A Versatile, Behavioral Method to Investigate Thyroid Hormone Effects on Cerebellar Function

Summary

Here we present a protocol for a versatile behavior test developed recently, the ladder beam test. This test has the advantage of detecting subtle cerebellar ataxia caused by a defect of thyroid hormone action in the central nervous system over the conventional behavior tests assessing motor performance.

Abstract

Thyroid hormone (TH) action is essential during the development of the central nervous system, including the cerebellum. In case of TH deficiency in early life such as congenital hypothyroidism, patients display neurological disorders such as cognitive retardation and motor deficits. There are various studies using mouse models with tissue- or cell-specific TH deficiency to investigate the role of TH in the cerebellum. Compared to generalized congenital hypothyroid mice, cerebellar cell-specific TH-deficient mice display milder and subtler ataxic features, making the assessment of motor function difficult when using conventional tests such as the rotarod test.

Due to the need for an alternative tool to assess motor function in TH-related animal models, we developed a versatile behavioral method called the "ladder beam test," in which we can design the various ladder tests depending on the severity of ataxia in model mice. We utilized transgenic mice expressing a dominant-negative TH receptor specifically in the cerebellar Purkinje cell, a sole output neuron in the cerebellar cortex modulating motor performance. The newly-built ladder beam test successfully detected robust impairments in motor performance in the transgenic mice at a greater level compared to the rotarod test. Disruption of motor learning was also detected in the ladder beam test but not in the rotarod test. The protocol with this novel behavioral apparatus can be applied to other animal models that may show mild ataxic phenotype to examine subtle changes in cerebellar function.

Introduction

Thyroid hormone (TH) is indispensable for brain development¹. In particular, its role in the cerebellum is critical because TH deficiency in early life causes aberrant cerebellar development^1,2. For example in congenital hypothyroidism, patients display a series of neurological retardation including cognitive and motor deficits³. To unveil the role of TH in cerebellar functional development, some studies have limited TH deficiency in a cerebellar cell-specific manner⁴. However, compared to generalized congenital hypothyroid mice, in which all tissues and cells are affected by TH deficiency, such cerebellar-specific models display so subtle ataxia that the conventional behavior tests, such as rotarod, footprint, and balance beam tests, barely detect the differences. Thus, to fully investigate the TH effects on cerebellar function, a new assessment tool is needed to detect a subtle change in the motor coordination of model mice.

The rotarod test is the most common tool for assessing motor coordination, originally developed by Dunham and Miya⁵ and later applied to an accelerating version by Jones and Roberts⁶. The latency to fall from the rotating rod is interpreted as the test for motor coordination, and its simplicity and conciseness make it commonly used among behavioral researchers studying motor function⁷. However, the ease of use of this test is a double-edged sword. Because the rod automatically rotates, mice can cling to and stay on the rotating rod without moving. Furthermore, mice may intend to fall off rather than keep balancing on the rotating rod. In either situation, the validity and reliability of the test are questionable for assessing "pure motor coordination"⁷. In other words, it does not accurately target the cerebellar function and involves other factors such as muscle strength for gripping.

Instead of the conventional tools for motor coordination assessment, here we present a novel behavior test called the "ladder beam test," which is recently developed in our laboratory. The horizontal ladder walking test was designed to assess cerebellum-related complex motor abilities: feed-forward prediction and integration of motion⁸. The test device was composed of four pieces of plexiglass with holes (Figure 1). The four plates were connected in parallel by screws and sticks inserted into the holes on the plates. Two outer plates were used to stabilize the device and two inner plates were used to design the various kinds of ladder rungs (Figure 2C). The width of the rung was adjusted depending on the animal size to minimize the animals' moving backward (Figure 2B). The distance from the start point to the goal was 110 cm. The device was located 60 cm above the bench and a safety cushion was set under the device (Figure 2A). The dark chamber was put near the goal to motivate the animals to move toward the goal (Figure 2A).

We examined the TH effects on cerebellar functional development by using transgenic mice expressing dominant-negative TH receptor (TR) in cerebellar Purkinje cells (Mf-1/FVB mice). In both rotarod and ladder beam tests, we observed cerebellar ataxic phenotype in Mf-1/FVB mice; however, the ladder beam test succeeded in detecting more significant differences than the the rotarod test (Figure 3). In addition, motor learning ability can be more thoroughly assessed in ladder beam test (Figure 3B,C). As a cellular background of such a behavioral phenotype, the induction of long-term depression (LTD) was inhibited and instead, long-term potentiation (LTP) was induced following an LTD-inductive stimulation in Mf-1/FVB Purkinje cells⁹. LTD is essential for motor coordination and motor learning in the cerebellum¹⁰. Many studies have reported motor deficits and inhibition of LTD in knockout or mutated mice for key regulator genes in cerebellar function, however, no studies have ever reported the induction of LTP following an LTD-inductive stimulation^11,12. Taken together, this phenomenon may be unique to Mf-1/FVB mice or TH-deficient mice (the same phenomenon was observed in adult-onset hypothyroid mice), suggesting that TH regulates cerebellar function differently from the other key proteins. If so, it is plausible that mice with abnormal TH action do not display cerebellar ataxia in the same way as other model mice. This again emphasizes the need for a specific method for assessing TH effects on cerebellar function. This paper presents a novel protocol to investigate TH effects on cerebellar function using the newly-built ladder beam test.

Protocol

The animal experimentation protocol in the present study was approved by the Animal Care and Experimentation Committee of Gunma University. All procedures for the care and treatment of animals were performed according to the Japanese Act on the Welfare and Management of Animals and the Guidelines for the Proper Conduct of Animal Experiments issued by the Science Council of Japan. The assembly drawing of the apparatus can be found in Figure 1.

1. Ladder setup

1. Horizontal ladder test (Figure 2C left)
   1. Insert aluminum sticks (2 mm diameter) into holes on the plexiglass walls.
      NOTE: The interval distance varies according to body size depending on the strain or sex of the tested mouse (see Table 1).
   2. Adjust the width of the ladder rungs by widening or narrowing the distance of the plexiglass walls.
      NOTE: The width may vary according to the body size depending on the strain or sex of the tested mouse (see Table 1).
2. Zigzag ladder test (Figure 2C right)
   1. Insert aluminum sticks (2 mm diameter) into every third hole in the horizontal and vertical directions on the plexiglass walls.
      NOTE: The interval distance varies by body size depending on the strain or sex of the tested mouse (see Table 1).
   2. Adjust the width of the ladder rungs by widening or narrowing the distance of the plexiglass walls.
      ​NOTE: The width may vary depending on the body size, strain, or sex of the tested mouse (see Table 1).
   3. Place the dark chamber in the target site.
   4. Place the cushion (can be made of a plastic bag stuffed with buffer materials) 30 cm below the ladder setup for safety.

2. Habituation (Day 0)

NOTE: It is recommended to take at least 1 day for habituating the mice to the ladder setup.

1. Place the mouse at the edge of the ladder (starting point).
2. Using an air puff, encourage the mouse to move forward on the ladder rung.
3. Repeat until the mouse gets habituated to walk on the ladder rungs.

3. Measurement of motor performance

1. Place the mouse at the edge of the ladder (starting point).
2. Record the time to complete crossing the ladder and the number of foot slips when the mice fail to place their paws or slip after placement on the aluminum sticks (see Figure 2D). If possible, record the video to check the missteps afterward.
3. After a 15 min interval, repeat the measurement (4-5 trials per day).
4. Repeat the experiment for 3-4 days for assessing motor learning.

4. Analysis

1. Calculate the average value of time and foot slips per day.
2. Compare the average value between the groups.

Representative Results

In the rotarod test, Mf-1/FVB mice showed a significant decrease in the latency to fall from the rotating rod over 3 consecutive days compared to the wildtype mice, indicating impaired motor coordination (Figure 3A). However, in terms of within-group results, Mf-1/FVB mice significantly improved their performance from Day 1 to Day 3, suggesting the preservation of motor learning (Figure 3A).

Both horizontal and zigzag ladder beam tests detected robust impairment in motor coordination in Mf-1/FVB mice as they spent significantly longer time to complete crossing the ladder and made more foot slips compared to wildtype mice (Figure 3B,C). More importantly, the score did not improve for 3 days in Mf-1/FVB mice, whereas wildtype mice showed improvement (Figure 3B,C). Unlike the rotarod test, Mf-1/FVB mice failed to exhibit any motor learning in the ladder beam test. These results indicate that the ladder beam test is superior to the conventional rotarod test for assessing not only motor coordination but also motor learning. There was no change in grip strength between wildtype and Mf-1/FVB mice (Figure 3D).

Figure 1: Assembly drawing of ladder beam test apparatus. 1. Cut out the plexiglass (110 cm x 20 cm, 110 cm x 10 cm). Drill a 2-mm hole (5 in vertical, 100 in horizontal, 1 cm intervals). 2. Insert thick metal sticks into holes at four corners to connect and stabilize the plexiglasses. 3. Inert aluminum sticks into holes to set up the ladder test. Stick adhesive rubber to the inserted part for stabilization. Please click here to view a larger version of this figure.

Figure 2: Experimental apparatus for the ladder beam test. (A) A side-view image of a custom-made apparatus for the ladder beam test. The safety cushion for mice was made of a cardboard box covered with a shock-absorbing plastic bag. (B) Back and top views of the ladder apparatus are shown with an adult-sized mouse. (C) Aluminum sticks can be inserted into holes on the wall to design the ladder test. Horizontal and zigzag ladder tests are shown as examples. (D) An image describes what the foot slip looks like on the ladder rung. Please click here to view a larger version of this figure.

Figure 3. Comparison between the rotarod and the ladder beam tests in assessing motor coordination and motor learning using Mf-1/FVB mice. (A) The results of the rotarod test for 3 consecutive days are shown for each group (wildtype, open circles; Mf-1/FVB mice, filled triangles). Mf-1/FVB mice stayed on the rotating rod for less time than the wildtype mice (**p < 0.01), although both groups improved their performance on Day 3 compared to Day 1 (# p < 0.001). (B) Horizontal ladder beam test. (Upper) Mf-1/FVB mice took much longer times to complete the task than wildtype mice (****p < 0.0001) and no improvement in the performance at Day 3 compared to Day 1, in contrast to wildtype mice (# p < 0.01). (lower) Mf-1/FVB mice made more foot slips than wildtype mice (****p < 0.0001). Only wildtype mice improved performance (# p < 0.01). (C) Zigzag ladder test. Mf-1/FVB mice spent a longer time crossing the ladder than the wildtype mice (***p < 0.001). (Upper) Mf-1/FVB mice spent a longer time completing the task than the wildtype mice (****p < 0.0001). Only wildtype mice improved performance from Day 1 to 3 (#p < 0.01). (D) No change was observed in grip strength between groups (p = 0.8998). Data are expressed as mean ± SEM. Please click here to view a larger version of this figure.

	C57bl/6J		FVB/NJcl
Age	3-6 weeks old	7 weeks old ~	3-6 weeks old	7 weeks old~
Male	Horizontal: 1-2 cm	Horizontal: 3-4 cm	Horizontal: 2cm	Horizontal: 4-5 cm
	Zigzag: 2 cm	Zigzag: 4 cm	Zigzag: 3 cm	Zigzag: 4 cm
	Width: 3 cm	Width: 4 cm	Width: 4 cm	Width: 5 cm
Female	Horizontal: 1-2 cm	Horizontal: 3-4 cm	Horizontal: 2 cm	Horizontal: 3-4 cm
	Zigzag: 2 cm	Zigzag: 3 cm	Zigzag: 3 cm	Zigzag: 3-4 cm
	Width: 3 cm	Width: 3-4 cm	Width: 4 cm	Width: 4-5 cm

Table 1: Recommended intervals and widths for the experiment. The distance of intervals and the width may vary depending on the strain, age, and sex of the tested mouse. The example of C57bl/6J mice and FVB/NJcl mice are shown.

Discussion

Our design drew on the past study by Metz and Whishaw, who reported the utility of "the ladder-rung walking test"^13,14. They designed the ladder-rung walking test to assess skilled walking and measure both forelimb and hind limb placing, stepping, and inter-limb coordination by requiring animals to walk from a starting point to a goal on a horizontal ladder^13,14. What makes our ladder walking test different from theirs is the versatility; the experimenter can change the ladder design and thus the complexity of the test by changing the location of the aluminum sticks on the walls with 5 holes in the vertical direction and 100 holes in the horizontal direction. "The zigzag ladder test" (Figure 2C right and Figure 3C) allows the experimenter to examine more complex movements that cannot be evaluated by the rotarod test. Such complex movements require the coordination of limb movements and likely reflect the cerebellar function¹⁵. The concern about neuromuscular strength affecting the performance in the zigzag ladder test can be excluded by performing the grip strength test (Figure 3D).

A more advanced tool for assessing cerebellar-specific function is presented by Cupid and colleagues¹⁶. The similarity to our model is that their ladder consists of high and low rungs, making it zigzag in the vertical direction. However, mice are required to walk on the higher rungs, which is equivalent to the "horizontal ladder test" (Figure 2C right) and stepping on the low rungs is considered a "misstep." Moreover, the ladder built by Cupid et al. can assess motor learning because a rung raised above the stepping surface ("an unexpected object," cued by a sound stimulus) makes mice learn to deal with it. The analysis software is built into the apparatus, allowing the automatic detection and recording of time, foot slips, and even back steps and jumps¹⁷. Although the model of Cupid and colleagues is such a handy, useful, and accurate tool for motor function, the apparatus is costly and thus, not always available to every laboratory. In contrast, our ladder beam test apparatus can be easily hand-made (requiring only three steps) and can be modified when building without considerable expense. The cost-effectiveness is another advantage of our ladder beam test.

Before beginning the test, it is important to determine the proper spacing of the ladder rungs depending on the body size, sex, and age of the mice. If the space is too short, there is no difference from the normal walking test or a gait pattern test, for example, the footprint test. If the space is too long, it is difficult for the mice to reach the next rung and walk on the ladder rungs. In either situation, the test would be unable to assess motor coordination and thus, cerebellar function. The ideal space is likely the distance that mice can continuously walk without any unnecessary efforts made by parts of the body other than the cerebellum. Moreover, the width of the ladder rungs should be carefully chosen. If it is too wide compared to the body size of the mice, mice can easily turn around and come back to the starting point. It is preferred to adjust the plexiglass walls on both sides so that they get close to mice but never touch them and disrupt their walking (Figure 2B). Adjusting the parts of the apparatus for each mouse is a critical step in establishing the validity of the test.

Generally, mice prefer dark and narrow conditions. Therefore, the dark chamber located at the goal site should motivate mice to complete the task. In case mice refuse to move forward or suddenly stop in the middle of the ladder, applying air pressure to mice from the back helps them start moving. However, too much air pressure may make mice stressed out or nervous, which may affect the test result. Habituating mice on the novel ladder rungs should be conducted well before starting the experiment. It usually takes 1 day to get mice habituated to a novel condition (being on the elevated ladder rungs). The training session may be cut off before the mice can improve their performance to prevent the training effect. We recommend separating the habituation day and test day to avoid the effect of fatigue in mice after the training session.

Our ladder beam test certainly has a few limitations. First, it is time-consuming to conduct the experiment. "Day 0" should be prepared for the habituation phase. In general, it takes ~15-30 min to get mice habituated and walk on the elevated ladder. If the experimenter has a large group, it may take an entire day to complete the habituation. Even after the habituation, the treated mice, especially those with motor deficits, may take a long time to complete crossing the ladder.

While conducting the test, the experimenters need to watch the mice to track the time and the number of foot slips. Alternatively, video recording is recommended for post experiment analysis. Recorded videos can also be analyzed using "the foot fault scoring system of the ladder rung walking test"¹⁸. Second, the throughput performance of the test is quite low. As mentioned earlier, the test requires a lot of time to conduct. In addition, only one mouse can be used per trial. In contrast, the rotarod test has higher throughput performance as the apparatus has 4-5 lanes and allows the experimenter to conduct the test with a group of mice at once. With the automatic tracking system such as the apparatus built by Mendes et al.¹⁵, it allows the experimenter to make several parallel lanes of a ladder and conduct the experiment with more than one sample.

Nevertheless, the present study emphasizes the superior part of the ladder beam test to the rotarod test using the transgenic mouse model expressing a dominant-negative TR specifically in the cerebellar Purkinje cells. Previous studies have indicated that congenital hypothyroid mice or transgenic mice with TH deficiency display impairment in motor coordination by the rotarod test^19,20. However, since they conducted the rotarod test only for 1 day, the TH effects on motor learning have been neglected. Motor learning is also important in the cerebellar function, requiring the proper synaptic plasticity in Purkinje cells¹⁰. The ladder beam test succeeded in detecting the impairment in not only motor coordination but also motor learning in Mf-1/FVB mice, whereas the rotarod test did not (Figure 3).

TH action in the cerebellum has yet to be elucidated. For future studies, the test is expected to be used to assess the cerebellar function of the various animal models of hypothyroidism or TH diseases. Furthermore, this test can be applied to other animal models that may show mild ataxic phenotype to examine subtle changes in cerebellar function.

Materials

Name	Company	Catalog Number	Comments
Air puff	DAISO
Aluminum sticks	CAINZ		2 mm diameter, number of sticks may vary depending on the ladder design. Aproximately 30 sticks may be required to build the horizontal ladder (4 cm interval).
Blutack	Bostik
Plexiglass	CAINZ		110 cm x 20 cm, 110 cm x 10 cm, 2 parts each
Screws	CAINZ