Abnormal brain iron metabolism in Irp2 deficient mice is associated with mild neurological and behavioral impairments

Abstract

Iron Regulatory Protein 2 (Irp2, Ireb2) is a central regulator of cellular iron homeostasis in vertebrates. Two global knockout mouse models have been generated to explore the role of Irp2 in regulating iron metabolism. While both mouse models show that loss of Irp2 results in microcytic anemia and altered body iron distribution, discrepant results have drawn into question the role of Irp2 in regulating brain iron metabolism. One model shows that aged Irp2 deficient mice develop adult-onset progressive neurodegeneration that is associated with axonal degeneration and loss of Purkinje cells in the central nervous system. These mice show iron deposition in white matter tracts and oligodendrocyte soma throughout the brain. A contrasting model of global Irp2 deficiency shows no overt or pathological signs of neurodegeneration or brain iron accumulation, and display only mild motor coordination and balance deficits when challenged by specific tests. Explanations for conflicting findings in the severity of the clinical phenotype, brain iron accumulation and neuronal degeneration remain unclear. Here, we describe an additional mouse model of global Irp2 deficiency. Our aged Irp2-/- mice show marked iron deposition in white matter and in oligodendrocytes while iron content is significantly reduced in neurons. Ferritin and transferrin receptor 1 (TfR1, Tfrc), expression are increased and decreased, respectively, in the brain from Irp2-/- mice. These mice show impairments in locomotion, exploration, motor coordination/balance and nociception when assessed by neurological and behavioral tests, but lack overt signs of neurodegenerative disease. Ultrastructural studies of specific brain regions show no evidence of neurodegeneration. Our data suggest that Irp2 deficiency dysregulates brain iron metabolism causing cellular dysfunction that ultimately leads to mild neurological, behavioral and nociceptive impairments.

Figure 1. Expression of iron homeostasis proteins in Irp2^−/− and WT mice.

A) Western blot analysis of tissue extracts from male Irp2^−/− and WT mice (n = 3 mice/genotype) using antibodies to detect Irp2, Irp1, ferritin (Ftl1), TfR1 and β-actin (loading control). A non-specific band migrating near Irp2 is observed in some Irp2^−/− lysates. B) Irp1 and Irp2 RNA-binding activity in lysates was assayed by RNA electrophoretic mobility shift assay using a ³²P-labeled ferritin IRE as a probe. 0.5% β-mercaptoethanol was added to samples to assay total RNA-binding activity.

Figure 2. Locomotion, motor coordination and nociception are impaired in Irp2^−/− mice.

Irp2^−/− mice display reduced horizontal locomotor activity (total distance traveled, number of turns, number of total line crossings, mean velocity and angular velocity), and B) reduced vertical exploratory activity (number of rearing and rearing latency) assessed by the modified-Hole Board test . C) Left panel, performance of Irp2^−/− and WT mice on the accelerating rotarod in four trials on four consecutive trials with 15 min inter-trial-interval; right panel, decreased mean latency of Irp2^−/− mice to fall off the rotarod (n = 4 trial; p = 0.055). D) 4-paw grip force test shows no difference in muscular strength between Irp2^−/− and WT mice. E) Hot plate test shows increased hind paw licking in Irp2^−/− mice. Data are given as the mean ± SEM; *p<0.05; ^** p<0.01, ***p<0.001, relative to WT; WT (n = 9) and Irp2^−/− (n = 10).

Figure 3. Iron accumulates in specific brain regions of Irp2^−/− mice.

Coronal sections from male Irp2^−/− and WT brains from rostral (top) to caudal (bottom) were selected to show iron accumulation in specific regions of brain. Ferric iron was detected using DAB-enhanced Perls' stain. Ctx cortex, CB cerebellum, CPu caudate putamen, cc corpus collosum, Gp globus pallidus, Hy hypothalamus, PAG, periaqueductal gray; SC superior colliculus, SN substantia nigra and Thl thalamus.

Figure 4. Iron accumulates in axons and oliogodendrocyte cell bodies of Irp2^−/− mice.

Increased DAB-enhanced Perls' iron stain is observed in small cells that have an eccentric nucleus characteristic of oliogodendrocyte morphology and in axons in A) cerebral cortex, B) caudate putamen, C) superior colliculus, D) substantia nigra, and E) cerebellum. Arrows, oligodendrocytes; S, striosomes; wm, white matter; gl, granule layer. Scale bars: 50 μm and 500 μm.

Figure 5. Iron is reduced in Purkinje neurons and in CA1 pyramidal neurons of Irp2^−/− mice.

A) Perls' DAB-enhanced iron staining in CA1 pyramidal neurons and Purkinje neurons in Irp2^−/− and WT mice (Pkj, Purkinje; gl, granular layer; wm, white matter). B) Quantification of Perls' staining of images in (A) was carried out by measuring staining intensity in a defined subsection of brains from Irp2^−/− (n = 10) and WT (n = 8) (Figure S3). Data are given as the mean ± SEM; *p<0.05; ***p<0.001 relative to WT. Scale bars: 25 μm and 50 μm.

Figure 6. Ferritin expression is increased in Irp2^−/− neurons.

Double immunofluorescence labeling of brain sections from male Irp2^−/− and WT mice using ferritin antibody with either calbindin (Purkinje specific) or NeuN (neuronal nuclei) antibodies. A) Calbindin immunofluorescence shows that Purkinje cell number and morphology (yellow arrows, processes), are normal in Irp2^−/− cerebellum. B) Ferritin and calbindin double immunofluorescence of cerebellum shows similar ferritin expression in Irp2^−/− and WT Purkinje neurons. C) Ferritin and NeuN double immunofluorescence of hippocampal CA1 pyramidal cell layer, and D) caudate putamen shows increased ferritin expression in neuronal cell bodies (yellow arrows) of Irp2^−/− mice compared to WT mice. Increased ferritin staining is observed in the neuropil in the hippocampus and caudate putamen of Irp2^−/− mice. Scale bars: 50 μm.