A neuron‐specific Isca1 knockout rat developments multiple mitochondrial dysfunction syndromes

Abstract

Background

Multiple mitochondrial dysfunction syndromes (MMDS) are rare mitochondrial diseases caused by mutation of mitochondrial iron–sulfur cluster synthesis proteins. This study established a rat model simulating MMDS5 disease in the nervous system to investigate its pathological features and neuronal death.

Methods

We generated neuron‐specific Isca1 knockout rat (Isca1 ^flox/flox ‐NeuN‐Cre) using CRISPR‐Cas9 technology. The brain structure changes of CKO rats were studied with MRI, and the behavior abnormalities were analyzed through gait analysis and open field tests, Y maze tests and food maze tests. The pathological changes of neurons were analyzed through H&E staining, Nissl staining, and Golgi staining. Mitochondrial damage was assessed by TEM, western blot and ATP assay, and the morphology of neurons was assessed by WGA immunofluorescence to detect the death of neurons.

Results

This study established the disease model of MMDS5 in the nervous system for the first time, and found that after Isca1 loss, the rats suffered from developmental retardation, epilepsy, memory impairment, massive neuronal death, reduced number of Nissl bodies and dendritic spines, mitochondrial fragmentation, cristae fracture, reduced content of respiratory chain complex protein, and reduced production of ATP. Isca1 knockout caused neuronal oncosis.

Conclusions

This rat model can be used to study the pathogenesis of MMDS. In addition, compared with human MMDS5, the rat model can survive up to 8 weeks of age, effectively extending the window of clinical treatment research, and can be used for the treatment of neurological symptoms in other mitochondrial diseases.

Using CRISPR/Cas9 technology and Cre‐LoxP system. This study established the disease model of multiple mitochondrial dysfunction syndrome (MMDS) in the nervous system for the first time, and found that after ISCA1 loss, the rats suffered from developmental retardation, epilepsy, memory impairment, motor coordination defect, abnormal brain structure, cell swelling, karyopycnosis, mitochondrial fragmentation, cristae fracture, reduced content of respiratory chain complex protein, and reduced production of ATP. Isca1 knockout may cause neuronal oncosis.

1. INTRODUCTION

Iron–sulfur (Fe‐S) proteins, which are involved in diverse fetal biological pathways such as the Krebs cycle, the mitochondrial respiratory chain and DNA repair, are conserved in most creatures. ¹ , ² In mammals, the mitochondrion (Mt.) is the essential organelle of Fe‐S protein biogenesis and maturation, and requires Fe‐S proteins for processes such as the electron transfer chain, lipoic acid synthesis, and iron metabolism. ³

NFU1, BOLA3, IBA57, ISCA2, and ISCA1 are involved in the last steps of the maturation of mitochondrial [4Fe‐4S]. ⁴ Gene mutations in these five proteins result in five types of multiple mitochondrial dysfunction syndromes (MMDS), named MMDS 1–5. The common symptom of MMDS is injury to organs that required a lot of energy such as heart, brain, and muscle, and most patients do not live past infancy. ² MMDS5 caused by mutation of Isca1 is the latest of five types of MMDS to be described. ⁵ So far, three homozygous variants at Gln87Lys, Val10Gly and Tyr101Cys from seven patients have been reported. ⁶ , ⁷ , ⁸ , ⁹ In all seven ISCA1 cases, clinical signs included early seizures in infancy, spasticity, nystagmus, sensorineural hearing loss, impairment of cerebrum and cerebellum and unfortunately led to early death. The molecular features of hyperlactatemia, impairment of lipoylation of the mitochondrial proteins and dysfunction of complexes II and IV were also detected.

Mt. is a vital organelle in the eukaryocyte, responsible for energy production, calcium signaling, reactive oxygen species (ROS) production, apoptosis, and other cellular processes. ¹⁰ Dysfunction of Mt. that caused by pathological mutations of Fe‐S proteins appear to share common mechanisms with other mitochondrial diseases. ¹¹ Therefore, although MMDS5 is a rare disease, it is important to investigate the mechanisms by which it impairs the function of certain tissues.

Our previous studies showed that systemic knockout of Isca1 causes early embryonic death in rats. ¹² The heterozygous myocardium‐specific Isca1 knockout rats exhibited dilated cardiomyopathy with larger chambers, cardiac dysfunction and myocardium fibrosis. ¹³ Myocardium‐specific ISCA1 homozygous knockout rats exhibited myocardial oncosis, heart failure, and early death within 10 days of birth. ¹⁴

Since the clinical symptoms of MMDS5 were mainly reflected in neurological impairment, in this study, we generated neuron‐specific Isca1 knockout rats (CKO). The rats exhibited clinical features corresponding to MMDS5 patients, such as dyskinesia, seizures, cognitive disorder, developmental retardation (dysplasia) and premature death. The impaired Mt. and synapses were observed in the CKO brain tissues. Furthermore, degeneration of electron transfer chain proteins and reduction of ATP production were detected, and oncosis may be the cause of neuronal death in CKO rats.

2. METHODS

2.1. Animals

The Isca1 ^flox/flox rats were produced in our laboratory. ¹⁴ The NeuN‐Cre rats were constructed by inserting the Cre under the promoter of the NeuN gene, which expressed the Cre enzyme specifically in neurons of rats (www.ratresource.com, GC00134).

The Isca1 ^flox/flox ‐NeuN‐Cre rats (named CKO in this study) were obtained by crossing Isca1 ^flox/flox and NeuN‐Cre rats, whereby the Isca1 could be deleted by the Cre‐Loxp conditional knockout system. The CKO rats were screened using tail DNA genotyping procedures. ¹⁴ DNA was extracted from the tail tissues of neonatal rats using the EasyPure® Genomic DNA Kit (China, Trans Gen Biotech, EE101‐22). Isca1 was amplified with primers of 5′ ATGGTTCCAGCACTTTGAAGG and 5′ AAGCTAATATGACAGTGGTGAGGC under the following conditions: 95°C 30 s, 59°C 30 s, and 72°C 2 min, for 35 cycles. The Cre was amplified with primers of 5′ CCTCTCAGATGTTGGAACTCTCT and 5′ GTGCCTTCTCTACACCTGCG under the following conditions: 95°C 30 s, 60°C 30 s, 72°C 2 min, for 30 cycles. The littermates with both Icas1 ^flox/flox and Cre PCR products were CKO rats.

All rats used in this study were from the Sprague–Dawley (SD) strain and were housed in a SPF facility. The breeding environment was maintained under barrier environmental standards with a 12 h light/dark cycle. Animals and associated handling procedures were in accordance with animal welfare requirements and approved by the Institutional Animal Care and Use Committee (IACUC) of the Institute of Laboratory Animal Science, Chinese Academy of Medical Sciences and Peking Union Medical College (MYW21011).

2.2. Behavior analysis

The behavior of wild type rats (WT) and CKO rats was first evaluated using Racine's scale, followed by gait analysis, the open field test, the Y maze test and the food maze test. The behavior analysis was performed in a SPF facility under standard conditions.

The Racine's scale evaluation was performed as described previously. ¹⁵ The parameters and corresponding scores are summarized in Table 1. The highest score among the parameters for each rat was considered as the score of this rat.

TABLE 1.

Racine's scale parameters and scores.

Parameter	No abnormality	Immobilization	Limb myoclonus	Tail held up stiffly	Falling down on its side	Tonic–clonic seizure
Score	0	1	2	3	4	5

Gait analysis of WT and CKO rats was performed as described previously. ¹⁶ Briefly, the rats' front and hind paws were coated with non‐toxic red or blue ink respectively. After a 2 min training period, each rat was allowed to walk along a narrow, paper‐covered track. The width and length of each stride was measured. The gait dysfunction was calculated from the width/length ratio of the stride.

The open field test was performed in an observation box (80 cm × 80 cm × 35 cm) divided into three zones: an outside zone, a middle zone and a center zone. ¹⁷ The WT and CKO rats were gently placed in the center of the open field and allowed to explore freely in the box. The motion trails of the rats were recorded for 5 min with a camera mounted above the three zones. Locomotor activity was analyzed with VisuTrack software (XinRuan, Shanghai, China).

The Y maze test was performed in rat Y maze apparatus (60 × 15 × 20 cm). The arms were marked A, B, and C for convenience. The rat was released at the end of one arm and allowed to explore freely in the three arms for 5 min. Entry to an arm was confirmed when the four limbs of the rat were within the arm. ¹⁸ The sequence that the rat entered different arms was recorded by VisuTrack software, which recognized a spontaneous alternation, such as A‐B‐C or C‐A‐B. Alterative rate (%) = alternative number/(total entry −2) × 100%. Between each test, the apparatus was cleaned with 75% ethanol.

Regarding the spatial memory test, the CKO rats were too weak to complete the Morris water maze test (MWM). We therefore designed a food maze instead of the MWM. The apparatus consisted of an observation box (120 cm × 80 cm × 30 cm). The inside of the box was divided into nine zones by 30 cm‐high walls. Six zones were connected by channels and three zones were impassable. All of the channels were marked with different shape objects to help the animals recognize passable and impassable channels. A release area was designed on one side of maze. A toy or food (sunflower seed or peanut) was placed in the zones at another side of the maze (Figure 3I). The function of the toy was to detect the accuracy of the rat's memory of a specific object, not just its memory of the location. In the training phase, the rats were fasted, but with water supplied, for 12 h. WT or CKO rats were placed into the release area and the hungry rats were allowed to explore freely in the maze for 30 min. As the hungry rats tried to find the food by exploring passable and impassable zones they acquired a spatial memory on the maze. The interval between the training phase and the experimental phase was 48 h. After 12 h fasting, each rat was placed in the release area and allowed to find the food zone. The toy or food was removed to avoid the smell of the food. Over a period of 1 min, the time to arrive at the entrance of the food zone was recorded using a camera. If the rat could not find the food zone within 1 min, the result is recorded as 60 s. After each test, the apparatus was cleaned with 75% ethanol.

FIGURE 3.

Behavioral abnormalities and memory impairment of CKO rats. (A) CKO rats with epileptic seizure at the eighth week. (B) Racine's Scale was used to score the seizures of CKO rats. (C) Representative graphs for the trajectories of WT and CKO rats in the open field test at the sixth week. (D) Representative graphs for the trajectories of WT and CKO rats in the Y maze test at the sixth week. (E) Total distances moved, speeds, time in the central area and the number of times entering the central area were compared between WT and CKO rats in the open field test. (F) Spontaneous alternation rates in the Y maze test were calculated and compared between WT and CKO rats. (G) Representative gait stamps of WT and CKO rats at the sixth week. (H) Stride width/stride length in the gait analysis were measured and compared between WT and CKO rats. (I) Representative graphs for the movement trajectory of WT and CKO rats in the food maze test at the sixth week. (J) Latency time to reach the food zone in the food maze were recorded and compared between WT and CKO rats.

2.3. Survival analysis

Survival analysis was performed as described previously. ¹⁴ The cumulative mortality of the littermates including homozygous CKO and wild‐type rats was recorded from birth to 10 weeks of age. The Kaplan–Meier curves were analyzed using GraphPad Prism8 software through comparison with the log‐rank test.

2.4. MRI examination

In vivo magnetic resonance imaging (MRI) was performed on a 7.0 T small animal MRI system (Varian, Palo Alto, CA, USA). For imaging, wild‐type and model rats (n = 4/group) were anesthetized with medical oxygen (1 L/min) containing 2% isoflurane. Fast spin‐echo (fsems) sequences were used to acquire coronal T2‐weighted MRI images with the following parameters: field of view = 35 × 35 mm; no. of slices = 20; slice thickness = 1 mm (zero slice gap); TR (repetition time) = 3000 ms; TE (echo time) = 18.0 ms; image matrix = 256 × 256; average = 10. Interesting regoins, such as corpus callosum, cortex, hippocampus, and ventricles, were plotted on T2 images to determine changes in signal intensity.

2.5. Pathology staining

The brain tissues from WT and CKO rats at 8 weeks of age were fixed with formaldehyde (10%) for 3 days at room temperature. The fixed tissues were then mounted in paraffin. Coronal sections of cerebrum and sagittal sections of cerebellum (4 μm in thickness) were prepared and stained with H&E following a standard pathology procedure. ¹⁹ The sections were observed using a digital slide scanner (3DHISTECH, pannoramic 250, Hungary) and the images were captured by Caseviewer image viewing software.

For Nissl staining, the coronal sections of cerebrum and the sagittal sections of cerebellum were dewaxed using a standard pathology procedure. The sections were then rinsed three times with distilled water for 5 min each time. The sections were immersed in 1% Toluidine Blue and stained for 40 min at 60°C. The staining solution was washed off with distilled water. The sections were then dehydrated in serial of ethanol, and cleared in xylene. The sections were finally mounted with mounting medium and observed using a digital slide scanner, and the images were captured by Caseviewer image viewing software.

Golgi staining was performed using a FD Rapid Golgi stain Kit following the manufacturer's instructions (NeuroTechnologies, Inc, PK401). Briefly, fresh brain tissues were dissected into 10‐mm thick blocks. The tissues were fixed in mixture of solutions A and B at room temperature for 2 weeks and then transfered to solution C at room temperature in the dark for 72 h. Sections 100 μm thick were prepared using a cryostat at −20°C. The sections were rinsed in double distilled (DD) water and stained in a mixture of solutions D and E for 10 min. The sections were then rinsed in DD water and dehydrated, cleared and mounted following a standard pathology procedure. Section images were captured using the digital slide scanner.

For transmission electron microscopy (TEM) observation, the fresh brain tissues were fixed in 2.5% glutaraldehyde for 12 h at 4°C and then transferred to osmium tetroxide buffer (1%) for 1 h. Ultrathin sections (100 nm) were obtained using an ultrathin sectioning machine. The mitochondrial structures of the neuronal cells were examined under JEM‐1230 transmission electron microscopy. ²⁰ For immunofluorescence staining, the sagittal or coronal sections of brain tissues were dewaxed and rehydrated using standard pathology procedures. The sagittal sections were stained with primary antibodies against ISCA1 (Thermo, PA5‐60121, 1:200) and goat‐anti‐rabbit secondary antibody. The coronal sections of cerebrum and the sagittal sections of cerebellum were stained with wheat germ agglutinin conjugated Texas Red (USA, Invitrogen, W7024, 1:200). All the sections were mounted with fluorescent mounting medium in DAPI (China, ZSGB‐BIO, ZLI‐9557). Section images were captured by the digital slide scanner.

2.6. Protein extraction and western blot analysis

Total protein lysates were prepared from rat brains as described previously. ²¹ Briefly, rats were anesthetized with 2% isoflurane by inhalation, and brain tissue was gently removed by meticulously peeling away the skull of rats with bone forceps. The brain tissues were homogenized in pre‐cooled RIPA (Beyotime, P0013b) containing protease and phosphatase inhibitor cocktail (Thermo, 78444) at a concentration of 0.01 g tissue/100 μL RIPA. The homogenized tissues were cooled on ice for 30 min, and then centrifuged at 10 000g at 4°C for 30 min. The supernatants were collected as total lysates of brains. The mitochondrial proteins were isolated according to the manufacture instructions with a Tissue Mitochondria Isolation Kit (China, Beyotime, C3606). Protein concentrations were determined using a BCA protein assay Kit (Thermo, 23 225). The protein samples were separated on a 10% SDS‐PAGE and transferred to 0.45 μm nitrocellulose membranes (Millipore, Immobilon NC). The NC membranes were incubated overnight at 4°C with primary antibodies (Table 2). The primary antibodies were washed off and then stored with appropriate secondary antibodies (Table 2) for 1 h at room temperature. The blots were visualized by ECL (Santa Cruz, sc‐2048) and imaged by Gel Imaging System (Bio‐Rad, ChemiDoc XRS+ Gel Imaging System). The images were quantitatively analyzed with Image J software.

TABLE 2.

Antibodies used in this study.

Antibodies	Concentration	SOURCE
ISCA1	1:500	Thermo, PA5–60121
GAPDH	1:10 000	Proteintech, HRP‐60004
TOM20	1:1000	Thermo, PA5‐52843
NDUFA9	1:1000	Abcam, ab14713
NDUFS3	1:1000	Abcam, ab110246
SDHB	1:200	Abcam, ab14714
Aconitase 2	1:1000	Abcam, ab110321
UQCRC2	1:1000	Abcam, ab14745
COX IV	1:1000	Abcam, ab16056
HRP‐conjugated anti‐rabbit IgG	1:10 000	ZXGB‐Bio, ZB‐2301
HRP‐conjugated anti‐mouse IgG	1:10 000	ZXGB‐Bio, ZB‐2305
HRP‐conjugated anti‐goat IgG	1:10 000	ZXGB‐Bio, ZB‐2306

2.7. ATP content analysis

ATP in the WT and CKO rats brain tissues was determined using an ATP content detection kit (Solarbio, BC0305) according to the manufacture's instructions. Briefly, the brain tissues were homogenized in extract buffer on ice at a concentration of 0.1 g/mL. The supernatants were separated by centrifugation at 10 000g at 4°C. Proteins in the supernatants were removed by extraction with chloroform. The aqueous phase was separated by a centrifugation at 10 000g at 4°C and the aqueous phase was collected for ATP assay. The standard samples and the brain tissue samples were respectively mixed with solution 1 and working solution in the 96‐well plate. The OD at 340 nm was detected immediately on a micro‐plate reader and the OD was designated A1. The plate was then incubated at 37°C for 3 min. The OD at 340 nm was detected again and designated A2. The ATP concentration was calculated using the formula: C = (A2‐A1)÷(ΔA _standard ÷C _standard) × V _total ÷W. Three duplicates were performed for each sample.

2.8. Statistical analysis

All experiments were performed at least 3 times and all samples were tested 3 times in duplicate. Data are expressed as mean ± S.D. Unless otherwise stated when only two‐group tests were performed, Student's t test was used for statistical analysis. A log‐rank (Mantel‐Cox) test was used for comparison of survival curves. In all cases, two‐tailed P values less than 0.05 were considered statistically significant. Statistical analysis was performed using GraphPad Prism software (version 8.0.2, San Diego, CA, USA).

3. RESULTS

3.1. Establishment of neuron‐specific Isca1 knockout rats

The expression pattern of ISCA1 was firstly observed in WT rats at 2 months of age by immunofluorescence staining with ISCA1 antibody (Figure 1A). The result showed that ISCA1 was expressed throughout the whole brain of WT rats. The neuron‐specific Isca1 knockout rats were produced by crossing Isca1 ^flox/flox with NeuN‐Cre rats. The third exon of the Isca1 gene was floxed in Isca1 ^flox/flox rats. The Cre gene was inserted downstream of the NeuN promoter, which is a neuron specific expression gene, and the Cre enzyme was specifically expressed in neurons of NeuN‐Cre rats. In CKO rats, the third exon of the Isca1 gene was removed by the Cre enzyme, which resulted in the deletion of the ISCA1 protein in neurons, with a knockout efficiency of 72.05% (Figure 1B–E). The CKO rats showed developmental retardation compared with the WT rats and died before 10 weeks of age (Figure 1F–H).

FIGURE 1.

Generation of neuron‐specific Isca1 knockout rats. (A) The expression of ISCA1 in a brain section from a WT rat detected by immunofluorescence staining. DAPI (blue) stained the nucleus, and TRITC (red) conjugated secondary antibody was used to detect the ISCA1‐specific antibodies. Scale bar: 1000 μm. (B) PCR was used for the genotyping of Isca1 CKO rats. Among the 12 littermates, numbers 2 and 6 are homozygote, 5 and 9 are wild types, and 3 is heterozygote. (C) Schematic diagram describing the strategy for generating CKO rats. (D) Western blots using an ISCA1‐specific antibody showed the absence of ISCA1 protein in the brain tissues of CKO rats. GAPDH was used as a loading control. (E) Representative images showed the absence of ISCA1 protein in different regions of WT and CKO rats by IHC staining at the eighth week (S1, primary somatosensory cortex; DG, dentate gyrus). Scale bar for overview, 1000 μm; scale bar for magnified view; 50 μm. (F) Body weights compared between WT and CKO rats. (G) A representative graph shows the appearance of WT and CKO rats at the eighth week. (H) Survival analysis for WT and CKO rats; n = 17/group.

3.2. Isca1 knockout caused abnormal brain structure

The geometric structure of the brains of WT and CKO rats at 6 weeks was examined by MRI. The results showed that the volumes of cerebrum and cerebellum of the CKO rats were significantly reduced and the lateral ventricles were obviously dilatated compared with WT rats (Figure 2A,B, n = 4, *p < 0.05, ***p < 0.001). In addition, T2 images showed abnormal hyperintensity in the hippocampal fissure and alveus of the hippocampus, which suggested that pathological changes existed in the hippocampus of CKO rats (Figure 2A, the 4th row). Gross anatomy examination and weight metering also detected a severe cerebellar atrophy in CKO rats (Figure 2C–E, n = 7, **p < 0.01, ***p < 0.001).

FIGURE 2.

Brain alterations in CKO rats. (A) Representative MRI images for serial coronal planes shows the brain abnormality of CKO rats at the sixth week. (B) Volumes of cerebrum, cerebellum, and lateral ventricle of WT and CKO rats calculated and compared on T2 images. (C) A representative graph showing the gross morphology of brains from WT and CKO rats at the eighth week. (D) Weights of cerebrum and cerebellum of WT and CKO rats were recorded and compared. (E) Cerebellum weight/cerebrum weight compared between WT and CKO rats.

3.3. Isca1 knockout caused dysfunctional behavior

The WT and CKO rats were observed from birth to 8 weeks of age and their natural behavior was scored using Racine's scale (Figure 3A,B, n = 6). The CKO rats showed limb myoclonus started at 4 weeks of age. Progressive dysfunctional behavior, such as a rat holding its tail up stiffly and/or falling down on its side, occurred with aging. Most of the CKO rats suffered tonic–clonic seizure and died before 10 weeks of age. The abnormal motor ability of CKO rats at 6 weeks of age was further analyzed using open field tests and gait tests. In open field test, the CKO rats showed reduced speed and distance of movement, and could not move along a straight line compared with WT rats. At the same time, the time spent in the central area and the number of times in the central area were significantly reduced, which may be related to anxiety (Figure 3C,E, n = 7, *p < 0.05, **p < 0.01, ***p < 0.001). The CKO rats showed obviously motor coordination defects compared with WT rats in gait test (Figure 3G,H, n = 4, ***p < 0.001).

Cognitive behavior was tested using a Y maze and a food maze. In the Y maze, the CKO rats at 6 weeks of age showed a significantly impaired working memory indicated by a reduction in spontaneous alternation rat compared with WT rats (Figure 3D,F, n = 5, **p < 0.01). In the food maze, CKO rats at 6 weeks of age showed a significantly impaired spatial memory indicated by an increase in latency time to reach the target zone through the maze compared with WT rats (Figure 3I,J, n = 7, *p < 0.05).

3.4. Isca1 knockout caused brain pathological changes

The coronal sections of cerebrum and the sagittal sections of cerebellum from WT and CKO rats at 8 weeks of age were stained with H&E and observed under a light microscope. Cavitation of brain tissue was observed throughout all brain sections of CKO rats and obvious pathological changes were found in primary somatosensory cortex (S1), hippocampus, posteromedial thalamic nucleus (VPM) and cerebellum. In the S1 area of CKO brain, karyopycnosis, cell swelling and tissue cavitation were clearly observed (Figure 4A). In the hippocampus (CA3 and DG) areas of CKO brain, serious disarray of pyramidal neurons, cell swelling, karyopycnosis and tissue cavitation were observed (Figure 4A). In VPM, serious cell swelling and cavitation was observed (Figure 4A). In the cerebellum of CKO brain, cell swelling and karyopycnosis of Purkinje and granulosa and tissue cavitation were observed (Figure 4B).

FIGURE 4.

H&E staining of brain tissue sections. (A) H&E representative images for the cerebrums of WT and CKO rat brain (left) at the eighth week. Magnified images (right) showed the neuron morphology alteration on S1, CA3, DG and VPM. (S1, primary somatosensory cortex; CA3, hippocampus; DG, dentate gyrus; VPM, ventral posteromedial thalamic nucleus). Scale bar for overview, 1000 μm; scale bar for magnified view, 100 μm. (B) H&E representative images for the cerebellums of WT and CKO rat brain (left) at the eighth week. Magnified image (right) shows the neuron morphology alteration on cerebellar cortex (Cc) and cerebellar medulla (Cm) of the cerebellum. Scale bar for overview, 1000 μm; scale bar for magnified view, 100 μm.

3.5. Isca1 knockout affected neuronal function  and synaptic morphology

Nissl staining can be used to determine the function of neurons. The coronal sections of cerebrum and the sagittal sections of cerebellum from WT and CKO rats at 8 weeks of age were stained with Nissl staining. The CKO brain sections were lightly stained compared with WT rats (Figure 5A,B). A serious reduction of Nissl bodies, cell swelling, and tissue cavitation were observed in the S1 area and hippocampus of CKO brain (Figure 5A). A significant decrease in the number of positive neurons was observed in CKO brain (Figure 5C, n = 3, *p < 0.05, **p < 0.01). These results suggested that the function of neurons was obviously damaged by the Isca1 knockout in rat brain.

FIGURE 5.

Nissl staining and Golgi staining. (A) Representative images show Nissl staining in coronal cerebrum sections of WT and CKO rats (left) at the eighth week. Magnified images (right) show the Nissl bodies in three brain regions of S1 cortex, CA3, and DG in WT and CKO rats (S1, primary somatosensory cortex; CA3, field CA3 of hippocampus; DG, dentate gyrus). Scale bar for overview, 2000 μm; scale bar for magnified view, 100 μm. (B) Representative images show Nissl staining of the sagittal cerebellum plane of WT and CKO rats (left) at the eighth week. Magnified images (right) showed the Nissl bodies in two brain regions of Cc and Cm in WT and CKO rats (Cc, cerebellar cortex; Cm, cerebellar medulla). Scale bar for overview, 500 μm; scale bar for magnified view, 100 μm. (C) Quantitative analysis of the numbers of positive neurons per ×400 field in S1 cortex, CA3, and DG and per ×1500 field in cerebellum. (D) Representative images show Golgi stained dendritic spines of the hippocampal granule cells of WT and CKO rat at the eighth week. Scale bar, 10 μm. (E) Spine density (number/10um) and three types of dendritic spines were counted using Image J, and the statistical comparison between WT and CKO rats was performed using GraphPad software.

The morphology of dendritic spines was analyzed by Golgi staining. Three different types of dendritic spines including mushroom, stubby and thin spines were observed and their numbers were counted. The results showed that the numbers of mushroom, stubby and thin spines were significantly reduced in CKO brain compared with WT brain (Figure 5D,E, n = 3 and 5 fields/rat, ***p < 0.001).

3.6. Isca1 knockout damaged mitochondrial structure and function

The morphology of Mt. was observed by TEM and the results showed that mitochondria were seriously damaged in CKO brain, indicated by impaired membranes, swelling, cristae fracture and vacuolus (Figure 6A). The degeneration of respiratory chain proteins was analyzed by western blot (Figure 6B). NDUFA9 and NDUFS3 in complex I and SDHB in complex II were significantly reduced in CKO brains compared with WT brains (Figure 6B,D–F, n = 4, *p < 0.05, ***p < 0.001). While Aconitase 2, UQCRC2 in complex III, and COX IV in complex IV were not significantly interrupted by Isca1 knockout (Figure 6B,C,G,H). The impairment of Mt. resulted in reduction of ATP content by 39% in CKO brains compared with WT brains (Figure 6I, n = 4, *p < 0.05). These results suggested that Isca1 knockout caused degeneration of complex I and complex II and resulted in damage to Mt. structure and reduction of ATP production.

FIGURE 6.

Altered mitochondrial structure and function in CKO rats. (A) Representative transmission electron microscope (TEM) images show abnormal mitochondria morphology in CKO rats at the sixth week. Scale bar, 500 nm. (B) The expression of six mitochondrial respiratory chain proteins in the brain tissues of WT and CKO rats at the sixth week was detected by Western Blot. (C–H) Relative expression of Aconitase 2, NDUFA9, NDUFS3, SDH B, UQCRC2, and COX IV quantified using Image J. TOMM20 was used as a loading control for mitochondrial proteins. (I) The ATP content in brain tissues was measured using a kit, and the ATP synthesis in CKO brains was significantly decreased compared with WT brains.

3.7. Isca1 knockout induced neuronal oncosis

The coronal sections of cerebrum and the sagittal sections of cerebellum from WT and CKO rats at 8 weeks of age were stained with wheat germ agglutinin (WGA) conjugated with Texas Red, which could mark cross‐sectional area or cavitation caused by cell oncosis. ¹⁴ The extent of cavitation was clearly observed in hippocampus (Figure 7A,B, n = 4, ***p < 0.001), cerebral cortex (Figure 7C,D, n = 4, ***p < 0.001) and cerebellar medulla (Figure 7E,F, n = 4, ***p < 0.001). These results suggested that Isca1 knockout caused neuronal oncosis in the brain.

FIGURE 7.

WGA immunofluorescence. Representative images show the neuron vacuolations present on the hippocampus (A), cortex (C), and cerebellum (E) of WT and CKO rats at the eighth week. Scale bar, 20 μm. WGA (red) stained the cell membrane and DAPI (blue) stained the nucleus. The number of vacuolations in hippocampus (B), cortex (D), and cerebellum (F) were counted.

4. DISCUSSION

Here, we established a neuron‐specific Isca1 knockout rat by crossing Isca1 ^flox/flox and NeuN‐Cre rats (Figure 1). The CKO rats developed MMDS5‐like clinical features, including developmental retardation (Figure 1F,G), tonic–clonic seizures (Figure 3A,B), abnormal movement and motor coordination defects (Figure 3C,G), impaired cognitive ability (Figure 3F,I,J), encephalanalosis (Figure 2) and early death (Figure 1H). As the rats aged, the symptoms gradually worsened. Compared with human onset from 3 months of age to 23 months of age, ⁶ , ⁸ CKO rats did not develop symptoms until after weaning, and this delay may be caused by ISCA1 knockout occurring only in neurons without affecting other organs. At the same time, the other symptoms present in MMSD5 patients such as nystagmus and sensorineural hearing loss were not observed in ISCA1‐CKO rats.

Taking advantage of CKO rats with MMDS5 clinical features, we further investigated the pathological changes caused by Isca1 knockout in brain. H&E staining showed that karyopycnosis, cell swelling and tissue cavitation were distinctly observed in the cortex, hippocampus, posteromedial thalamic nucleus and cerebellum of CKO rats (Figure 4A,B). Nissl staining showed that the state of neurons was obviously affected by the Isca1 knockout in brain and the number of neurons decreased (Figure 5A–C). The number of dendritic spines were significantly reduced in CKO brain compared with WT brain (Figure 5D,E). This result suggested that Isca1 knockout could induce degeneration and death of neurons.

The iron–sulfur cluster biogenesis in eukaryotic mitochondria starts with the synthesis of [2Fe‐2S] by ISCU2, followed by GLRX5 dimeric translocates [2Fe‐2S] to an accessory protein system that involves several proteins (ISCA1, ISCA2, IBA57, FDX2, IND1 (NUBPL), BOLA3, and NFU1). The ISCA1‐ISCA2‐IBA57 complex plays an important role in the synthesis of [4Fe‐4S];FDX2 transfers electrons; NFU1, IND1, BOLA3, which are specialised Fe‐S‐targeting proteins, deliver [4Fe‐4S] clusters to receptor proteins such as complex I, complex II and lipoyl synthase. ² , ³ , ⁴ , ²² Loss of function of the ISCA1‐ISCA2‐IBA57 complex due to loss of ISCA1 may be the molecular mechanism of this study. ²³ , ²⁴ Not enough [4Fe‐4S] in mitochondria is transferred to NDUFS3 in mitochondrial complex I, SDHB in mitochondrial complex II, lipid acyl synthase, apoproteins and many other proteins requiring [4Fe‐4S]. We found that the core subunits of mitochondrial complex I NDUFA9 and NDUFS3, and the iron–sulfur protein subunits of mitochondrial complex II SDHB were significantly reduced in CKO brains compared with WT brains (Figure 6B), although Aconitase 2, UQCRC2 and COXIV were not changed. Considering the importance of electron transfer chain proteins for the maintenance of mitochondrial function, ²⁵ , ²⁶ , ²⁷ the expression level changes of some key mitochondrial proteins may be enough to have serious consequences. This corresponds with the results showing reduced ATP production and seriously damaged morphology of Mt. that were observed using TEM.

Neuronal oncosis may be the reason for cavitation after ISCA1 knockdown (Figures 4 and 5A,B). Oncosis is a type of necrocytosis caused by ATP depletion and failure of the ionic pumps of the plasma membrane. ²⁸ The main features of oncosis including cellular swelling, organelle swelling, blebbing and increasing membrane permeability, finally result in cavitation. We used WGA to stain cell membranes and observe the morphology of neurons. ²⁹ The brain sections from CKO rats demonstrated a large number of swollen cells that had clear boundaries, and excluded the possibility that cavitation was caused by mechanical or accidental damage.

Although mutations to five genes have been identified as the cause of MMDS in humans, there are few studies of the genes in mammals. MMDS1–5 have different clinical features. MMDS1 specifically causes pulmonary hypertension ³⁰ , ³¹ ; MMDS2 causes dilated cardiomyopathy ³² ; MMDS3 has a milder clinical phenotype, which may be related to the residual efficiency of mutant IBA57; MMDS4 and MMDS5 have similar symptoms such as seizures in infancy, spasticity, nystagmus, sensorineural hearing loss. The different symptoms of MMDS1–5 may be due to the different functions of the corresponding proteins during [4Fe‐4S] synthesis. So far, only two genomic modified rat models have been reported. The rats with a NFU1 mutation developed pulmonary hypertension, and did not show any symptoms in terms of brain pathology. ³³ Our previous work focused on developed heart failure in myocardium‐specific Isca1 knockout rats, ¹⁴ but research on the nervous system is very important. Isca1 neurological specific knockout rats can mimic the disease phenotype of human MMDS in the nervous system, and this rat model has the advantage of an extended therapeutic window compared to human premature mortality that could be widely used in clinical therapeutic studies.

Looking to the future, although gene therapy is a useful method for genetic diseases, ³⁴ , ³⁵ symptomatic medication, such as anticonvulsant medications to prevent seizures, may be a more viable strategy for now. ¹⁰ , ³⁶ A ketogenic diet is regarded as a safe and effective way to treat children who have epilepsy and respiratory chain defects. ³⁷ Similar dietary supplements include antioxidants, agents that regulate mitochondrial electron transfer flux, nitric acid precursors and so on. ³⁸ An interesting therapy is protein replacement, and one successful case is the treatment of a multiple metabolic mitochondrial disease called MNGIE syndrome, caused by a loss‐of‐function mutation in the gene encoding thymidine phosphorylase, by supplementing thymidine phosphorylase encapsulated by erythrocytes. ³⁹

AUTHOR CONTRIBUTIONS

All listed authors meet the requirements for authorship. LFZ and LZ conceived and designed the experiments and wrote the main manuscript text. HXS performed most of the experiments. DL contributed to Western Blot and immunofluorescence. XLQ contributed to the measurement of ATP content analysis. YHL and XZ contributed to establishment of the animal models and sequence analysis. JL, SG and XG contributed to the animal breeding and management. WD contributed to behavior analysis. WC contributed to microinjection technique.

FUNDING INFORMATION

The present work was supported in part by The National Key Research and Development Program of China (2022YFF0710702), National Natural Science Foundation (31970508) and CAMS Innovation Fund for Medical Sciences (CIFMS, 2021‐I2M‐1‐034).

CONFLICT OF INTEREST STATEMENT

Lianfeng Zhang and Dan Lu are Editorial Board members of AMEM and co‐authors of this article. To minimize bias, they were excluded from all editorial decision‐making related to the acceptance of this article for publication.

ETHICS STATEMENT

All the procedures were approved by the Animal Care and Use Committee of the Institute of Laboratory Animal Science, Peking Union Medical College.

Sheng H, Lu D, Qi X, et al. A neuron‐specific Isca1 knockout rat developments multiple mitochondrial dysfunction syndromes. Anim Models Exp Med. 2023;6:155‐167. doi: 10.1002/ame2.12318