Mutation of the attractin gene impairs working memory in rats

Abstract Objective Attractin (ATRN) is a widely expressed member of the cell adhesion and guidance protein family in humans that is closely related to cellular immunity and neurodevelopment. However, while previous studies in our laboratory have confirmed the effect of ATRN mutations on long‐term memory, its specific role and the molecular mechanism by which it influences spatial cognition are poorly understood. Methods This study aimed to examine the effect of ATRN mutations on working memory in water maze with a novel ATRN‐mutant rat generated by the CRISPR/Cas9 system; the mutation involved the substitution of the 505th amino acid, glycine (G), with cysteine (C), namely, a mutation from GGC to TGC. The changes in myelin basic protein (MBP) expression in rats were also analyzed with the western blot. Results The ATRN‐G505C(KI/KI) rats exhibited significant increases in the required latency and distance traveled to locate the escape platform in a Morris water maze test of working memory. In addition, the expression of MBP was reduced in ATRN‐mutant rats, as shown in the western blot analysis. Conclusion Our results indicate that ATRN gene mutations may directly lead to the impairment of working memory in the water maze; this impairment may be due to the inhibition of MBP expression, which in turn affects the spatial cognition.

Previous studies have summarized working memory as an interactive mode of behavior and brain function that temporarily holds and manipulates information during a series of cognitive tasks, such as language comprehension, learning, and reasoning (Kent, 2016;Miller et al., 2018). Thus, working memory can be considered the basis and prerequisite for the successful execution of complex behaviors. Further investigation of the effects of ATRN mutations on working memory could provide a better understanding of the role of ATRN in spatial memory.
To examine the relationship between ATRN and working memory, a novel ATRN-mutant rat generated via the CRISPR/Cas9 system was used to investigate the effect of ATRN gene mutations on working memory in rats. This study also observed changes in myelin basic protein (MBP) expression in the rats. Therefore, our findings suggest corresponding behavioral changes associated with ATRN mutations in rats. The novel ATRN-G505 C mutant rat can serve as an ideal model for further study of the multiple functions of ATRN.

Animals
The rats used in this study were all male and were mated with a generation of heterozygous ATRN rats provided by Nanjing University. Genotypes were identified 2 weeks after birth, and behavioral tests were performed at 60 days. Briefly, transcript 201 of rat ATRN (attractin, Gene ID: 83526) was selected to create the ATRN mutants, and a G505C mutation was introduced into exon 9. Specifically, the 505th amino acid (aa) glycine (G) was altered to cysteine (C); that is, GGC was altered to TGC . A synthetic primer was inserted into introns 8-9 for subsequent genotyping. The animal muta- (CAGGACTGTGCACATGAATAG) and 3444-ATRN-wt-tR2 (GAGAT-ACAGAGAGACTAGTGC). All animal procedures were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publications No. 8023, revised 1978) and were approved by the Animal Care and Use Committee of Jiangsu University. All animals were housed under a reversed light-dark cycle (lights on 20:00-8:00) in a room with controlled temperature (21 ± 1 • C) and humidity (55 ± 5%) and were allowed free access to food and water, as described in previous studies from our laboratory.

Water maze working memory test
The maze was a circular pool 160 cm in diameter and 75 cm in height that contained water at 20 ± 5 • C. The interior of the swimming pool was painted black, and there was a removable escape platform 10 cm in diameter placed 2 cm below the surface of the water, not visible to a rat in the pool, which provided a base for the rats to stand on (and thereby cease swimming). The working memory water maze task for successive trials were different (see Figure 1a for more details).
Rats were given four trials per day. At the beginning of the test, the rats were placed into the water at the starting position. The trial ended when the rat reached the escape platform, and the escape latency was recorded for each trial. If a rat failed to escape within 60 s, the maximum escape latency was recorded as 60 s, and it was guided to the platform by the experimenter. The rat was then allowed to stay on the escape platform for 15 s before being returned to its home cage for 20 s before the start of the next trial, resulting in an intertrial interval of 35 s. After the trials for the day were completed, the rats were dried with a towel and returned to their home cage. Three independent measures, (1) swimming distance (cm), (2) swimming velocity (cm/s), and (3) escape latency (s), were tracked and digitized, and information was stored for subsequent analysis (Pouzet et al., 2002). At the end of the water maze experiment, the brains of the rats were removed within 2 h of the final trial and stored at 4 • C for later use.

Western blot analysis
Western blotting and semiquantitative analyses were performed in accordance with previously described procedures . In  Figure 2A(h)). Moreover, the trials × genotype interaction was not significant on each day (all p s > .05). Taken together, these findings suggest that the different escape latencies of the three groups were most likely related to the differences in genotype.
We further compared the swimming trajectories of rats in each group with different platform positions. We found that ATRN-G505C(KI/WT) and wild-type rats dynamically altered their swimming paths according to the platform locations to rapidly reach the platform; however, the search strategy of the ATRN-G505C(KI/KI) rats appeared to be static over the 9-day experiment ( Figure 2B). When these rats were placed in the water, they first sought the platform along the edge of the pool and only then ventured into the middle region; exploration of the edge of the pool often consumed a substantial amount of time.
Therefore, these results suggest that ATRN-G505C(KI/KI) rats took longer to find the platform when it was not located at the edge because of a failure to update learning strategies or working memory and not because of spatial memory impairments. Thus, ATRN gene mutations could cause working memory impairments.

Effect of the ATRN mutation on MBP expression in the brain
The effect of the ATRN mutation on CNS homeostasis was examined by comparing the expression of MBP in the hippocampus and cerebellum among the three groups of rats. We found that MBP lev- but the expression did not significantly differ between these two groups (p > .05). This confirmed that ATRN mutations suppress MBP expression in the brain and may thereby affect CNS function.

DISCUSSION
In this study, the ATRN homozygous mutant rats performed worse than the ATRN heterozygous mutant rats: They did not seem to develop correct spatial perception despite repeated training in the water maze working memory test, although the heterozygous rats formed somewhat reliable memories of the platform location during this time.
Although heterozygous rats did not perform as well as the wild-type rats, we concluded that the working memory impairment due to heterozygous mutation in ATRN was not as severe as that caused by homozygous mutations; this could indicate that partial mutations in ATRN do not completely disrupt spatial memory or that other compensatory mechanisms are employed. Additionally, the reduced expression of MBP in ATRN homozygous rats suggests that the stability of myelin structure and function in the CNS is disrupted, which may trigger the disruption of spatial perception and memory seen in this mutation.

F I G U R E 2
Performance on the working memory test on each day of training (9 days in total) and the representative swimming paths for the three groups on Days 3, 6, and 9. The ATRN gene is related to pigmentation, central nervous system, and immune regulation. Therefore, the mutation of ATRN gene will affect the homeostasis of central nervous system. Zitter rats developed tremor at 3 weeks of age and flaconic hind limb paralysis at about 6 months of age. The main pathological manifestations were progressive myelotopia and vacuolization of the central nervous system (Kuramoto et al., 2001). The initiation of myelination and the basic structure of the myelin sheath are normal, but over time, the density of myelin fibers and layers is abnormally low, as well as abnormal or elongated myelination. The splice site at the 12th exon-intron junction of the ATRN gene was mutated to be homozygous. Clinical and neuropathological phenotypes were saved by expressing the WT membrane form of ATRN instead of the soluble form. Another ATRN spontaneous rodent mutant, the black tremor hamster, is characterized by tremors in the trunk and back legs secondary to the abnormally thin myelin sheath of all central axons (Nunoya et al., 1985). When 10 kb is inserted into ATRN, the molecule is homozygous, resulting in complete loss of the membrane isoform (Kuramoto et al., 2002). Finally, several mouse ATRN mutants were characterized by fur color changes and neurodegenerative changes Gunn et al., 2001). ATRN mg-6J mice with the deletion of most of the ATRN gene (exon 1-27) genome had mahogany-colored coat, extended gait, and tremors. Dyskinesia can be severe, affecting the extent of feeding, and some mice have died by 3-4 weeks of age. Seizures observed in some ATRN mg-3J homozygous mice are characterized by sudden freezing of motion followed by rollover. Beginning at 2 months, the mutant developed severe vacuolation of the brain, brainstem, cerebellar granular layer, and spinal cord, which severely affected myelination in the central nervous system.
In this study, experimental manipulation and individual animal differences, as well as the test environment, will affect the experimental results. Therefore, during the experiment, blind detection was used to minimize the effects of experimenter factors, that is, the analysis participants were different from the actual participants, who did not know the specific group of rats. Of course, in order to avoid large individual differences between rats, it is necessary to ensure the unity of the breeding environment, including the fact that the breeder and the actual experimental operator must be the same person, to prevent the distortion of the experimental results caused by the stress of the rats on the experimental personnel during the experiment. In terms of experimental environment, in addition to ensuring appropriate water temperature and light intensity, researchers should also try to avoid interference to rats. Therefore, during the experimental process, the researchers should control the time whether they put the rats in the water or guide the rats on the platform.
Due to spatial and temporal resolution limitations in previous studies, only estimates of the highest density of the most active neural elements (cells and fibers) in brain regions involved in spatial memory representations, whether sensory, motor, emotional, or associative, could be calculated. However, these methods are clearly insufficient for characterization of the fine-grained details and distribution of neural elements involved in specific memories. Recent studies have revealed the complexity of the structural connectivity between brain areas underlying these memories (Brincat et al., 2018;Reinert et al., 2021 In addition, reverse transcription-polymerase chain reaction (RT-PCR) analysis revealed elevated Nrg-1 and ErbB4 mRNA expression in the phosphatidylserine emulsion-treated group; high Nrg-1 and ErbB4 expression levels are associated with better myelination , while Nrg-1 gene knockdown in mice lead to schizophrenia-like behavior (Cong et al., 2022). These results imply that ATRN mutations may reduce myelination and impair learning and memory in rats by inhibiting BDNF/TrkB and Nrg-1/ErbB4 signaling. ATRN is important for myelination, so ATRN may be added to the growing list of genes associated with low myelin leukodystrophy, which seems to be limited to the central nervous system, so ATRN may be a new clinical breakthrough in the study of this disease. The findings of the present study are also consistent with previous studies in our lab , suggesting the novel ATRN mutant rat can serve as an ideal model for further study of the multiple functions of ATRN.

CONFLICT OF INTEREST
The authors have no conflict of interest to declare.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.