (HG: 0.58 ± 0.12 mm²; HG16wS: 0.59 ± 0.09 mm²; P = 0.77, ns), with ANOVA showing main effects of NG housing duration (F_3,72 = 14.49; P < 0.001) and early rearing environment (F_1,72 = 59.93; P < 0.001) and an interaction between these two factors (F_3,72 = 13.98; P < 0.001; Figs. 5 and 6C). Overall, our findings indicate that early exposure to HG prevented the expansion of forepaw cutaneous representational zones normally taking place in NG over the first postnatal months, despite exposure to NG from the end of the third postnatal week.

Fig. 5.

Enduring effect of early HG rearing on cutaneous forepaw maps. Forepaw representations shown were obtained in a rat representative of the group of animals reared in normogravity from birth (NG, NG1wS, NG8wS, and NG16wS groups; left) and an age-matched rat representative of those born and reared in hypergravity during the first 3 postnatal weeks at P22 (HG group) and after 1 wk (HG1wS group), 8 wk (HG8wS group), or 16 wk (HG16wS group) of housing in normogravity (right). Note the “freezing effect” of early HG rearing, which prevented the expansion of forepaw skin sectors, whereas nail expansion was maintained despite exposure to normogravity for 16 wk.

Fig. 6.

Effect of early HG rearing on age-dependent changes in the representational organization of forepaw cutaneous cortical maps. Cortical areas (in mm²; means ± SD) of the total cutaneous maps (A) and different sectors of the maps (B–D) as a function of age (in wk) in rats born and reared in HG until the age of 3 wk are shown in rats housed for 1, 8 and 16 wk in NG (open squares) after the HG rearing and in age-matched rats reared in NG from birth (filled circles). *P < 0.05, statistically significant differences between NG and HG rats at the same time points. Horizontal bars apply to rats reared in NG and indicate statistically significant differences between time points. Note that none of the map representational zones in the HG rats increased as a function of age, in contrast with those recorded in NG rats.

Neuronal responsiveness.

Directional sensitivity of cortical neurons to upward vs. downward nail displacement was analyzed in rats reared in HG for 3 wk and then housed in NG environment until the end of the 19th postnatal week (HG, HG1wS, HG8wS, and HG16wS rats). Repeated-measures ANOVA showed no main effect of duration of NG housing (F_3,36 = 0.67; P = 0.58, ns) but a main effect of directional sensitivity (F_2,72 = 106.88; P < 0.001), as well as an interaction between these two factors (F_6,72 = 16.48; P < 0.001). The HG-induced changes in directional sensitivity of cortical neurons characterized by an increased proportion of U nail displacement responses were still observed when rats reared in HG were housed for 1 wk in NG. HG1wS rats displayed similar proportions of neurons within each category of response compared with HG rats (U: P = 0.68; D: P = 0.92; and U+D: P = 0.45; Table 2). This long-lasting effect of the HG early rearing was shown to dissipate between the second and eighth weeks spent under normal gravity conditions: HG8wS rats displayed a majority of neuronal responses elicited by U+D nail displacement (P < 0.001 when U+D is compared with U or D responses), whereas the rest of the neuronal populations exhibited either U or D unidirectional sensitivity in proportions that are not statistically different (P = 0.12). The proportions recorded in HG8wS rats are similar to those recorded in aged-matched NG8wS rats (U: P = 0.10; D: P = 0.86; and U+D: P = 0.89) and in HG16wS rats (U: P = 0.99; D: P = 0.87; and U+D: P = 0.93, compared with U, D, and U+D proportions found in HG8wS).

In the rats housed in NG from birth and in HG rats exposed to NG from the end of the 3rd week until the 19th week, the mechanical thresholds of neuronal responses assessed with the von Frey filaments were found to be stable until the end of the first month (NG vs. NG1wS: P = 0.22; HG vs. HG1wS: P = 0.96) and to increase thereafter, reaching a maximum at the end of the 11th postnatal week (Table 3). No difference between animals reared in NG or HG was observed at this time (P = 0.12) and at the end of postnatal week 19 (NG16wS vs. HG16wS: P = 0.085, ns). ANOVA showed a significant effect of NG housing duration (F_3,72 = 57.59; P < 0.001) but no effect of early rearing condition (F_1,72 = 3.09; P = 0.08, ns) on neuronal responsiveness, and no interaction between these two factors (F_3,72 = 2.56; P = 0.06, ns). Therefore, these age-dependent profiles of neuronal response thresholds within the glabrous skin cortical sectors were not affected by the early rearing gravity conditions.

Size of cutaneous RFs.

Regarding the size of glabrous skin RFs (Fig. 7A), ANOVA performed between our experimental groups disclosed a significant effect of both NG housing duration (F_3,72 = 17.19; P < 0.001) and early rearing conditions (F_1,72 = 14.17; P < 0.0001), as well as an interaction between these two factors (F_3,72 = 17.32; P < 0.001). Further analysis revealed that the HG-induced increase in relative glabrous RF size was reversed as early as the end of the first week of housing in NG environment (HG: 5.14 ± 0.48%; HG1wS: 3.76 ± 0.30; P < 0.001), thus becoming similar to that recorded in NG (3.58 ± 0.57%; P = 0.79, ns) or NG1ws rats (3.72 ± 0.67%; P = 0.84, ns). The hairy skin RFs, not modified by HG, were similar in all our experimental groups as shown by ANOVA, which yielded no effect of NG housing duration (F_3,72 = 1.78; P = 0.16, ns) or early rearing condition (F_1,72 = 3.40; P = 0.07, ns), as well as no interaction between these two factors (F_3,72 = 0.17; P = 0.92, ns; Fig. 7B).

Fig. 7.

Reversible effect of early HG rearing on the size of glabrous skin RFs. A: examples of glabrous and hairy skin RFs of which the locations on the forepaw were used to reconstruct the cutaneous maps shown in Fig. 5. B: age-dependent changes in the mean sizes of glabrous and hairy skin RFs recorded in the forepaw maps. Note that the HG-induced increase in glabrous RF size observed at P22 was reversed as early as the 1st week of housing in NG. Same conventions as in Fig. 5.

Forepaw prints.

The enduring effect of early HG rearing on forepaw prints was evaluated under NG conditions 1 and 2 wk after NG housing. The weight of HGp rats after 1 wk of NG (67.50 ± 7.82 g) and that of age-matched NGp rats (61.09 ± 4.03 g) were similar (P = 0.08, ns). The same observation applies for HGp rats exposed for 2 wk in NG (97.55 ± 8.86 g), whose weight was similar to that of age-matched NGp rats (93.32 ± 10.19 g; P = 0.25, ns). Repeated-measured ANOVA performed on forepaw prints at 1 and 2 wk after the HG rearing in HGp rats and in age-matched NGp rats disclosed no effect of early rearing environment (F_1,18 = 0.07; P = 0.80, ns) but a main effect of subsequent NG housing duration (F_1,18 = 18.46; P < 0.001), with no interaction between these two factors (F_1,18 = 0.004; P = 0.95, ns). Post hoc tests indicated that forepaw prints increased between the first and second weeks in both groups (HGp: P < 0.02; NGp: P < 0.02) and that they were similar in HGp rats and age-matched NGp animals when tested 1 wk (HGp: 34.30 ± 5.09 mm²; NGp: 34.84 ± 5.93 mm²; P = 0.82, ns) or 2 wk (HGp: 41.40 ± 3.91 mm²; NGp: 41.73 ± 5.95 mm²; P = 0.89, ns) after HG rearing.

Influence of Developmental HG Experience on Adult Use-Dependent Plasticity of Cortical Maps

Organization of S1 forepaw maps.

We sought to determine the influence of early developmental HG experience on later use-dependent plasticity. For this purpose, we used the environmental enrichment (EE) paradigm, which has been shown to strongly modify the S1 forepaw map organization in adult rats (Coq and Xerri 1998). Age-matched rats raised in NG or HG during the first 3 postnatal weeks were subsequently housed in EE for 16 wk. ANOVA performed on the data recorded in the four experimental groups (NG16wS, NG16wE, HG16wS, and HG16wE) yielded a main effect of early rearing (F_1,36 = 163.25; P < 0.001) but no effect of adult housing conditions (F_1,36 = 0.86; P = 0.36, ns) on the overall forepaw cutaneous representation, with an interaction between these two factors (F_1,36 = 9.11; P < 0.01). Further post hoc analysis showed an EE-induced expansion in the rats reared in NG during the first 3 postnatal weeks compared with the rats housed in a standard environment during the same period (NG16wE: 2.66 ± 0.29 mm²; NG16wS: 2.30 ± 0.35 mm²; P < 0.01). By contrast, no such increase of the cutaneous maps was found when rats were exposed to EE after rearing in HG (HG16wS: 1.39 ± 0.33 mm²; HG16wE: 1.20 ± 0.17 mm²; P = 0.15, ns; Fig. 8). ANOVA yielded a main effect of both early rearing and adult housing conditions on the glabrous skin representation (F_1,36 = 362.13; P < 0.001 and F_1,36 = 16.35; P < 0.001, respectively), with an interaction between these two factors (F_1,36 = 11.31; P < 0.01). The post hoc tests showed an expansion of cortical zones representing the glabrous skin only in rats raised during the early postnatal period in NG (NG16wS: 1.27 ± 0.23 mm²; NG16wE: 1.64 ± 0.19 mm²; P < 0.001) but not in HG (HG16wS: 0.49 ± 0.06 mm²; HG16wE: 0.52 ± 0.08 mm²; P = 0.64, ns). ANOVA performed on hairy skin sectors showed an effect of both early rearing (F_1,36 = 80.62; P < 0.001) and adult environment conditions (F_1,36 = 5.92; P < 0.02), with no interaction between these two factors (F_1,36 = 1.47; P = 0.23, ns). Further post hoc analysis indicated that hairy skin areas were not modified by EE in rats reared in HG (HG16wS: 0.19 ± 0.07 mm²; HG16wE: 0.23 ± 0.07 mm²; P = 0.40, ns), whereas they increased in the rats reared in NG (NG16wS: 0.47 ± 0.12 mm²; NG16wE: 0.60 ± 0.17 mm²; P < 0.01) during the first 3-wk postnatal period, although to a lesser extent than glabrous skin sectors. Moreover, ANOVA yielded no effect of early rearing (F_1,36 = 0.25; P = 0.62, ns) but an effect of adult housing conditions (F_1,36 = 10.38; P < 0.01) on the surface area of the nail sectors, with no interaction between these two factors (F_1,36 = 0.02; P = 0.90, ns). The post hoc tests showed no EE-induced changes in nail regions both in rats reared in NG (NG16wS: 0.56 ± 0.18 mm²; NG16wE: 0.43 ± 0.16 mm²; P = 0.09, ns) or in HG (HG16wS: 0.59 ± 0.09 mm²; HG16wE: 0.45 ± 0.07 mm²; P = 0.06, ns) during the first 3-wk postnatal period. Taken together, these findings show that the EE increased skin sectors in rats reared in NG, but not in HG, during the first 3 postnatal weeks. In addition, the expansion of nail representation induced by early HG was not modified by a subsequent exposure to the EE.

Fig. 8.

Influence of early HG rearing on use-dependent plasticity of forepaw cutaneous maps in young adults. Examples are shown of electrophysiological maps obtained in age-matched rats raised in either NG or HG during the first 3 postnatal wk and subsequently housed in an enriched environment (EE) (NG16wE and HG16wE rats, respectively) or a standard environment (NG16wS and HG16wS rats, respectively) for 16 wk. Note that the expansion of glabrous skin sectors induced by the EE in rats reared in NG was not found in rats reared in HG during the first 3 postnatal weeks. Furthermore, the HG-induced expansion of nail representation was not reversed by subsequent EE housing.

Neuronal responsiveness.

Analysis of mechanical thresholds for activating neurons with von Frey filaments assessed in our five experimental groups showed a main effect of the adult environment conditions (F_1,36 = 19.62; P < 0.001) but no effect of the developmental rearing conditions (F_1,36 = 0.30; P = 0.59, ns) on neuronal responsiveness, and an interaction between these two factors (F_1,36 = 5.74; P < 0.02). Post hoc tests indicated no effect of EE on the neuronal response thresholds in rats reared in NG (NG16wS: 0.393 ± 0.049 g; NG16wE: 0.427 ± 0.068 g; P = 0.16, ns), in contrast to the significant increase in rats reared in HG (HG16wS: 0.343 ± 0.025 g; HG16wE: 0.459 ± 0.061 g; P < 0.001).

Size of cutaneous RFs.

As for the RFs recorded within the glabrous skin cortical zones, ANOVA showed a significant effect of the adult housing conditions (F_1,36 = 19.87; P < 0.001) but no effect of the early rearing conditions (F_1,36 = 3.96; P = 0.06, ns) and no interaction between these two factors (F_1,36 = 3.74; P = 0.06, ns; Fig. 9). Further analysis indicates an EE-induced decrease in glabrous RF sizes in rats reared in NG (NG16wS: 3.39 ± 0.31%; NG16wE: 2.48 ± 0.24%; P < 0.001) but not in those reared in HG (HG16wS: 3.40 ± 0.32%; HG16wE: 3.04 ± 0.75%; P = 0.19, ns). Moreover, ANOVA yielded no significant effect of early rearing (F_1,36 = 0.31; P = 0.58, ns) and an effect of adult housing (F_1,36 = 4.51; P < 0.05) conditions on hairy skin RFs, as well as no interaction between these two factors (F_1,36 = 0.72; P = 0.40, ns). Post hoc tests indicated that the hairy skin RFs were not modified by the EE either in rats reared in NG (NG16wS: 7.21 ± 1.01%; NG16wE: 7.93 ± 2.05%; P = 0.37, ns) or in those reared in HG (HG16wS: 6.42 ± 1.78%; HG16wE: 8.10 ± 2.08%; P = 0.17, ns) during the early postnatal period of development.

Fig. 9.

Influence of early HG rearing on use-dependent plasticity of glabrous skin RFs in young adults. Examples are shown of glabrous and hairy skin RFs corresponding to the cutaneous maps shown in Fig. 8. HG during the first 3 postnatal weeks prevented the EE-induced decrease in glabrous RF size found in rats reared in NG.

Taken together, our data indicate that the cortical map changes induced by early postnatal HG rearing impeded the remodeling of forepaw representations induced by EE in young adult animals raised in NG from birth.

DISCUSSION

Summary of Results

In this mapping study of the maturation of somatosensory maps (see Fig. 10), we first provide evidence that the forepaw cortical representations in 2- and 3-wk-old rats are adultlike in their somatotopic organization, but with lower mechanical thresholds than in the adults. A continuous expansion of each of the constituent territories of the map was observed until the 19th postnatal week. Second, we report both short- and long-term effects of early and transient HG exposure on the development of somatotopic forepaw cortical maps in rats subsequently exposed to NG. Despite preservation of the surface area of the forepaw map, HG was found to shape the somatotopic organization of the cortical forepaw map, by expanding the nail representation at the expense of the skin territories in which ventral skin cutaneous RFs were found to be enlarged. In addition, nail responses were biased toward an increased sensitivity to nail extension. The early imprint of HG on the map organization was not modified by a subsequent exposure to NG during the 4-mo period examined. However, after 1 wk of housing in NG, neuronal RFs recorded in HG rats displayed sizes very similar to those recorded in NG rats, whereas directional sensitivity to nail displacement developed within 8 wk. Third, we found that early and transient HG exposure prevented the use-dependent reorganizational changes induced by enriched environment in adult animals reared under NG conditions from birth.

Fig. 10.

Synopsis of findings on the development of the forepaw representation in the S1 cortex and main effects of HG.

Adultlike organization of forepaw maps in 2- or 3-wk-old rats.

This study reports new findings on the development of forepaw somatosensory maps based on high-density electrophysiological maps. The organization and the development of representation in the forepaw barrel subfield (FBS) in the S1 cortex of the rat has been documented in very few studies, in contrast to whisker representation, which has been extensively investigated. Barrel-like structures in the FBS develop around P5 (McCandlish et al. 1989; Rhoades et al. 1990). Experiments carried out to detect the earliest time when the S1 cortex is driven by somatosensory inputs indicate that responses are elicited in the snout subfield as early as 12 h after birth, in the forepaw area by the end of P1, and in the hind paw area by the middle of P2, thus following a lateral-to-medial gradient (McCandlish et al. 1993). However, increases in amplitude and complexity of wave shape and decreases in response latencies were described over subsequent postnatal days, until P14. Seelke et al. (2012), who examined the time course of emergence of the topography of entire body maps, have shown that early in development, most of S1 is occupied by vibrissae/face representations. These authors observed that from P10, representations of body parts, including the forelimbs, emerge with a crude topography, whereas a clear topographical organization began to appear by P15, along with a reduction of RF sizes. From P20, body maps were found to be similar to those recorded in the adult. Our results are in accordance with these data: the maps recorded at P15 in two rats and at P22 in larger samples appeared to be adultlike in their topographical organization (see Coq and Xerri 1998, 1999; Xerri and Zennou-Azogui 2003).

Short-term effects of early exposure to hypergravity on the development of forepaw somatosensory maps.

HG was found to shape the forepaw map by decreasing both the cortical zones serving the glabrous and hairy skin surfaces, whereas that of the nail expanded within cortical sectors normally allocated to skin surfaces. The specificity of the HG effect was evidenced by the fact that this expansion did not occur in rats exposed to HG and subjected to partial deprivation of movement-induced nail inputs resulting from clipping. Interestingly, proprioceptive areas that normally surround the cutaneous maps replaced the lost nail representations that were found to substitute for the skin sectors in the HG rats. In addition, HG clearly modified the overall directional sensitivity to nail displacement of the S1 neurons, as shown by a population response biased toward upward deflection, presumably resulting from counterreaction forces exerted on the nails during upright standing and walking of the HG animals in their home cages.

The alteration of the organization of forepaw somatosensory maps resulting from early HG suggests that this environment strongly modifies the cortical integration of afferent signals from somatosensory receptors. It is worth noting that rat pups are nursed over the first 3 postnatal weeks, with a daily time devoted to lactating behavior that decreases from about 80%, during the first days, to 30% on the 18th day postnatal (Rosselet et al. 2006). The motor system matures slowly over the first postnatal weeks, and the pups do not walk spontaneously before the end of the second week of life. Altman and Sudarshan (1975) reported that, in the rat, walking and running predominate by the end of the second postnatal week. Consistently, clear topographical organization of the forelimb cortical map appears by P15 in S1 (Seelke et al. 2012). These findings suggest that the impact of HG on the forces exerted on the forepaws during the locomotor behavior was predominant during the third postnatal week.

Previous studies have shown that changes in gravity during development influenced motor development in rodents. Bojados et al. (2013) found that exposure to hypergravity before P10, i.e., the acquisition of locomotion, induces postural changes in mice, in particular a more extended ankle joint during locomotion (extension bias). As a mirror situation, rats exposed to microgravity from P14 to P30 exhibited changes in flexor muscle responses (Walton et al. 2005a). In rats born and reared during 100 days in HG, the contractile properties and phenotype of hindlimb extensor muscle fibers were found to be modified (Picquet et al. 2002). Muscle characteristics were reinforced as the soleus, an antigravity muscle, became slower (100% expression of myosin heavy chain slow isoform I), whereas its agonist, the plantaris muscle, presented faster contractile behavior. Collectively, these data sustain the view that joint and muscle proprioception is altered by gravity constraints both during development and in the adult, and that this alteration, which presumably reflects an adaptation to these constraints, is very likely to induce enduring changes in the processing of somatosensory information within S1. However, it is worth mentioning that in our rats born and reared in HG for 3 wk, video monitoring did not reveal impaired behavior, consistent with observations reported by Bouët et al. (2004), who used an experimental device similar to ours.

It is widely recognized that early sensory experience plays a critical instructive role in the structural and functional maturation of the sensory systems (Fox and Wong 2005; Hensch 2004). According to this view, a process of competition for cortical space between sensory inputs shapes and refines the organization of the neural circuits underlying the visual, auditory, and somatosensory cortical maps (Hensch 2004). Patterns of activity elicited by sensory experience are translated into networking of synaptic connections (Katz and Shatz 1996). Our results confirm that the topographically organized somatosensory forepaw maps are sculpted by early sensory experience. They also reveal an experience-driven competition between somatosensory submodalities that critically shapes cortical maps during the early sensory events of postnatal life. We propose that hypergravity, by increasing the pressure exerted on the ventral surfaces of the paws and enhancing the antigravity muscle activity, modifies the pattern of somatosensory afferent signals, thereby leading to the emergence of a distinctive forepaw somatosensory map. The balance between cutaneous submodalities was modified, as evidenced by the contraction of the map territories allocated to skin surfaces that was paralleled by an expansion of cortical sectors serving nail areas. This expansion of nail representation can be attributed to a relative increase in nail stimulation in HG, as also suggested by a change in directional sensitivity reflected by the greater percentage of responses to upward deflection. Interestingly, drastic reduction of tactile input resulting from nail clipping in HGn⁻ rats resulted in a prominent shrinkage of cortical sectors serving the nails that unexpectedly was compensated for, not by an increase in skin surface representations but by an expansion of proprioceptive sectors. This finding suggests that somatosensory submodalities converging on cortical neuronal targets do not have equivalent effectiveness in competing for cortical space. The propensity of proprioceptive representation for “invading” cutaneous territories of the somatosensory cortical maps has been repeatedly documented in our previous studies on adult experience-dependent plasticity following tactile impoverishment or sensorimotor restriction (Coq and Xerri 1999), as well as during aging (Coq and Xerri 2000). An additional argument in favor of competition between cutaneous and proprioceptive inputs comes from the finding that in humans, a rapid increase in cortical proprioceptive activity elicited by stimulation of muscle afferents from the first dorsal interosseus occurs after transient cutaneous deafferentation of the cutaneous territory overlying the corresponding muscle following anesthesia of the radial nerve (Tinazzi et al. 2003). In the present study, the emergence of proprioceptive sectors could be confused with an increase in the threshold of cutaneous responses. Accordingly, a decreased response sensitivity would be expected to occur in the spared cutaneous territories of the HG maps. However, our data show a lack of change in the mechanical thresholds recorded in the shrunken glabrous skin representational zones. The reduced glabrous skin representation observed in both HG and HGn⁻ rats more likely reveals a redistribution of effective inputs within these cortical sectors of the forepaw maps due to the HG environment. The dynamic balance between competing inputs distributed within thalamocortical and intracortical networks is underpinned by the convergence of cutaneous and proprioceptive inputs onto single layer IV neurons of the S1 cortex (Chapin and Lin 1984; Gioanni 1987; Lamour and Jobert 1982; Sievert and Neafsey 1986). This convergence enables an experience-dependent, behaviorally relevant allocation and refinement of somatosensory cortex representational territories.

In this study, the spatial resolution of the contracted forepaw cutaneous maps was found to have deteriorated after early rearing in HG, as indicated by the increase in the sizes of the glabrous RFs, whereas hairy RF sizes were not affected. Previous studies on adult experience-dependent plasticity of somatosensory maps have described RF sharpening or expansion in response to specific task learning or changes in subject-environment interaction (see Xerri 2008 for review). These studies have underscored the influence of the temporal patterning, i.e., the degree of local synchrony of concurrent sensory inputs on the representational segregation of cortical somatosensory maps through receptive field remodeling (Allard et al. 1991; Armstrong-James et al. 1994; Byl et al. 1996; Diamond et al. 1993; Godde et al. 1996; Recanzone et al. 1992; Rosselet et al. 2008; Wang et al. 1995). A consistent finding in all these studies was that temporally synchronized stimulation of large skin territories induced RF enlargement, whereas temporally distributed skin stimulation resulted in RF decreasing. Along the same lines, in an earlier study, we performed severe sensorimotor deprivation by using single-forelimb casting in rats. This procedure led to an over-reliance on the nonconstrained forelimb for postural balance and increased mechanical pressure on the ventral skin surfaces of the uncast forepaw (Coq and Xerri 1999). Although the forced use of the noncast forelimb did not affect the area of the corresponding cutaneous map, it led to enlargement of glabrous skin RFs, probably as a result of concomitant stimulation of larger than usual glabrous skin territories. In the present study, the 2G environment induced broader simultaneous contact of forepaw ventral skin surfaces with the cage floor during stance and locomotion, as confirmed by the increased paw prints. Accordingly, the glabrous skin RF enlargement recorded in HG young rats can be attributed to the spatial aggregation of formerly segregated thalamocortical inputs through unmasking/strengthening of formerly subthreshold synapses. The cutaneous RF enlargement found in HG rats is consistent with the decrease in GABA immunoreactivity recorded in the limb representation of the rats S1 cortex after 14 days of exposure to HG (D'Amelio et al. 1998). Furthermore, an expansion of cutaneous RFs in the S1 cortex was found when GABA-mediated local inhibition was antagonized (Alloway et al. 1989; Chowdhury and Rasmusson 2002; Dykes et al. 1984; Tremere et al. 2001).

Enduring effects of early exposure to hypergravity.

In this experiment, we show that postnatal changes in somatosensory map organization examined after 1, 8, and 16 wk in NG was impacted by neonatal exposure to HG for 3 wk. Gradual increase in the area of the constituent regions of the forepaw maps was found to be typical of rats reared in NG. Interestingly, the HG-induced contraction of skin representation and concomitant increase in nail areas found after HG rearing for 3 wk were maintained over the following 16 wk of housing in NG. Therefore, this early experience impeded normal development of somatosensory maps after exposure to NG, thereby indicating a lack of adaptation of the S1 cortex to new somatosensory experience occurring in NG.

These enduring effects of HG rearing on somatosensory maps could be attributed to an impaired development of posture and locomotion that would tend to maintain changes in the patterns of somatosensory signals conveyed to the S1 cortex. Consistently, previous studies on the effects of altered gravity on the development of motor functions have shown enduring changes in locomotion (Bojados et al. 2013; Walton et al. 1992), swimming behavior (Walton et al. 2005a), and development of surface righting (Walton et al. 2005b) and thus suggested the existence of a critical period in motor development. Persisting extension bias was observed by Bojados et al. (2013) in 2-mo-old mice centrifuged from conception to P10 or P30. In addition, rats born and reared until the age of 3 mo in HG were found to exhibit a lower body position, an enlarged support surface, and an exaggerated foot dorsiflexion (Bouët et al. 2003, 2004). These rats also showed alterations in locomotor pattern when exposed to NG. They exhibited a faster locomotor rhythm (increased number of steps per second; Bouët et al. 2003, 2004). However, all these alterations disappeared within 3 wk, reflecting a gradual adaptation to NG. It is important to underline that in all these studies, the HG-induced alterations to the rodents' posture and locomotion were reported after a longer exposure to HG than in the present study, during a period of adaptation to NG. In our study, we did not observe the postural changes reported by Bouët et al. (2003, 2004), and we found only a transient increase in exploratory behavior. Also of relevance for potential environmental stress, no changes in corticosterone levels were measured in rats exposed to 2G (Moran et al. 2001). Therefore, the available evidence suggests that the long-term cortical changes induced by the 3-wk early exposure to HG were not accounted for by enduring behavioral alterations.

Regarding descending pathways, Gimenez y Ribotta et al. (1998) showed a delayed and modified development of serotonergic projections to the spinal cord of rats exposed to HG from the 11th day after gestation to the 15th day postnatal that appeared to be smaller than in controls. The organization and ultrastucture of these projections were durably affected, even after 8 mo in NG. However, no behavioral correlates were provided in this study, leaving open the question of the functional consequences of alterations in the organization of the serotonergic projections. As previously mentioned, Picquet et al. (2002) showed that soleus antigravity muscle of rats conceived, born, and reared in HG until 100 days expressed 100% slow myosin phenotype. In a subsequent study, the same authors showed that in these animals, NG from 100 days to 115 or 220 days did not transform the muscle phenotype, suggesting the existence of a critical period in muscle phenotype determination (Picquet 2005). One may hypothesize that these persisting changes in descending pathways and muscle properties during subsequent exposure to NG were likely to lastingly influence the cortical integration of proprioceptive inputs, and thus the corresponding representations in the S1 map. By contrast, the paw print measurements of rats exposed to NG after HG rearing were identical to those of rats reared in NG, suggesting that the flows of cutaneous information to the S1 cortex were similar in NG and in HG rats subjected to normogravity. This hypothesis is supported by the normalization of the sizes of ventral skin RFs in HG rats as early as the end of the first week of housing in NG and that of the directional sensitivity of the responses to nail displacement between the second and eighth weeks of normogravity. However, such normalization of neuronal responses within cortical sectors serving cutaneous inputs does not preclude the enduring imprint of early HG environment on the cortical integration of somatosensory afferents. This is suggested by persistent alterations of the forepaw map organization such as the decreased area of ventral skin sectors and the increased nail sectors, as well as the lack of age-related expansion of hairy skin zones of the forepaw maps. Furthermore, it is noteworthy that the HG-induced freezing effects lasted at least 16 wk after exposure to normal gravity. Collectively, these findings sustain the view that the first 3 postnatal weeks represent a time window that is critical for the developmental segregation of somatotopic maps. Importantly, forelimb motor maps were first detectable at the end of the second postnatal week (Young et al. 2012). Moreover, during rat locomotion development, pivoting predominates during the second half of the first week, crawling during most of the second week, and walking or running by the end of the second week (Altman and Sudarshan 1975). Therefore, the behaviorally induced influence of HG on the forepaw map formation reported in the present study was predominant during or even confined to the third postnatal week. The present findings support the view that the maturation of somatosensory maps is strongly influenced by use-dependent changes taking place within a narrow critical postnatal time window.

The long-lasting effects of HG reported herein may reflect enduring changes in neuronal network connectivity during early development. The enduring effects of early HG rearing are at variance with the reversibility of alteration found in adult somatosensory maps following various manipulations (see for reviews Barnes and Finnerty 2010; Buonomano and Merzenich 1998; Xerri 2008), such as experimental syndactyly (Clark et al. 1988), limb immobilization (Coq and Xerri 1999), or natural episodic behavior (nursing; Rosselet et al. 2006). These HG-induced sustained alterations provide an additional argument in favor of a critical period in the maturation of somatosensory maps, previously shown for the development of the whisker-barrel system (see Erzurumlu and Gaspar 2012 for review).

In this article, we report the persistent effect of early experience on the shape and area of somatosensory map territories. By contrast, other response characteristics, namely, the spatial selectivity, i.e., the size of glabrous skin RFs, and the directional sensitivity to nail displacement, were found to display an adaptation to NG. This finding is in accordance with an earlier study in adult rats in which we showed that the spatial selectivity of S1 neurons resulting from ongoing segregation/desegregation of inputs converging onto neuronal targets was reshaped on a timescale closely dependent on behaviorally driven changes in somatosensory inputs (Rosselet et al. 2006). The reversibility vs. retention of different neuronal response properties, after the animals were exposed to NG, shows that the use-dependent synaptic mechanisms leading to masking/unmasking of less/more effective inputs or to synaptogenesis at maturity are expressed within limits of representational malleability stabilized by somatosensory experience during a critical developmental period. This view is substantiated by our findings that the potential for experience-dependent plasticity in the adult was constrained by somatosensory experience during development. In the present study, HG rearing was found to impede the use-dependent map remodeling induced by enriched environment in adult animals reared in normogravity, originally described by Coq and Xerri (1998). It seems that early HG experience shaped the neural network underpinning the forepaw cortical maps in such a way that adult experience-dependent plasticity was limited. Interestingly, the influence of tactile interactions between parents and offspring during early life on intrinsic connections within S1 and callosal connections has been highlighted by a recent study in rodents (Seelke et al. 2016). Therefore, one cannot rule out the possibility that subtle HG-induced changes in nursing behavior that were not detected using video monitoring could have induced alteration in cortical connectivity.

The present study provides evidence that HG is a relevant paradigm to investigate experience-dependent cortical plasticity without sensory deprivation. Our findings advocate the view that the potential for somatotopic map remodeling induced by novel sensory experience in adulthood is constrained by construction of the underlying neuronal networks within a narrow critical period of development.

GRANTS

This work was supported by grants from Ministère de l'Enseignement Supérieur et de la Recherche and Centre National de la Recherche Scientifique.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Y.Z.-A. and C.X. conception and design of research; Y.Z.-A. and C.X. performed experiments; Y.Z.-A., N.C., and C.X. analyzed data; Y.Z.-A. and C.X. interpreted results of experiments; Y.Z.-A., N.C., and C.X. prepared figures; Y.Z.-A. and C.X. drafted manuscript; Y.Z.-A. and C.X. edited and revised manuscript; Y.Z.-A., N.C., and C.X. approved final version of manuscript.