Acute intermittent porphyria (AIP), due to half-normal hydroxymethylbilane synthase activity,is characterized by acute life-threatening neurologic attacks whose etiology remains unclear. To date, only 3 patients confirmed to have homozygous dominant AIP (HD-AIP) have been described (hydroxymethylbilane synthase genotypes R167Q/R167Q and R167W/R173Q).
To investigate the genetic, biochemical, clinical, and neuroradiologic features of a severely affected infant with HD-AIP.
Clinical, imaging, and genotype/phenotype studies were performed.
The proband, homoallelic for hydroxymethylbilane synthase mutation R167W, had approximately 1% of normal hydroxymethylbilane synthase activity, elevated porphyrins and porphyrin precursors, severe psychomotor delay, and central and peripheral neurologic manifestations. When expressed in vitro, the R167W mutant enzyme had less than 2% of normal activity but was markedly unstable, consistent with the proband’s severe phenotype. Mitochondrial respiratory chain enzymes were normal. Neuroradiologic studies revealed a unique pattern of deep cerebral white matter injury, with relative preservation of the corpus callosum, anterior limb of the internal capsule, cerebral gray matter, and infratentorial structures.
This severely affected patient with HD-AIP expanded the phenotypic spectrum of HD-AIP. His brain magnetic resonance imaging studies suggested selective cerebral oligodendrocyte postnatal involvement in HD-AIP, whereas most structures developed prenatally were intact. These findings indicate that the neurologic manifestations result from porphyrin precursor toxicity rather than heme deficiency and suggest that porphyrin precursor toxicity is primarily responsible for the acute neurologic attacks in heterozygous AIP and other porphyrias.
Acute intermittent porphyria (AIP), an autosomal dominant inborn error of heme biosynthesis, is characterized by life-threatening, acute neurologic attacks.1 Affected heterozygotes have half-normal hydroxymethylbilane synthase (HMBS) activity and accumulate the porphyrin precursors δ-aminolevulinic acid (ALA) and porphobilinogen (PBG), particularly during acute neurologic attacks.1- 3
The acute attacks are precipitated by metabolic, hormonal, and environmental factors that induce hepatic 5-aminolevulinate synthase (ALAS1) activity.1,2 With increased ALAS1 activity, porphyrin precursor (ALA and PBG) levels increase, and the half-normal hepatic HMBS activity in heterozygous patients is apparently insufficient to prevent the pathological precursor accumulation, leading to the neurologic symptoms. Acute attacks are characterized by severe abdominal pain, vomiting, constipation, hypertension, tachycardia, and bladder dysfunction, presumably due to an autonomic neuropathy. Motor weakness and sensory involvement presumably result from the peripheral axonal neuropathy, and mental symptoms do not have clear cerebral morphological findings.1 Although the pathogenesis of the neurologic manifestations is poorly understood, the 2 leading hypotheses are heme/hemoprotein deficiency in nerve cells and the neurotoxicity of ALA, PBG, and/or other porphyrin metabolites.4 It has been suggested that ALA, a γ-aminobutyric acid (GABA), glutamate, and aspartate analogue, may be neurotoxic by interaction with GABA receptors and/or inhibition of glutamate uptake.1,4 Of note, 0.1 mg/dL of ALA inhibits 50% of GABA binding to synaptic membranes.5
Whereas heterozygous AIP is the most common hepatic porphyria, with approximately 1 in 20 000 affected in the United States and Europe,1 homozygous dominant AIP (HD-AIP) is extremely rare, having been documented at the biochemical/genetic level in only 3 patients, a Dutch child6,7 and 2 English siblings.8 A putative fourth patient was described,9 but no enzymatic or molecular diagnostic data were reported (Table 1).
Here, we describe a more severely affected Spanish patient who further delineates the HD-AIP phenotype. Expression and characterization of the HMBS mutations causing HD-AIP in the proband and previous patients permitted preliminary genotype-phenotype correlations. Moreover, the unique neurologic and neuroradiologic findings in these patients provided insight into the pathogenesis of HD-AIP and the acute neurologic attacks in heterozygous AIP.
Blood samples were collected from the family members with HD-AIP after obtaining informed consent. Lymphoblasts were established by standard techniques. Erythrocyte ALA-dehydratase activity and erythrocyte protoporphyrin and urinary ALA and PBG concentrations were determined with commercial kits (Biosystems SA, Barcelona, Spain). Urinary porphyrins and erythrocyte HMBS activities were determined.10,11 Hydroxymethylbilane synthase mutation identification, expression, and characterization were performed as described.12,13 To determine if the R167W mutation caused abnormal splicing, the lymphoblast HMBS heterogeneous nuclear RNA was isolated, reverse transcribed, and PCR amplified, and the transcripts were analyzed.14
Magnetic resonance imaging (MRI) was performed at 0.5 Tesla using sagittal T1-weighted, axial proton density-weighted, and axial T2-weighted pulse sequences.
The male proband was the third child of asymptomatic heterozygous AIP Spanish parents who were second cousins. His parents were healthy and had no history suggestive of AIP. An older brother had symptoms similar to those of the proband at 2 months of age and died at 27 months of age with the diagnosis of cerebral palsy. A 5-year-old sister was healthy. At birth, the proband weighed 3.4 kg, appeared normal, but had dark-orange urine. At 3 months of age, psychomotor retardation was evident; he was unable to hold up his head or follow objects. He gained head control at 8 months of age and sat without support at 12 months of age. He never pulled to standing or walked unassisted. He had no history of seizures. Surgery was performed at age 9 months for capsular cataracts. Metabolic studies and cranial MRI and computed tomographic scans were initially reported as unremarkable at 10 months of age.
When referred at 13 months of age, the proband was 76 cm tall (10th-25th percentile), weighed 9.2 kg (10th percentile), and had a fronto-occipital circumference of 45.5 cm (10th-25th percentile). He had coarse facial features with a long philtrum and protruding tongue, bilateral capsular cataracts, right palpebral ptosis, dorsal kyphosis, and coarse wiry hair. He had mild palpable hepatosplenomegaly. Neurologic examination confirmed psychomotor delay and revealed occasional dystonic head positioning and dystonic arm movements, normal appendicular muscle tone and strength, but axial hypotonia, reduced tendon reflexes, and bilateral Babinski responses. Pupillary response to light was normal; left mydriasis and bilateral cataracts were noted. The patient did not fix or follow. Results from a funduscopic examination were normal.
Dark-orange urine was noted, and elevated ALA and PBG suggested a homozygous porphyria.1 Deficient erythrocyte HMBS activity and half-normal levels in his consanguineous parents confirmed the diagnosis of HD-AIP (Table 2).
Follow-up evaluation at 2.3 years of age revealed progressive developmental deterioration. The patient died at 3.3 years of age of sudden death while sleeping.
Results from an electroencephalogram at 13 months of age were normal. Sensory nerve conduction velocities were absent in ulnar, sural, and plantar nerves, while peroneal motor conduction velocity was decreased (24 m/s; normal, >40 m/s), compatible with a mixed neuropathy with greater sensory involvement. Open nerve and muscle biopsies showed significantly decreased numbers of myelinated fibers in the sural nerve and atrophy of type I and II fibers in the quadriceps muscle, suggestive of denervation. Electromyographic studies of the tibialis anterior demonstrated fibrillations and sharp waves, compatible with a neurogenic process. Brainstem auditory evoked potentials showed increased latencies in waveforms PIII (pons; right, 4.8 milliseconds; left, 4.8 milliseconds; normal, <4.09 milliseconds), PIV (lateral lemnisci; right, 6.1 milliseconds; left, 6.0 milliseconds; normal, <5.03 milliseconds), and PV (midbrain; right, 7.2 milliseconds; left, 7.2 milliseconds; normal, <6.08 milliseconds), suggesting brainstem involvement. Visual evoked potentials showed bilateral delay (P100 latency; right, 124 milliseconds; left, 138 milliseconds; normal, 100 milliseconds). Despite these increased latencies, both sets of evoked potentials demonstrated preserved wave amplitudes. The electroretinogram showed reduced amplitude and prolongation of latencies.
Noncontrast brain MRI at 13 months of age revealed symmetrical prominence of the ventricles and cisterns. T2-weighted MRI showed symmetrical signal increase within the cerebral hemispheres and external/extreme capsules, reduced caliber of the posterior corpus callosum (with no signal change), and a symmetrical delay in myelination comparable to those at 8 months of age. There was striking symmetrical sparing and (nearly) normal signal intensity in the brainstem tegmentum, cerebellum, cerebral peduncles, optic chiasm/tracts, anterior limb of the internal capsule, fornix, anterior corpus callosum, cingulum, fronto-occipital fasciculus, subcortical arcuate U fibers, basal ganglia, thalami, and cortical gray matter. There was no white matter vacuolation/cavitation.
Noncontrast T2-weighted MRI at 28 months of age (Figure 1) showed increased cerebrospinal fluid spaces, greater loss of cerebral volume, reduced size of the posterior corpus callosum, and prominent diffuse signal increase in the cerebral white matter. Limited additional myelination met criteria for 8 to 10 months of age, indicating greater retardation from expected myelination. The antero-inferior temporal lobe and frontal deep white matter showed relative sparing. The striking sparing of the other structures persisted. However, small vacuoles/cavities now aligned along the bundles of periventricular white matter.
Noncontrast magnetic resonance images at age 28 months. T2-weighted magnetic resonance images in the axial (A-C) and coronal (D) planes show the normal cerebellum and brainstem, normally myelinated corpus callosum, prominent involvement of the white matter of the external/extreme capsules and cerebral hemispheres, relative sparing of the white matter of the anterior-inferior temporal lobes, and sharply marginated periventricular foci of vacuolation/cavitation oriented along the fiber bundles.
At 13 months of age, blood counts (Table 1), reticulocytes, erythrocyte sedimentation rate (16 mm/h), and routine serum chemistries were normal. Serum lactic acid concentration was elevated (45.4 mg/dL [5 mmol/L]; normal, 9-16 mg/dL [1-1.8 mmol/L]), as was the serum pyruvic acid (2.5 mg/dL [284 μmol/L]; normal, 0.36-0.9 mg/dL [40.9-102.2 μmol/L]). Serum creatinine phosphokinase (232 U/L; normal, 24-195 U/L) and aldolase (11.4 U/L; normal, 0-7.6 U/L) activities were slightly increased, while serum carnitine, vitamins A, E, and B12, and folic acid levels were normal. Serum copper and ceruloplasmin, serum and urinary amino acids, mitochondrial respiratory chain enzymes in muscle, leukocyte lysosomal enzyme activities, and serum phytanic acid and long chain fatty acids were normal, as were results from hepatic and renal function tests. Cerebrospinal fluid analyses revealed essentially normal glutamic acid (0.04 mg/dL [2.7 μmol/L]; normal, 0.01-0.03 mg/dL [0.5-1.8 μmol/L]), tryptophan (0.37 mg/dL [1.8 μmol/dL]; normal, ≥0.2- 1.2 mg/dL [≥1-6 μmol/dL]), and albumin (0.19 g/dL [1.9 g/L]; normal, 0.07-0.16 g/dL [0.7-1.6 g/L]), and elevated lactic acid (351 mg/dL [39 mmol/L]; normal, 110-210 mg/dL [12.2-23.3 mmol/L]) and immunoglobulin G (28 mg/L; normal, 3-15 mg/L) with no cells present. Urinary urobilinogen was 2 mg/dL, and urinary bilirubin was moderately increased.
The proband had approximately 1% of normal mean erythrocyte HMBS activity, while his parents had approximately half-normal activities (Table 2). The proband’s urinary ALA, PBG, uroporphyrin I and III and coproporphyrin I, and erythrocyte protoporphyrin were markedly elevated (Table 2), whereas his asymptomatic parents’ ALA and PBG were normal.1
The proband was homoallelic for a C to T transition in codon 167 of exon 10, predicting an arginine to tryptophan substitution (designated R167W). Both parents were heterozygous for this mutation. Because this mutation altered the first base of exon 10, studies were directed to determine if mutation altered normal splicing. Comparison of the reverse transcription–polymerase chain reaction products from the proband and normal individuals revealed a single 295–base pair band, indicating the absence of aberrant splicing.
To compare HMBS mutant enzymes causing HD-AIP, expression vectors for R167Q, R167W, R173Q, and R173W were constructed and expressed in Escherichia coli. All 4 mutant enzymes had less than 2% of expressed normal activity (Table 3) and were relatively stable, except the R167W enzyme, which was markedly unstable (Figure 2), consistent with the proband’s more severe phenotype.
Thermostability of hydroxymethylbilane synthase (HMBS) activity expressed in Escherichia coli by pKK-HMBS normal and mutant constructs. Results are expressed as percentage of initial activity based on mean of 3 independent assays as described.12,13
The Spanish proband and the 3 previously described patients with enzyme/mutation-confirmed HD-AIP6,8 had a chronic, progressive neurodegenerative disease, although the presence, onset, and severity of neurologic symptoms varied (Table 1). The Spanish proband was more severely affected, with an earlier clinical onset, more rapidly progressive neurodegenerative course, and earlier demise at 40 months of age (Table 1). The proband and previously reported patients had elevated urinary ALA and PBG, erythrocyte protoporphyrin, and markedly reduced erythrocyte HMBS activity (<2% of normal mean) (Table 1). In the Spanish proband, PBG, but not ALA, was elevated in the cerebrospinal fluid, which may reflect the greater tissue uptake and/or solubility of ALA.
The HMBS mutations in the Dutch (R167W/R173Q),7 English (R167W/R167Q),8 and Spanish (R167W/R167W) patients occurred in exon 10 at CpG dinucleotides, known as mutational hotspots, and altered highly conserved arginines in the enzyme’s active site, which interact with PBG and the acidic side chains of the enzyme’s dipyrromethane cofactor.15 Expression studies revealed that all the mutants had less than 2% of mean normal activity; however, the R167W enzyme was markedly more heat labile (Table 3). Thus, the markedly decreased activity and stability of the R167W enzyme presumably accounted for the Spanish proband’s more severe phenotype.
The proband’s neurologic impairment included a peripheral neuropathy, progressive loss of cerebral volume, delayed myelination, a strikingly abnormal pattern of myelin signal intensities, and, later, multifocal vacuolation/cavitation of the periventricular white matter. Of note, the brain MRI of another patient with HD-AIP had a very similar pattern (R.J.D., oral communication, September 1998). The MRI changes in HD-AIP suggest a primary process affecting cerebral myelination, with neuronal/axonal sparing. Notably, there were no MRI lesions of the gray matter architecture, including the cerebral cortex, cerebellar cortex, or subcortical gray matter structures at 13 or 28 months of age or at 6 years of age in another patient with HD-AIP. Instead, there was a unique pattern of white matter involvement. The pathological process spared many tracts that myelinate prenatally or shortly after birth, including those in the brain stem and cerebellum, but affected tracts that myelinate in the later postnatal period.16 Furthermore, the selective white matter damage was associated with arrest of myelin maturation at the 8- to 10-month milestones and later with progressive vacuolation/cavitation in the periventricular white matter. Thus, later-made myelin either was not formed normally or was subject to early and progressive injury. However, time of myelination cannot be the sole factor because certain long association tracts (eg, fronto-occipital fasciculus) were spared although they myelinate late. The relative preservation of the corpus callosum and long descending tracts also suggests a process that affects primary myelin or myelinating cells.
Of the 2 major hypotheses for the acute neurologic attacks, toxic injury due to accumulated ALA, PBG, and/or other porphyrin metabolites may be relevant to the pathogenesis of HD-AIP. Based on the MRI studies described here, ALA-mediated neurotoxicity appears to be the most likely mechanism causing the pathologic features and clinical manifestations of HD-AIP. In addition to the distribution of cerebral white matter abnormalities, which affected regions particularly vulnerable to hypoxia-ischemia and excitotoxicity (eg, periventricular white matter), immature oligodendrocytes are highly susceptible to glutamate-dependent and other excitotoxic processes.17 The homology of ALA with glutamate and aspartate could selectively affect white matter regions that mature postnatally, predominantly those in which flow and oxygenation are at a borderline level. The molecular relationship of ALA with GABA also cannot be excluded since oligodendrocytes express GABA receptors18 and GABA is known to increase intracellular calcium concentrations in oligodendrocyte precursors.19 The fetal brain may be protected from the potential toxicity/injury of the soluble/diffusible porphyrin precursors because circulating and cellular ALA, PBG, and other metabolites may cross the placenta to be excreted in the mother’s urine. Postnatally, the persistently elevated concentrations of the porphyrin precursors and/or metabolites may be neurotoxic to myelination and/or to myelinating cells.
If heme deficiency was important in the pathogenesis of HD-AIP, then it would be expected to cause early neural pathologic features and clinical manifestations, perhaps similar to those in mitochondrial disorders that cause acute hepatic encephalopathies. However, the white matter abnormalities seen in the patients with HD-AIP differ from those reported in mitochondrial disorders.20 In addition, the gray matter structures did not appear physically affected in the patients with HD-AIP, indicating that the mitochondrial electron transport hemoproteins were not functionally impaired in neurons or oligodendrocytes. Also, the muscle mitochondrial respiratory chain enzymes from the proband and the British patients with HD-AIP8 were normal. Moreover, the early-onset, extraneural clinical manifestations characteristic of mitochondrial respiratory chain defects were not present in the proband or the previously reported patients with HD-AIP.6- 8
In sum, these findings suggest that the unique neuropathologic features of HD-AIP result from postnatal toxic injury, primarily of white matter structures, presumably due to the persistently elevated levels of ALA, PBG, and/or other porphyrin precursors or porphyrin metabolic products. In patients with heterozygous AIP, the overproduction of ALA and PBG may be responsible for the acute neurologic attacks, analogous to the acute hepatic encephalopathies due to various hepatic-specific inborn errors of metabolism. Of direct relevance, an allogeneic liver transplant in a 19-year-old female AIP heterozygote normalized her urinary ALA and PBG levels in 24 hours and completely eliminated her chronic neurologic attacks (37 in 29 months pretransplant) for more than 18 months posttransplant.21 These findings argue strongly that porphyrin precursor toxicity, and not neural heme deficiency, is responsible for the acute neurologic attacks in the heterozygous hepatic porphyrias, and they support our neuroradiologic findings indicating that postnatal porphyrin precursor toxicity is responsible for the neurologic manifestations in HD-AIP.
Correspondence: Robert J. Desnick, PhD, MD, Department of Human Genetics, Box 1498, Mount Sinai School of Medicine, Fifth Avenue and 100th Street, New York, NY 10029 (email@example.com).
Accepted for Publication: February 4, 2004.
Author Contributions:Study concept and design: Kaufmann and Desnick. Acquisition of data: Solis, Martinez-Bermejo, Astrin, Bishop, and Desnick. Analysis and interpretation of data: Solis, Martinez-Bermejo, Naidich, Kaufmann, Bishop, and Desnick. Drafting of the manuscript: Astrin and Desnick. Critical revision of the manuscript for important intellectual content: Solis, Martinez-Bermejo, Naidich, Kaufmann, Astrin, Bishop, and Desnick. Obtained funding: Desnick. Administrative, technical, and material support: Solis, Martinez-Bermejo, Astrin, and Desnick. Study supervision: Solis, Martinez-Bermejo, Bishop, and Desnick.
Acknowledgment: We thank Jose A. Garcia Penas, MD, for referring the proband; Jose M. Abelairas, MD, for expert ophthalmologic evaluation; Maria J. Becedas, BS, Weiming Xu, PhD, and Gregory Young, BS, for technical assistance; and C. Warren Olanow, MD, Premysl Ponka, MD, PhD, and Hugo M. Moser, MD, for thoughtful discussions. This study was supported in part by grant 5 RO1 DK26824 (Dr Desnick) from the National Institutes of Health, Bethesda, Md, and 5 MO1 RR00071 from the National Center for Research Resources for the Mount Sinai General Clinical Research Center, New York, NY.
Solis C, Martinez-Bermejo A, Naidich TP, Kaufmann WE, Astrin KH, Bishop DF, Desnick RJ. Acute Intermittent PorphyriaStudies of the Severe Homozygous Dominant Disease Provides Insights Into the Neurologic Attacks in Acute Porphyrias. Arch Neurol. 2004;61(11):1764-1770. doi:10.1001/archneur.61.11.1764