A Sensitive Period for the Development of Motor Function in Rats: A Microgravity Study

Neonatal rats were launched into the microgravity environment of LEO on two separate space shuttle flights: a 9-day flight, the National Institutes of Health Rodent Experiment 3 (NIH-R3, STS-72, Endeavour), and a 16-day flight, Neurolab (STS-90, Columbia). Rats were launched on P7 in the 9-day mission and on P8 in the 16-day mission (due to a 24-h weather delay in the latter mission). The experimental designs for both experiments compare motor skills in the neonatal rats flown in LEO with those in age-matched control animals that remained on Earth during the space shuttle missions.

Time-pregnant Sprague-Dawley dams were obtained from Taconic Farms (Germantown, NY), delivered to Kennedy Space Center (KSC), and individually housed in standard vivarium cages. These animals were placed on the same diet used in the spaceflight: food bars from Teklab Diet (American Institute of Baking). Each dam was verified to be pathogen free. A number of criteria were used to select litters, including water and food bar consumption of the dams, dam weight gain, maternal behavior, adequate weight gain of the pups, and optimal timing of birth in relation to the scheduled launch date. Dams that consumed at least 20 g over the first 24 h of transition to food bars and drank from a Lixit were eligible for selection. Dams were observed every 6 h to determine the birth window. Pups born within the birth window to the dams that had adequate weight gain were eligible for selection. Mixed litters of male and female pups were selected for the 9-day mission. The National Aeronautics and Space Administration (NASA) decided that only female animals would be included in the 16-day mission to eliminate the variable of sex.

Two groups of rats were formed for the 9-day mission: flight rats flown in space and housed in flight cages (FLT, n=10, 4 females and 6 males), and age-matched ground controls (AGC) housed in flight-like cages (n=9, 4 females and 5 males). Four groups of female animals were formed for the 16-day mission: flight rats (FLT, n=6), age-matched ground control animals housed in flight-like cages (AGC, n=8), age-matched ground control animals housed in standard vivarium cages (VIV, n=8), and a group of younger animals (Basal, n=7) that were tested on P8 to assess baseline skills. The basal animals were P8 because they were only included in the 16-day mission experimental design. Although age-matched ground control animals remained at the KSC, the terms “launch” and “landing” are applied to all groups. Animals were P7 or P8 on the day of launch and P16 or P24 on the day of landing. The 16-day mission launched after a 24-h delay during which the animals remained in the orbiter (FLT group) or life science support facility (all ground controls).

Local government authorities, the NASA and the New York University Institutional Animal Care and Use Committee (IACUC) approved all experimental procedures. The animals were cared for and sacrificed according to NIH, NASA, and New York University School of Medicine guidelines. Animals in the Neurolab mission were used for anatomical studies after humane sacrifice (DeFelipe et al., 2002; Raymond et al., 2000).

For each trial, an animal was placed supine on a large (2.5′ by 2.5′) foam pad. The head was held straight or tilted approximately 90 degrees about the roll axis. When an animal was immobile and relaxed, it was quickly released. This procedure was repeated up to 20 times for each head position. A single high-speed camera recorded the animal's movements (NAC, 200 or 500 fps). After the 9-day mission, surface righting was tested daily in the KSC life science support facility from the day of landing (return day [R]0) until 6 days later (R0–R6). After the 16-day mission (Neurolab), surface righting was tested 2, 3, 5, 7, 10, and 23 days after landing at KSC.

The posture of the animal during surface righting and the time required to complete the movement were evaluated. Posture during each turn was assigned to one of three categories (Pellis et al., 1991; Walton et al., 2005c): (1) axial righting (Figure 1A), (2) corkscrew (Figure 1B), and (3) U-posture (Figure 1C). During our analysis, we also observed an extreme form of ventroflexion, where the animals sat upright. This was designated as an “L” posture (Figure 1D), as previously reported (Walton et al., 2005b). For each animal, the occurrence of each posture was expressed as a percentage of the total number of trials for that day. These categorical data are expressed as the group mean of the individual animal means.

Three values for surface righting times were determined using frame-by-frame analysis: when the animal was free to move (Start), when the head rotated 90 degrees (Head), and when both hindlimbs touched the ground to complete the movement (Hindlimbs). Using these data, the times to turn the head (Head-Start) and the hindquarters (Hindlimbs-Head), and to complete righting (Hindlimbs-Start) were calculated. Only trials in which the entire turn could be seen clearly (between 12 and 15 trials per animal per day) were included in the data analysis.

Swimming measurements and analysis followed the same protocol as the older animals on the 9-day and 16-day missions (Walton et al., 2005b). Since they were participating in other experiments, not all of the animals were available for the swimming test every day. Individual animals were videotaped while swimming on R0–R5, R7 (controls), R8 (flight), R9, and R14 after 9 days in LEO or on R1–R3, R7, R10, R16, R20, and R23 after 16 days in LEO. For each swimming trial, an animal was placed carefully at one end of a swimming lane (30″ × 12″) in a tank maintained at 32–35°C. A platform at the other end of the lane allowed the animals to leave the water. Animals were able to complete between 8 and 12 trials during the 2-min period allotted for swimming on each test day. Swimming was videotaped at 60 frames per second (fps). After swimming, the animals were towel-dried and returned to their cages, where they groomed, ate normally, and were quickly accepted by the other animals.

Trials were selected for analysis in which animals could be seen clearly and were not touching the side of the tank. After the 16-day mission (but not the 9-day mission), frame-by-frame analysis was used to evaluate behavior of the animals when they were first placed in the water. Three parameters were measured: (1) posture before beginning to swim, (2) movements used to initiate swimming, and (3) interval between placement in the water and the first swimming stroke. Frame-by-frame analysis was also used to measure swimming stroke duration in the animals from both missions. Each stroke was divided into two phases: (1) power phase, those frames during which the hindlimb moved backward, and (2) return phase, those frames during which the hindlimb moved forward. A sum of the power phase frames and return phase frames gave the number of frames in each stroke. The temporal resolution was limited by the duration of one frame (16.67 ms at 60 fps).

Surface righting, swimming, and pre-swimming behavior were compared between flight and control animals using the grand means for each animal group. Statistical analysis was carried out in the SPSS environment (SPSS Statistics, IBM) using the independent sample t-test (for one independent variable) or one-way analysis of variance (ANOVA) with the least significant difference (LSD) method for post hoc analysis (for more than one independent variable). A level of p<0.05 was considered to be statistically significant. The litter that each animal came from was not known in every case. Thus, as in our other publications (Walton et al., 2005a,bc), we considered that independent observations could be made on these animals, independent of their litter. Values in the text are mean ± s.e.m. Three levels of significance are marked in the figures: * p<0.05, ** p<0.01, and *** p<0.001.

The use of axial surface righting differed between flight and control animals on the first day of testing after landing (Figure 2). In the 9-day mission (Figure 2A), on the day of landing, there was a significant difference between groups in the frequency of axial (mature) righting as determined by one-way ANOVA (F(2, 23)=20.28, p=0.0001). Post-hoc tests revealed that the AGC control group used axial righting more often than launch-age (basal) animals (p=0.0001) or flight animals (p=0.0001). There was no difference between the basal and the flight animals.

After the 16-day mission, on R2, there was a significant difference between groups in the frequency of axial righting (F(3, 24)=8.455, p=0.001) (Figure 2B). Post-hoc tests revealed that the control groups (AGC and VIV) used axial righting significantly more often on the first day of testing compared to basal animals (p=0.007 and p=0.0001, respectively). The control groups also used axial righting more frequently than flight animals, and this reached significance with respect to the vivarium group (p=0.002). There was no difference between the basal and the flight animals, as in the 9-day mission.

Next, the frequency of axial righting was compared in flight and control animals as a function of time after landing (Figure 3). After returning to 1 g, the flight animals from the 9-day mission increased their use of axial righting from R0 to R5 (Figure 3A□. Differences between the two groups of animals were significant on R0–R6. Flight animals reached control levels on R7. Significant differences were found between flight and control animals from the day of landing through day 6 (see Supplementary Table 1 for t values).

A different pattern was seen after the 16-day mission (Figure 3B). The flight animals did not show an increase in use of axial righting over the 23-day testing period after landing. There was no statistically significant difference between days by the flight animals (Figure 3B, ●). There was a statistically significant difference between groups as determined by one-way ANOVA on all test days except on R10 (when a fewer number of animals were available for testing). A post-hoc test revealed a significant difference between flight and AGC control animals on all test days except on R2 and between flight and VIV control animals on all days (see Supplementary Table 2 for statistics). The proportion of axial righting used by control animals did not change over most of the testing period for either mission (Figure 3).

To visualize the immature righting tactics the animals used after landing, the mean occurrence of each righting tactic, expressed as a percentage, was plotted for both missions as a function of age for control animals (Figure 4A & B) and as a function of days after landing for the flight group (Figure 4C & D). In control animals from both missions, axial righting was the predominant tactic and the use of corkscrew righting and righting involving ventroflexion remained stable post-flight (Figure 4A & B).

After the 9-day mission, corkscrew righting was the predominant immature tactic used by the flight animals (Figure 4C, gray). Compared to flight animals, corkscrew righting was used significantly less often by control animals on R1, R2, and R5, and righting with ventroflexion was used less often by control animals on R0, R2, and R5 (Supplementary Table 1).

After the 16-day mission, righting with ventroflexion was the predominant immature tactic used by the flight animals (Figure 4D, black). There was a statistically significant difference for this set of immature tactics on each test day between flight and the AGC and the VIV control groups, as determined by one-way ANOVA (Supplementary Table 2). In contrast to the findings for the 9-day mission, there was no significant difference in the use of corkscrew righting between the flight and either of the control groups.

The increased use of ventroflexion by the flight animals after the 16-day mission (Figure 4) prompted us to ask if head position at the beginning of the righting reflex influenced the tactic used, as we had observed in older rats from this mission (Walton et al., 2005c). In these experiments, the animals were released from the supine position with the head held straight or rotated 90 degrees about the roll axis. Ground control animals used axial, corkscrew, or U-posture righting on R2 when they were released from the head straight position (Table 1). When they were released from the head rotation position, they used axial or corkscrew righting, but not U-posture righting (Table 1). Head position did not have a significant effect on which tactic was used by the control animals; however, tilting the head did lead to a reduction in the use of “U” and “L” positions across groups (Table 1). In contrast, the initial head position had a significant effect on righting tactics in the flight animals (Table 1). Indeed, head rotation significantly increased the use of axial righting, decreased the use of U-posture righting, and eliminated the use of L-posture and U and C righting.

To determine whether the speed of surface righting was influenced by microgravity, righting times were compared for the various tactics used at launch (P8) and on R2 (P26) after the 16-day mission (Table 2). Data from the 9-day mission have been reported previously (Walton, 1998). The study found that both control and flight animals were significantly faster than basal animals for all of these tactics (p<0.001). Differences between flight and control animals did not reach significance, but it is interesting to note that flight animals were faster than the controls.

Pre-swimming behavior was studied after the 16-day mission. When the flight and control animals were first placed in the water, they started to swim, floated before beginning to swim, or tried to get out of the water by climbing up the tank (Figure 5). There were no significant differences among the three groups of animals with respect to the frequencies of any of these behaviors, although flight animals floated less often than either group of controls until R10.

Figure 5

While floating, the flight animals twisted about their cephalocaudal axis. To understand this better, we measured the angle between the plane of the animal's back and the horizontal surface of the water. On R1, the “body angle” while floating was greater in the flight (30.50±5.8°) than in the AGC (7.86±1.844°) or VIV (10.83±3.0°) control animals as determined by one-way ANOVA (F(2, 15) = 11.399, p=0.001). A post-hoc test revealed significant differences between flight and AGC (p=0.0001) and between flight and VIV (p=0.002) animals. From R2 onwards, the flight animals sank less in the water while floating, and their body angles while floating were not significantly different than those of the control animals.

We next measured how long the animals floated before they began to swim. On the first day they were put in the water (R1, P24), the flight animals floated for a significantly shorter period of time (2.24±0.25 s) than the control animals before starting to swim (AGC (4.86±1.14 s); VIV (4.32±0.38 s) as determined by one-way ANOVA (F(2, 17) = 7.390, p=0.004). Post-hoc tests revealed a difference between flight and AGC (p=0.004) and between flight and VIV (p=0.002) animals. However, after R1, there were no significant differences among the three groups of animals.

Stroke duration (measured as the time needed to complete one hindlimb stroke) was used as an estimate of swimming speed after the 9-day (Figure 6A) and 16-day mission (Figure 6B). Over the testing period, the swimming rate of the control animals from either mission did not change significantly. Following the 9-day mission, three flight animals tried to leave the water on R0 and did not swim until R1. Differences between flight and control animals were seen only on day 5 (Figure 6A). On that day, control animals swam faster than flight animals (t(15) = 2.898, p<0.05).

Figure 6

After the 16-day mission (Figure 6B), there was a statistically significant difference in stroke duration among the groups as determined by one-way ANOVA on R2 (F(2,18)=20.308, p=0.0001), R3 (F(2,17)=11.638, p=0.001), R7 (F(2,16)=5.367, p=0.016) and R16 (F(2, 17)=6819, p=0.007). The post-hoc test revealed an effect of spaceflight as well as a cage effect for this parameter. Indeed, on R2 flight animals swam faster (shorter stroke duration) than both VIV (p=0.0001) and AGC (p=0.0001) control animals. On R3, R7, and R16, flight animals swam faster than the vivarium controls (p=0.0001, p=0.005, and p=0.004, respectively). The AGC controls swam faster than the VIV controls on R3 (p=0.003) and R16 (p=0.010), indicating a cage effect. There was no significant difference among the groups after R16.

In order to compare swimming in the animals from both NASA missions, stroke duration was plotted as a function of age in days after landing (Figure 6C). Swimming speed increased from that measured before launch (basal, P8) in flight as well as in control animals. Note that the flight animals from the 16-day mission swam faster (shorter stroke duration) than those from the 9-day mission even though they were close in age (near P25 and P30).

Ankle and knee joint angles were measured for the flight and control animals on the day of landing (R0) and on R1, R7, and R14 following the 9-day mission (Table 3). While swimming, the flight animals tended to move their foot over a wider arc than controls (greater ankle angle range), but this difference was only significant on R7. There were no differences in the knee angle between the two groups.

Differences in the body weight between flight and control animals were seen after the 9-day mission. Because both female and male animals were included in the mission, their weights are plotted separately (Figure 7). On the day of landing, both female and male flight animals were lighter than their control counterparts (female, t(7)=3.911, p=0.028; male, t(8)=5.379, p=0.001). There was no difference in the rate of weight gain over the first month after landing between flight and control animals (Figure 7B). The weight of the animals over three time periods is shown for female (Figure 7C) and male (Figure 7D) animals. The only significant difference in weight between flight and control animals was seen in the females for the 11–15-day post-flight period (t(6)=2.51, p<0.046).

Figure 7

Due to the difference in weights between female and males after the 9-day mission, NASA decided to include only female animals in the 16-day mission. An analysis of their body weights showed significant differences among the three animal groups on the day of landing (R0) (Figure 8A). There was a statistically significant difference between groups as determined by one-way ANOVA (F(2, 19)=66.892, p=0.0001). A post-hoc analysis revealed that flight rats were lighter than controls in flight-like cages (AGC) (p=0.0001) and in vivarium cages (VIV) (p=0.0001). There was also a cage effect. Controls in the AGC cages were lighter than controls in VIV cages (p=0.004).

Figure 8

There was a significant difference among the groups in weight gain over the first month after landing (F(2, 85) = 5.032, p =.009). Post-hoc analysis revealed a larger gain by flight animals compared to vivarium (p=0.006) and AGC (p=0.009) controls (Figure 8B). Nevertheless, there was a continued weight difference between flight and control animals and between the two control groups over the first two weeks after landing (Figure 8C).

The limitations inherent to any spaceflight study may be overcome by using young mammals that are weaned as previously described (Walton et al., 2005a). On the contrary, serious problems are faced when the animals are not weaned (Maese and Ostrach, 2002). In this study, the problems were mostly a result of cage design, which did not provide enough gripping surface. The most important consequence of this was an inability of the dams to adequately nurse the pups. Inadequate nursing led to high mortality among the pups, particularly for those in the Research Animal Holding Facility (RAHF) cages used for the Neurolab mission, where the mortality rate was 50% compared to only 10% for the R3 mission, where bisected Animal Enclosure Modules (AEM) were used (Maese and Ostrach, 2002). Other consequences of inadequate housing and poor maternal–offspring interaction were poor thermoregulation, reduced olfactory and tactile stimulation, and inadequate anogenital stimulation of the pups. The younger animals from the Neurolab mission were available for study only because the crew were able to keep some pups alive by directly hand-feeding, hydrating, and administering antibiotics (Maese and Ostrach, 2002). These findings highlight the importance of in-flight experimental studies and careful hardware design.

This study reports data from young animals that spent either 9 or 16 days in microgravity after developing in a 1 g environment during the first week of life. To understand the effect of microgravity on motor development, these results must be compared to those of older animals on the same NASA missions, which had developed for twice as long in a 1 g environment before entering microgravity (Walton et al., 2005b; Walton et al., 2005c).

Overall, these findings show that postnatal motor system development was appropriate to the animal's environment, in spite of the difficult conditions encountered by the animals on the 16-day flight. Indeed, in the microgravity environment, there was no stimulus for the surface-righting reflex and the data suggest that the normal development of this reflex stopped. This was reported previously in older animals that spent either 9 or 16 days in microgravity (Walton et al., 2005c). Here, the animals entered microgravity at P7 or P8. Therefore, in all four cases, when they returned to the 1 g environment, the flight animals used immature tactics more often than age-matched ground controls. The effect of microgravity on vestibular development reflexes was also seen in swimming as the flight animals in all four groups tried to float when they were immersed in water. After the 16-day mission, swimming studies also showed that flight animals had a steeper body angle than the control animals. Some effects were transient, but others lasted for as long as the animals were studied, providing insight into sensitive and critical periods of development.

The animals in this study depended on nursing for their nourishment. This proved to be challenging in the microgravity environment due to housing. This was seen in the difference in body weight between flight and control animals on the day of landing after both the 9-day and 16-day mission.

Although it is likely that the flight animals were malnourished, the question of interest is the contribution of this factor to the differences in motor behavior seen between flight and control animals. Indeed, malnutrition can have a significant effect on postnatal motor function (Horiquini Barbosa et al., 2016). Nevertheless, the differences in the behavioral measures used in this study are consistent with having spent time in LEO, rather than with the flight animals being malnourished. The most direct argument in support of this view is that the differences seen in the young flight animals are similar to those seen in the older flight animals that had the same weight as controls at landing (Walton et al., 2005b; Walton et al., 2005c). Undernutrition is not necessarily associated with changes in cortical development (Fernandez et al., 1993), and although it can reduce basal dendritic length and branching, environmental stimulation can compensate for this impairment (Pascual et al., 1996).

Another possible confounding factor is that the younger animals experienced more stress than the older ones due to differences in cages, noise, lighting conditions, and experience in microgravity. In a related study, neonatal rats subjected to maternal separation stress from P2 to P21 showed some delay in the acquisition of air righting, surface righting, and postural righting reflexes (Mesquita et al., 2007). However, these observed effects were transient, and surface righting was restored by P3 in males and by P5 in females, even when the stress continued. In contrast, our results suggest that lack of gravitational input during a critical period of development beginning on P7 or P8 can produce long-term effects on righting tactics. Therefore, we contend that the effects seen cannot simply be due to stress.

Space-based biological studies have limitations that affect interpretation of the findings. The most important of these in conducting behavioral studies is perhaps the limitation of the number of animals that may be studied. Limitations in data collection are due to both the constrained space available on the space shuttle and the post-flight availability of the animals. Nevertheless, these findings provide a first look at the relative importance of age and mission length on some aspects of postnatal motor development.

Although these findings must be considered preliminary, they are consistent with the idea that postnatal nervous and muscular system development are adaptations to the environment, and that spaceflight offers a unique environment for further study.