The fmr1 KO zebrafish model with the impairment in FMRP expression, demonstrated similar pathological, physiological and behavioral abnormalities, and abnormal phenotypes as human FXS patients. Therefore, we turned our research interest for studying the possible role of FMRP on the development of social behavior and tried to clarify the relationship between social behavior and anxiety level and ensured development stage shoaling behavior. Furthermore, we wanted to study the possible therapeutic effect of the dietary supplement of linseed oil, a highly short-chain n-3 PUFAs, and fish oil, which has abundant long-chain n-3 PUFAs. We anticipated those dietary supplement may rescue the behavioral abnormalities including hyperactivity, abnormal anxiety-level, inhibitory avoidance learning impairment and shoaling preference behavior difference in fmr1 KO zebrafish. Results obtained from the present study will not only expand our knowledge of fmr1 function but also beneficial for the development of novel therapeutic strategies to the FXS patient.
Materials and Methods 1. Animals
The fmr1 knockout (KO) zebrafish, caring the mutation in the fmr1hu2787 allele, was acquired from the Wellcome Trust Sanger Institute Zebrafish Mutant Resource. The mutation caused a premature termination at codon position 113 via a C432T change in allele (den Broeder et al., 2009). The behavioral tests were used adult male fish or larvae fish to analyze. Tupfel long fin strain (TL) was used as the control background.
Animals of the present study were maintained according to standard guideline (Westerfield, 2007) The experimental procedures were approved and supervised by the Institutional Animal Care and Use Committee (IACUC) of National Taiwan Normal University.
2. Genotyping
Transgenic zebrafish were subjected for genotyping for confirmation.
Tissues were collected from the fish fin and the genomic DNAs were extracted by using commercialized DNA purification kit (Genemark, Taipei, Taiwan). Quantification of DNA was performed by using a nano spectrophotometer for determine the optical density (OD) at 260 nm and 280 nm (NanoDrop Technologies, Inc. Wilmington, DL). A derived cleaved amplified polymorphic sequences (dCAPs) assay was applied.
The fmr1 gene sequence was amplified by using forward primer and reverse primer (Table 1) which product size was 222-bp. The PCR cocktail contains 200ng genomic DNA, 0.5 mM dNTP, 1µM of forward and reverse primer, 1unit Prozyme DNA polymerase (Protech Enterprise,
Taipei, Taiwan) and 1X PCR buffer. The condition of PCR was summarized as following.
Cycle 1 40 1
Time 4 min 30→30→20 (sec) 20 sec
Temperature (°C) 94 94→60→72 72
The amplified sequences were further confirmed by using specific restriction enzymes. Briefly, the PCR product was cleaved by 1 unit RsaI restriction enzyme and the condition was 1X restriction enzyme buffer for overnight (New England BioLabs® Inc., UK). The amplified cDNA products were subjected to restriction enzyme digestion. The cDNA derived from WT DNA template contained two fragments (193-bp and 29-bp). Conversely, the amplified sequence was not split if the template DNA was acquired from hu2787 mutation strain. The PCR products were verified by using electrophoresis in 3% agarose gel which was prepared from TBE buffer (Ng et al., 2013).
3. Western blot analysis
The fmr1 KO zebrafish were further confirmed by using western blotting to evaluate the expression of FMRP. Zebrafish were sacrificed and brains were dissected out, homogenized and centrifuged for protein extraction. We followed the procedure of NuPAGE® Technical Guide for sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (Invitrogen, USA). After electrophoresis, proteins were transferred to PVDF membrane (Millipore, Bedford, MA). The non-specific binding
was blocked with 5% nonfat milk, 0.05% Tween and 1X PBS in room temperature for 1 hour. Rabbit polyclonal anti-FMRP (GTX125996, 1:1000; Genetex, Irvine, USA) antibody and mouse polyclonal anti-beta-actin antibody were applied in 4 °C for overnight.
Chemiluminescence assay was used (Bioman Scientific Co. Ltd., Taiwan) to visualize the signals and protein contents were determined by a LAS3000 digital imaging system (Fujifilm, Tokyo, Japan).
4. Behavioral analysis
Several behavioral paradigms were applied to determine the fmr1 KO effects on development of social behavior and possible therapeutic effects of n-3 PUFAs supplement. For example, the locomotor activity was used to analyze the spontaneous motor activity; novel tank was applied to quantify anxiety-like behavior; social score was detected by shoaling and shoaling preference behavior; and inhibitory avoidance task (IA) was also applied for studying the formation of associative memory. The detail procedures of each paradigm were summarized as following.
4.1. Locomotor activity
For larva, animals were placed into a 6 cm petri dish with water depth kept in 0.4 cm height for larva behavioral test. For adult fish, a transparent acrylic cylinder (20 cm in height and 24 cm in diameter) with water depth kept with 3.5 cm was used. EthoVision video tracking system (Noldus Information Technology, Leesburg, VA, USA) was applied for data acquisition and analyzing. The recording has two stages including a
5 min adaptation and following with a 10 minutes recording. The total moving distance within the 10 minutes recording was used for analyzing the locomotor activity of the zebrafish.
4.2. Novel tank test
In this behavioral test, zebrafish was placed into a transparent rectangular tank (6 cm length, 2 cm width and 6 cm in height), which water depth kept with 4 cm for larval fish test or a transparent trapezoidal tank (26 x 22 x 6x 11.5 cm, top x bottom x width x height) and water depth kept with 11.5 cm for adult fish analysis. Furthermore, two tanks were divided into upper and lower water zone. Each zone contained 1/2 of the total volume in the two different tanks. During the experiment, fish were freely swimming in the tank for 15 minutes, and we divided three different 5 minutes from total detect duration. We calculated the duration in upper zone and transferred to percentage, which value indicates anxiety level in fish (Egan et al., 2009).
4.3. Social behavior
To explore the development of social behavior, we followed behavioral procedure from our previous study (Hsu et al., 2014). There were three compartments in the behavior chamber for determining the shoaling and shoaling preference behavior in zebrafish (Fig 2).
Compartments were separated by transparent acrylic plates, and each side compartment was placed with different group of fishes, either zebrafish and/or medaka. Animals were placed into the center area (6.4 x 3.0 x 2.5
x 1.0, length x width x height x depth) and time spent in each side compartment was quantified. Briefly, the center compartment was further divided into three different zones. Both side zones, which closed to both side compartments, occupied 1/5 of the volume of center compartment and the center zone occupied 3/5 of the volume of center area. We analyzed larval fish from 6 days post fertilization (dpf) to 28 dpf. The experimental designs used here including group of zebrafish on both side (ZF-ZF), groups of medaka on both end (MK-MK), or place the zebrafish in one end with medaka in the other end (ZF-MK). The social activity was detected by using an EthoVision video tracking system (Noldus Information Technology, Leesburg, VA, USA) for 10 minutes, and we calculated the duration of swimming time in three different zones and transferred the time to percentage of shoaling “(duration in both side zones)*100 / total duration” and/or shoaling preference score “duration in closing zebrafish zone/ total duration for shoaling”.
The n-3 PUFAs diet supplement studies were aimed to clarify whether supplement had rescue effects on social behavior abnormalities in fmr1 KO zebrafish. The apparatus was designed for testing adult fish, with center area size (24 x 12 x 12 x 4.5 cm, length x width x height x depth). In addition, the duration of experiment was 25 minutes. The first 10 minutes is called acclimation phase, both screens were blocked by white acrylic board for avoiding the visual contact between the subject and other group of fishes. After that, both acrylic boards were removed and the fish could observe the different groups of fishes (ZF-MK) in each end. We detected the time of the zebrafish spend on the three different
zones by EthoVision video tracking system (Noldus Information Technology, Leesburg, VA, USA) The testing time was set for 15 minutes. In the same way for larval study, we calculated the percentage of shoaling and shoaling preference score for qualifying the therapeutic effects of n-3 PUFAs supplement (Engeszer et al., 2004; Engeszer et al., 2007).
4.4. Inhibitory avoidance (IA)
In this research, the apparatus (28 x 12 x 17, length x width x height) was divided into two different zones, a shallow water zone (2 cm in depth) and a deep water zone (8 cm in depth). They were separated with a white, opaque guillotine door between shallow and deep compartment (Fig. 3).
Based on the natural preference of zebrafish for a deep environment comparing to a shadow one (Darland and Dowling, 2001), the procedure contained a training phase and a testing phase. Animals were allowed to acclimate to the behavior chamber via a habituation session. They were placed in the shallow chamber for 5 min; the white, opaque guillotine door was then removed, and the fish were allowed to swim freely between the two compartments for another 5 min. In the training session, the fish were placed in the shallow compartment, allowing them to swim for 1 min before the guillotine door was opened. Once the fish entered the deep compartment, the guillotine door was closed, and a mild electric shock was applied to the deep compartment for 5 s. Animals were then tested for the avoidance learning 24 hours later. During the testing phase, we used the similar procedure with training phase but omitting the electrical shock. The duration of fish moved from shallow to deep
chamber escape latency) was recorded and analyzed. The maximum duration for testing is 300 seconds.
5. Gene expression quantitative analysis 5.1. RNA extraction
Zebrafish was sacrificed with ice water, and we removed their telencephalon or liver and put in a microcentrifuge tube. 500 µl TRIzol®
Reagent (Invitrogen, USA) was added into the tube and we homogenized tissue by using a pestle. 100 µl chloroform (Sigma, USA) was added into the sample and completely mixed which kept on the ice for 10 min. Then, they were put into a centrifuge (Eppendorf, Germany) and centrifuged at 13000 rpm for 30 minutes, and the temperature was kept on 4 °C. Next, the aqueous phase was drawn to another clear tube, and we added 250 µl Isopropanol into the samples and gently mixed them. They were stored into a refrigerator keeping on -20 °C overnight. After that, the samples were centrifuged at 13000 rpm for 15 minutes, and the parameter of temperature was set on 4 °C. We removed supernatant and added 500 µl 75% ethanol (EtOH) into the tube. The samples were centrifuged for 5 minutes, the parameters were kept on 13000 rpm and 4 °C. After finishing centrifugation, we removed EtOH and dried them by using an oven which set on 60 °C until pallet transferred to transparent. Finally, the samples were dissolved through adding 20 µl DEPC water, and we quantified each sample by using NanoDrop ND-1000 (Thermo Scientific, USA) and stored at -80 °C (Alimuddin et al., 2005; Cheng et al., 2015).
5.2. Reverse transcription
We used the High-Capacity cDNA Reverse Transcription Kits (Applied Biosystems, USA) to transfer RNA into complimentary DNA (cDNA). There were two steps in the experiment. The first formula was displayed on below table:
Items Volume (µl)
RNA sample (2 µg) X
10X RT Buffer 2.0
10X RT Random Primers 2.0
25X dNTP mix (100 mM) 0.8
MultiScribeTM Reverse Transcriptase 1.0
Nuclease-free H2O 14.2-X
Total volume 20
The volume was kept 20 µl in each sample, and we put them into the PCR machine. The protocol was set at 25 °C for 10 min, 37 °C for 120 min, 85 °C for 5 min, and kept temperature at 4 °C. Finally, we stored cDNA (20X, 100 ng/µl) at a -20 °C refrigerator.
5.3. Quantitative real-time polymerase chain reaction (q-PCR) In order to investigate the difference of gene expression in telencephalon and liver between WT and fmr1 KO fish, we selected fmr1, grin1b (glutamate receptor, ionotropic, N-methyl D-aspartate 1b), gria2a (glutamate receptor, ionotropic, AMPA, 2a), htr2a, htr2b (5-hydroxytryptamine receptor, 2a and 2b), htr2cl1, htr2cl2 (5-hydroxytryptamine receptor 2C, G-protein-coupled-like 1 and 2),
fads2 (fatty acid desaturase 2), elovl5, elovl4a, elovl4b (ELOVL fatty acid elongase 5, 4a, and 4b) to be our target genes. eef1a1l1 (eukaryotic translation elongation factor 1 alpha 1, like 1) was used to be an internal control. Primer designs were placed on table 1.
The cDNA was diluted to 1X concentration (5 ng/µl). The q-PCR synthesis mix reagents included 4 µL diluted cDNA, 5 µL 2X SYBR green dye, and 1 µL each primer pair (forward and reverse, 10 µM). The method was analyzed by using a LightCycler480 system (Monocolor hydrolysis UPL-probe, Roche Applied Science). The reaction contained four steps set on next page protocol:
Cycle Temp (°C) Duration (sec) Ramp Rate
Incubation 1 95 20 4.4
Amplification 45 95 3 4.4
60 30 2.2
Melting
Curve 1
95 5 4.4
65 60 2.2
97 0 0.11
Cooling 1 40 30 2.2
In order to analyze the relative expression value for each gene. First, threshold cycle (CT) of target gene subtracted CT of internal control gene (ΔCT). Next, the ΔCT in each sample subtracted the average CT of the control sample (WT) called ΔΔCT. Finally, the relative fold expression levels were calculated by the formula 2T-ΔΔC and presented on the result.
6. Analyzing the fatty acids composition in fmr1 KO zebrafish
To analysis the composition of fatty acids in zebrafish, a protocol developed by Cheng et al in 2015 was applied. Briefly, we sacrificed the zebrafish and total fatty acids of whole fish body were extracted. Folch’s lipid extraction method were used to extract lipid content with mixing organic solvent (chloroform: methanol 2:1), containing butylated hydroxyanisole (BHA) 0.05 mg (Sigma, USA), and aqueous solution, containing 10 mL of 30 mM magnesium chloride (MgCl2) during overnight (Folch et al., 1957). Then, the organic layer was collected and evaporated with a vacuo concentrator (Eyela, Japan). One mL of 50%
potassium hydroxide (KOH) (Merck, Germany), 15 mL 90% alcohol (Merck), and boiling stones were added and incubated them at 90°C for 40 min to saponify crude lipid. After saponification, the solvent was kept to cool temperature and purified the samples with 30 mL water and 30 mL ethyl ether (repeated three times). Ten mL of 2 N hydrogen chloride (HCl) (Merck, Germany), 30 mL ethyl ether, and 2 drops of methyl orange (Merck, Germany) were added into samples and were gently shaken to reduce saponified lipid. After 1 min to steady the samples, the organic layers were collected and washed to neutral pH value using water, and evaporated through vacuo concentrator. The purified lipid samples were methyl esterified by addiction of 5 mL of boron trifluoride - methanol (BF3 – MeOH) 7% from the 14% solution (Sigma, USA) and boiling stones at 90°C for 20 min. we were then added 5 mL hexane (Sigma, USA) into the samples for 1 min. The organic layers were separated using saturated saline, and absorbed excess water through
sodium sulfate (Na2SO4) (Merck, Germany). The fatty acid methyl esters were dried and diluted with 400 µL highly pure hexane. Finally, Agilent 5975C Series GC-MSD (Agilent, USA) was used to analyze our lipid samples. The analytic condition was modified from previous study (Abu and Oluwatowoju, 2009), and the Agilent column was 30 mm*0.25 mm with a film thickness of 0.25 mm. We calculated their percentage of fatty composition to analyze differences between dietary treatments.
6.1. Fatty acids composition in zebrafish body
The lipid of zebrafish body was extracted and subjected for analyzing the content of fatty acid difference between WT and fmr1 KO zebrafish.
The percentage of fatty acid was determined by using GS-MS system.
6.2. Fatty acids composition in different n-3 PUFAs dose
The dose of n-3 PUFAs in diet were modified from previous study which significantly rescued behavioral abnormalities in fmr1 KO mice (Pietropaolo et al., 2014). Different doses of n-3 PUFAs (8% linseed oil, 4% linseed oil + 4% fish oil, and 8% fish oil) and Antemia nauplii were supplemented. The control diet was used as the commercial diet (Taikong, Taiwan), its composition of ingredient was displayed on table 2. Different dose of n-3 PUFAs diets were adjusted their type of lipid from commercial diet. The fatty acid composition among different groups were extracted and analyzed by using GC-MS system.
7. Behavior analysis after n-3 PUFAs supplement
In order to investigate the possible therapeutic effects of n-3 PUFAs supplement in adult fmr1 KO zebrafish, we chose 4% linseed oil + 4%
fish oil supplement diet to test therapeutic effects on their abnormal behaviors including hyperactivity, higher anxiety level, higher level of shoaling preference, and impairment on fear learning. The experimental procedure and grouping were described on below flow chart.
Results
1. Qualitative analysis of the genotype
The genotyping results showed three different genotypes including wild-type (WT), heterozygous strain (+/-) and homozygous (+/+). In the WT strain (+/+), its PCR product has an endonuclease cleavage site, two fragments, 193-bp and 29-bp (not showed in this figure) were obtained after enzyme digestion, in contrast, the PCR product in fmr1 KO strain (-/-) do not have cleavage site so its fragment size kept in 222-bp, and heterozygous strain (+/-) contained both 222-bp, 193-bp, and 29-bp fragments (Fig 4A). Moreover, the expression of FMRP (72 kDa) was not detected in the fmr1 KO zebrafish brain, the β-actin (42 kDa) was used as the internal control (Fig 4B) in this experiment.
2. Shoaling behavior and shoaling preference in wildtype and fmr1 KO zebrafish at 14 and 28 dpf
There were three experimental design including a zebrafish-zebrafish (ZF-ZF), medaka-medaka (MK-MK), and zebrafish-medaka (ZF-MK).
The expectancy value is 40 % which means no shoaling preference behavior were found. In ZF-ZF design (both side areas were placed with zebrafish), we used one sample t test to compare the expectancy value (40%). According to statistics result (WT-6 dpf, p = 0.9195; WT-14 dpf, p = 0.0073*; WT-28 dpf, p < 0.0001*; KO-6 dpf, p = 0.4563; KO-14 dpf p
= 0.0004*; KO-28 dpf < 0.0001*), it showed the development of shoaling behavior begin on 14 dpf in both wildtype and fmr1 KO zebrafish.
Moreover, we used the two-way ANOVA to analyze the influence of
genotype and time. The results indicated that there were significant effects in both genotype (F= 6.5017, p = 0.0131*) and age (F=49.4885, p
< 0.0001*) but there was no interaction between them (F = 1.4261, p = 0.2476). We used Dunnett-Hsu for post-hoc analysis in genotype effect, and used WT as the control group. It demonstrated that the percentage of shoaling in fmr1 KO was significantly higher than WT (p = 0.0131*). On the other hand, Tukey-Kramer post hoc analysis (α = 0.05) was used. A significant effect of age on the development of shoaling behavior was found.
In MK-MK design (both sides were placed with medaka), the statistical methods were the same as mentioned in figure 5A. The statistic results were (WT-6 dpf p = 0.0028*, WT-14 dpf p = 0.0415*, WT-28 dpf p = 0.0006*, KO-6 dpf p = 0.2705, KO-14 dpf p < 0.0001*, KO-28 dpf
<0.0001*), both wildtype and fmr1 KO zebrafish showed shoaling behavior developed in 14 dpf, however, the value in WT-6 dpf was lower than 40%. In two-way ANOVA analysis, the results indicated that there was significant effect in genotype (F= 17.1915, p < 0.0001*) and age (F= 38.2841, p < 0.0001*) but there were significant effects on both genotype and time (F = 2.1615, p = 0.1232). We used Dunnett-Hsu post-hoc for analyzing. It demonstrated that the percentage of shoaling in fmr1 KO was significantly higher than WT (p < 0.0001*). Tukey-Kramer post hoc (α = 0.05) was used to analyze the duration effects. The result showed the shoaling behavior increased incrementally with time (Fig 5B).
Our results demonstrated that shoaling behavior is found in 2 week-age-old larval of both group. In the subsequent experiment, we
between WT and fmr1 KO zebrafish. We excluded the 6 dpf group since no shoaling behavior was detected at this stage. An alternative design was used with placed zebrafish in one end and medaka on the other end (ZF-MK). Results showed both WT and fmr1 KO zebrafish displayed shoaling behavior in 14 and 28 dpf. The statistic values are summarized as following: WT-14 dpf p = 0.0003*, WT-28 dpf p < 0.0001*, KO-14 dpf p < 0.0001*, KO-28 dpf p < 0.0001*. Moreover, it also indicated a significant effect in genotype (F= 5.8615, p = 0.0201*), age (F=14.8873, p = 0.0004*), and there was no interaction between genotype and age was found (F = 0.0083, p = 0.928). The percentage of shoaling in both genotypes in 28 dpf was significantly higher than WT fish in 14 dpf (Fig 6A). Furthermore, we also explored the shoaling preference behavior in wildtype and fmr1 KO zebrafish. Comparing with expectancy value is 0.5, the statistical result (WT-14 dpf p = 0.6589, WT-28 dpf p = 0.1688, KO-14 dpf p = 0.5938, KO-28 dpf < 0.0001*) demonstrated that a precocious development in shoaling preference behavior in fmr1 KO fish.
In two-way ANOVA, the result showed there was no significant difference on the effect of genotype (F= 0.7369, p = 0.3958) but there was significant effect on age (F=0.0027, p = 0.0027*), and a significant interaction between genotypes and age was found (F = 4.1172, p = 0.0491) using two-way ANOVA. In addition, we used Tukey-Kramer (α = 0.05) for analyzing (Fig 6B). The preference score in fmr1 KO fish was significant higher than WT in 28 dpf. The results were consistency with our previous study (Hsu et al., 2014).
3. Locomotor activity in larval zebrafish
To determine locomotor activity in larval zebrafish of different genotypes, we evaluated distance moved of WT and fmr1 KO zebrafish at the age of 14 dpf and 28 dpf. Two-way ANOVA was used for determining the influence of genotype and development age. Result showed significant differences on the effects of genotype (F = 16.7401, p
To determine locomotor activity in larval zebrafish of different genotypes, we evaluated distance moved of WT and fmr1 KO zebrafish at the age of 14 dpf and 28 dpf. Two-way ANOVA was used for determining the influence of genotype and development age. Result showed significant differences on the effects of genotype (F = 16.7401, p