doi:10.1093/brain/awh401 Brain (2005), 128, 711–722 Neuropathological, biochemical and molecular findings in a glutaric acidemia type 1 cohort Christopher B. R. Funk, 1,5 Asuri N. Prasad, 6 Patrick Frosk, 3 Sven Sauer, 7 Stefan Ko ¨lker, 7 Cheryl R. Greenberg 3,4 and Marc R. Del Bigio 1,2,5 Correspondence to: Marc R. Del Bigio MD PhD FRCPC, Canada Research Chair in Developmental Neuropathology, Department of Pathology, University of Manitoba, D212-770 Bannatyne Avenue, Winnipeg MB, R3E 0W3, Canada E-mail: [email protected]Departments of 1 Pathology, 2 Human Anatomy and Cell Science, 3 Biochemistry and Medical Genetics and 4 Pediatrics and Child Health, Faculty of Medicine, University of Manitoba, 5 Manitoba Institute of Child Health, Winnipeg, 6 Department of Pediatrics, University of Western Ontario, London, Ontario, Canada and 7 Department of General Pediatrics, Division of Metabolic and Endocrine Diseases, University Children’s Hospital, Heidelberg, Germany. Summary Glutaric acidemia type 1 (GA-1) is an autosomal recessive disorder characterized by a deficiency of glutaryl-CoA dehydrogenase (GCDH) activity. GA-1 is often associated with an acute encephalopathy between 6 and 18 months of age that causes striatal damage resulting in a severe dys- tonic movement disorder. Ten autopsy cases have been pre- viously described. Our goal is to understand the disorder better so that treatments can be designed. Therefore, we present the neuropathological features of six additional cases (8 months – 40 years), all North American aboriginals with the identical homozygous mutation. This cohort dis- plays similar pathological characteristics to those previ- ously described. Four had macroencephaly. All had striatal atrophy with severe loss of medium-sized neurons. We present several novel findings. This natural time course study allows us to conclude that neuron loss occurs shortly after the encephalopathical crisis and does not progress. In addition, we demonstrate mild loss of large striatal neur- ons, spongiform changes restricted to brainstem white matter and a mild lymphocytic infiltrate in the early stages. Reverse transcriptase-PCR to detect the GCDH mRNA revealed normal and truncated transcripts similar to those in fibroblasts. All brain regions demonstrated markedly elevated concentrations of GA (3770–21 200 nmol/g pro- tein) and 3-OH-GA (280–740 nmol/g protein), with no evid- ence of striatal specificity or age dependency. The role of organic acids as toxic agents and as osmolytes is discussed. The pathogenesis of selective neuronal loss cannot be explained on the basis of regional genetic and/or metabolic differences. A suitable animal model for GA-1 is needed. Keywords: autopsy; glutaric acid; 3-hydroxyglutaric acid; striatum; molecular genetics Abbreviations: ChAT = choline acetyltransferase; DAB = diaminobenzidine; H&E = haematoxylin and eosin; GA = glutaric acid; GA-1 = glutaric acidemia type 1; GCDH = glutaryl-CoA dehydrogenase; GFAP = glial fibrillary acidic protein; HLA-DR = human leucocyte antigen-DR; NMDA = N-methyl-D-aspartate; 3-OH-GA = 3 hydroxyglutaric acid; RT-PCR = Reverse transcription polymerase chain reaction Received July 14, 2004. Revised December 10, 2004. Accepted December 13, 2004. Advance Access publication February 2, 2005 Introduction Glutaric acidemia type 1 (GA-1) is an autosomal recessive disorder of amino acid metabolism caused by the deficiency of functional glutaryl-CoA dehydrogenase (GCDH) activity (Christensen, 1993), an essential enzyme in the catabolic path- ways of L-tryptophan, L-lysine and L-hydroxylysine (Goodman et al., 1977; Baric et al., 1998; Goodman and Frerman, 2001). Lack of functional GCDH activity typically leads to the accumulation of glutaric acid (GA), 3-hydroxyglutaric acid (3-OH-GA) and glutaconic acid in the blood, urine, CSF and brain tissue (Stokke et al., 1975; Goodman et al., 1977; Baric et al., 1998). GA and 3-OH-GA might induce an imbalance in glutamatergic and GABAergic neurotransmission (Wajner et al., 2004) and 3-OH-GA might act through excitotoxic NMDA receptors to produce a neurotoxic effect (Ko ¨lker # The Author (2005). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved. For Permissions, please email: [email protected]Downloaded from https://academic.oup.com/brain/article/128/4/711/284316 by guest on 10 February 2022
12
Embed
Neuropathological, biochemical and molecular findings in a glutaric
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Received July 14, 2004. Revised December 10, 2004. Accepted December 13, 2004. Advance Access publication
February 2, 2005
IntroductionGlutaric acidemia type 1 (GA-1) is an autosomal recessive
disorder of amino acid metabolism caused by the deficiency
of functional glutaryl-CoA dehydrogenase (GCDH) activity
(Christensen, 1993), an essential enzyme in the catabolic path-
ways of L-tryptophan, L-lysine and L-hydroxylysine (Goodman
et al., 1977; Baric et al., 1998; Goodman and Frerman, 2001).
Lack of functional GCDH activity typically leads to the
accumulation of glutaric acid (GA), 3-hydroxyglutaric acid
(3-OH-GA) and glutaconic acid in the blood, urine, CSF and
brain tissue (Stokke et al., 1975; Goodman et al., 1977; Baric
et al., 1998). GA and 3-OH-GA might induce an imbalance in
glutamatergic and GABAergic neurotransmission (Wajner
et al., 2004) and 3-OH-GA might act through excitotoxic
NMDA receptors to produce a neurotoxic effect (Kolker
# The Author (2005). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved. For Permissions, please email: [email protected]
Dow
nloaded from https://academ
ic.oup.com/brain/article/128/4/711/284316 by guest on 10 February 2022
et al., 2004a,b), although this is not supported by all experi-
mental data (Freudenberg et al., 2004; Lund et al., 2004). In
one autopsy case, 3-OH-GA was relatively more abundant in
the striatum (Kolker et al., 2003), where the neuronal damage
is most severe.
GA-1 affected children are clinically characterized by
macrocephaly appearing at, or shortly after, birth and initial
normal development interrupted by abrupt onset of dystonia
and choreoathetosis, which then remain relatively static.
Neurological abnormalities usually appear between 6 and
18 months of age, often in conjunction with a febrile illness.
Intellect seems to be relatively preserved (Goodman and
Frerman, 2001). The profound neurological sequelae may
lead to death in early childhood; however, some individuals
survive for many years. A minority of biochemically affected
individuals may remain asymptomatic or experience an insi-
dious onset of mild neurological abnormalities. The brains of
children affected with GA-1 exhibit wide Sylvian fissures and
enlarged frontal ventricles due to caudate atrophy. There are
only 10 published autopsy reports of GA-1 (Goodman et al.,
ration and rare perivascular cuffing (up to 5 cells thick) was
present in the putamen. The tail of the caudate adjacent to the
hippocampus appeared normal. Rare vacuoles were seen in
the white matter of a single gyrus and vacuoles were fairly
abundant in the central tegmental tract of the brainstem, but
not elsewhere.
Case 3This male’s parents were known heterozygotes for the GA-1
mutation. He is Case 12 in pedigree C described in the clinical
report of this cohort (Haworth et al., 1991). Following full-
term birth in 1989, he presented at 7 months of age with
developmental delay and relatively sudden onset of dystonia
and seizures. A CT scan showed hypointensity of the caud-
ate and putamen, and widening of the Sylvian fissures. He
became severely impaired, was treated with Phenobarbital
and required placement of a feeding tube. He was placed in
a chronic care institution. At 16 months, he developed fever
and died suddenly. Autopsy showed acute glottitis and pneu-
monitis. His head circumference was on the 22 percentile.
The external appearance of the brain was unremarkable. The
caudate and putamen were atrophic and the lateral ventricles
were mildly enlarged. Microscopically, the striatum exhibited
loss of medium-size neurons, plump reactive astrocytes,
reactive microglia, scattered calcospherites and rare CD3
immunoreactive lymphocytes in the caudate. The tail of the
caudate adjacent to the hippocampus appeared normal. There
were scattered pyknotic neurons in the cerebral cortex.
Case 4This male was born at 39 weeks by Caesarean section and was
found, on genetic screening in 1999, to have GA-1 (he is
Case 3 in Greenberg et al., 2002). His development was
delayed slightly. At 5.5 months of age, he developed fever
with onset of dystonia and athetoid limb movements as well
as seizure activity. A CT scan showed enlarged frontal horns
of the lateral ventricles and widened Sylvian fissures, but no
generalized atrophy (Fig. 1). Caudate atrophy was worse
at 10 months. He was treated with Phenobarbital and
Fig. 1 Photographs showing a CT scan of the head in thehorizontal plane obtained 8 months before death (upper) and acoronal slice of the brain (lower) from Case 4. Both exhibitflattening of the head of the caudate nucleus along the wall of thelateral ventricle (arrows) and mild enlargement of the lateralventricles. Note that the cortical thickness is essentially normal.
714 C. B. R. Funk et al.
Dow
nloaded from https://academ
ic.oup.com/brain/article/128/4/711/284316 by guest on 10 February 2022
topiramate, but failed to thrive and had multiple respiratory
infections. At 15 months, he was unable to sit but had some
head control and visual interaction. During a febrile illness at
18 months, he stopped breathing. Autopsy revealed laryngitis
and dehydration. His head circumference had been on the
52 percentile at birth, 89 percentile at 3 months, 63 percentile
at 8 months and was on the 50 percentile at the time of death.
The brain exhibited mild widening of temporal and frontal
sulci, and pronounced widening of the Sylvian fissures. No
histological abnormalities were apparent in the cerebral cor-
tex of the temporal lobe tips or frontal lobes. The caudate
nuclei were small, yellowish and firm. There was symmetric
lateral ventricle enlargement (Fig. 1). The head of the caudate
and the putamen exhibited severe neuronal loss with pro-
nounced reactive astrocytes. Only rare reactive microglia
were identified. The tail of the caudate adjacent to the hip-
pocampus appeared normal. The cerebral white matter exhib-
ited no vacuoles, no damaged axons were identified using
amyloid precursor protein immunohistochemistry, and no
myelin debris could be demonstrated with the Marchi
method. Rare vacuoles were present in the central tegmental
tract at the level of the midbrain.
Case 5This female born by forceps delivery at 37 weeks gestation,
in 1983, had generalized hypotonia in infancy. She is Case 1
in pedigree A described in the clinical report of this cohort
(Haworth et al., 1991). At 6 months of age, she developed
seizures and choreoathetoid movements. Thereafter she
developed severe spastic quadriparesis. She was unable to
walk and developed severe scoliosis. A CT scan at 5 years
showed enlarged frontal horns and slightly wide Sylvian
fissures. She died of acute pneumonia at age 7 years. Her
head circumference had been on the 25 percentile at birth,
52 percentile at 6 months, 3 percentile at 19 months and
25 percentile at the time of death. Her brain exhibited mild
gyral flattening, slightly enlarged ventricles and severe striatal
atrophy. The caudate and putamen had a near complete loss of
medium size neurons (Fig. 2) and some GFAP immunoreactive
astrocytes; the nucleus accumbens was spared. Scattered
hypertrophic astrocytes and very rare HLA-DR immuno-
reactive microglia were present in the internal capsule—
although there were no vacuoles. There were small foci of
spongiform change in the frontal and insular cortex.
Case 6This male born at 40 weeks gestation, in 1953, was apparently
slow to breathe. He had three siblings that died in infancy
from pneumonia. He is Case 14 in pedigree E described in
the clinical report of this cohort (Haworth et al., 1991). At
12 months of age, he developed a high fever and was admitted
to hospital with dystonia. At 6 years, he was documented
to have severe flexor spasticity of the arms and legs,
choreoathetoid movements and oral dyskinesia. Seizures
were reported only early in life. He resided in a nursing
care facility for his entire life. At no time was brain imaging
conducted and head circumference information was not
available. He required many hospitalizations for aspiration
pneumonia. On one such admission at age 40 years, the
treatment was complicated by pseudomembranous colitis
from which he died. His brain had a normal external appear-
ance. The frontal horns of the lateral ventricles were moder-
ately enlarged. There was marked atrophy of the caudate and
putamen with neuronal loss and abundant corpora amylacea.
There was no significant immunoreactivity for GFAP or
HLA-DR. The anterior nucleus accumbens was spared.
The white matter of the cerebrum and brainstem was well
myelinated and no vacuoles were identified. Subtle chronic
astrogliosis, demonstrable with modified phosphotungstic
acid haematoxylin (PTAH) stain, was evident in the inferior
olivary nuclei but there was no obvious neuron loss. Very rare
empty basket cells were present in the cerebellum, suggestive
of mild Purkinje cell loss. Frontal cortex tissue was examined
in all cases, none provided evidence of obvious cell loss or
reactive changes.
Comparative analysisQuantitative comparison of the striatum in GA 1 cases and
controls showed statistically significant (P < 0.05) loss of
Fig. 2 Photomicrographs of the striatum showing (A) normalneuron density (neurons indicated by arrows) in an age-matchedcontrol and (B) severe loss of neurons in the same area of GA-1Case 5. (H&E stained sections, 403 objective magnification,bar = 20 mm).
Neuropathology in GA-1 715
Dow
nloaded from https://academ
ic.oup.com/brain/article/128/4/711/284316 by guest on 10 February 2022
medium sized neurons in the dorsal caudate, ventral caudate,
dorsalputamenandventralputamen(Fig.3).Thedorsal regions
of the caudate and putamen were more severely affected,
although this was not statistically significant. There was no
age-dependent trend (Fig. 4) in the quantity of neurons in
the GA-1 cases, suggesting that maximal neuron loss had
occurred within 2 months of onset of symptoms, which was
the time of death after encephalopathical crisis in Case 1. Large
neurons, which are normally much less abundant than the
medium-sized neurons, were significantly fewer in the ventral
putamen of the GA1 patients with similar trends in all areas of
the striatum (Fig. 5). Immunostaining for GFAP (Figs 6 and 7)
demonstrated the presence of reactive astrocytes in all areas of
the striatum, with a tendency to greater staining in the dorsal
striatum. Analysis of reactive microglial activation (Fig. 8)
demonstrated significant HLA-DR immunoreactivity in
only the three youngest cases and mild infiltrate in the fourth
case, suggesting that it only persists a few months after the
acute episode. Overall, the differences approached statistical
significance (P = 0.0725 and 0.0699, respectively; paired t-test
versus age-matched controls) only in the dorsal and ventral
putamen.
Loss of GABA and calbindin immunoreactivity confirmed
that the population of medium sized neurons is decreased in
GA-1 (data not shown). The relative absence of NeuN label-
ling confirmed that neurons were lost and not simply atrophic.
Qualitative inspection of ChAT and tyrosine hydroxylase
immunoreactivity in large cholinergic and dopaminergic
neurons indicated little, if any, difference between GA-1
Fig. 3 Counts of medium sized neurons (mean 6 SE) in fourregions of the striatum. A significant loss of medium sized neuronswas seen in dorsal and ventral areas of the caudate and putamen;paired t-test versus age-matched controls, *P < 0.008.
Fig. 6 Photomicrographs showing scant immunoreactivityfor GFAP in dorsal caudate of a control case (A) and abundantGFAP-positive reactive astrocytes in GA-1 case 1(B). (DAB detection of anti-GFAP with haematoxylincounterstain; bar = 50 mm).
Fig. 4 Neuron density as a function of age. Note that themagnitude of neuron loss is similar regardless of the age of theGA-1 patient.
Fig. 5 Counts of large neurons (mean 6 SE) in the striatum.There was a marginally significant loss in the ventral caudate;paired t-test versus age-matched controls, P = 0.0422.
716 C. B. R. Funk et al.
Dow
nloaded from https://academ
ic.oup.com/brain/article/128/4/711/284316 by guest on 10 February 2022
cases and controls (data not shown). Synaptophysin immun-
oreactivity in striatum was not significantly different between
cases and controls (data not shown), indicating the preserva-
tion of input axons to the striatum.
The presence of GAPDH amplification product in both
GA-1 cases and controls indicated the presence of undegraded
mRNA despite long post-mortem delays to autopsy and the
lengthy interval between autopsies and molecular study.
However, there was limited frozen tissue stored and striatal
tissue was not recoverable. Thus, proper quantification of the
relative proportion of normal and mutant GCDH transcripts
was not possible. Nonetheless, we observed abundant mutant
and normal sized GCDH transcripts in the frontal cortex of
Cases 3 and 5 (the most abundant frozen tissue available)
(Fig. 9). A similar pattern has been seen in fibroblasts and
lymphoblasts of affected patients (Greenberg et al., 1995).
Organic acid analysis demonstrated marked elevations of
GA and 3-OH-GA compared with controls. There was no
evidence of striatal specificity or age dependency. There
was a slight elevation of GA in one control brain, which
likely can be considered a non-specific change related to
agonal events (Table 2).
Fig. 7 Proportionate area with DAB precipitate (measured by NIHimage analysis; mean 6 SE) when labelled with primary antibodyagainst GFAP; paired t-test versus age-matched controls*P < 0.016, **P < 0.005.
Fig. 8 Photomicrographs showing HLA-DR immunolabellingof reactive microglia in the putamen. The age matched controlcase (A) exhibits no cells while the GA-1 sample (Case 1, B)exhibits abundant activated cells. (DAB detection of anti-HLA-DR with haematoxylin counterstain, bar = 50 mm).
Fig. 9 PCR amplification products from frontal lobe samplesincubated without (–RT) and with reverse transcriptase ( + RT),respectively, are shown from Case 3 (lanes 1 and 2), Case 5 (lanes3 and 4), Control 1 (lanes 5 and 6) and Control 2 (lanes 7 and 8).The upper band in lanes 2, 4, 6 and 8 represents a normal sizedfragment of 393 bp and the lower band in lanes 2 and 4 representsthe truncated message 26 bp shorter than the normalsized-fragment. Lane 9 is the l00 bp ladder.
Table 2 Analysis of organic acids in brain tissue samplesfrom four GA-1 and three control cases