Site-specific somatic mitochondrial DNA point mutations in patients with thymidine phosphorylase deficiency (original) (raw)
COX in skin fibroblasts. COX activity was decreased in the skin fibroblast lines of MNGIE patients compared with control cells (P = 0.04) (Table 1). The skin fibroblasts with normal COX activity came from a mildly affected individual (MN1-4). In contrast to COX, activity of citrate synthase (a mitochondrial matrix enzyme and a measure of mitochondrial mass) was similar in the two groups (Table 1). When COX activity was normalized to citrate synthase, the defects of COX activity in patient skin fibroblasts were more evident (P = 0.02) (Table 1).
Mitochondrial enzyme activities in cultured skin fibroblasts from MNGIE patients and controls
Levels of dUrd in MNGIE patients. Plasma levels of dUrd were undetectable in the controls (n = 20) and heterozygous TP mutation carriers (n = 14) (i.e., concentrations were below 0.05 μM) but were markedly elevated in all patients analyzed (14.2 μM ± 4.4 μM, mean ± SD; range 5.5–24.4; n = 25).
mtDNA point mutations in MNGIE patients. Based upon our observation that TP-deficient skin fibroblasts have decreased COX activity (Table 1) but do not show depletion or multiple deletions of mtDNA by Southern blot analyses (data not shown), we suspected that mtDNA point mutations might be present in cells from MNGIE patients. We therefore sequenced the 22 mtDNA-encoded tRNA genes from cultured skin fibroblast DNA from six MNGIE patients (MN1-4, MN3-1, MN4-2, MN4-3, MN7-1, and MN7-2). We identified three T-to-C transitions: at nucleotide 4,370 in tRNAGln, at nucleotide 5,814 in tRNACys, and at nucleotide 15,956 in tRNAPro (16). Interestingly, all three mutations were preceded by three A residues (i.e., 5′-AAAT to 5′-AAAC).
In addition, we sequenced the entire mtDNA of skeletal muscle and cultured skin fibroblasts from two MNGIE patients (MN4-2 and MN7-2). To identify maternally inherited polymorphisms, we also sequenced peripheral blood leukocyte mtDNA from the mothers of both patients. We did not find any mutations in the skeletal muscle mtDNA of the patients. In contrast, we identified 22 additional mtDNA point mutations in cultured skin fibroblasts from the two MNGIE patients (Table 2). Interestingly, 18 of these point mutations (82%) (nucleotides 1,040, 2,233, 3,653, 4,028, 5,517, 5,628, 7,055, 7,440, 8,108, 8,367, 9,110, 9,758, 10,432, 11,324, 11,781, 13,879, 14,166, and 15,831) were T-to-C transitions. Fifteen of these 18 T-to-C transitions (83%) were preceded by at least one A residue (Table 2), and seven transitions (39%) were preceded by at least three A residues.
mtDNA point mutations in fibroblasts of MNGIE patients identified by sequencing
To quantitate the levels of the mtDNA mutations and to identify additional 5′-AAAT to 5′-AAAC transitions, RFLP analyses were performed on DNA extracted from six cultured skin fibroblast lines, 13 peripheral blood leukocyte samples, and four autopsy tissue samples from MNGIE patients. Of the 203 5′-AAAT sites in mtDNA, eight that were amenable to RFLP analysis were screened along with one 5′-AAT site. We found six additional heteroplasmic mutations in skin fibroblast mtDNA: four were 5′-AAAT to 5′-AAAC mutations located at nucleotide 616 in tRNAPhe, at nucleotide 3,386 in subunit 1 of nicotinamide adenine dinucleotide dehydrogenase-1, at nucleotide 10,221 in nicotinamide adenine dinucleotide dehydrogenase-3, and at nucleotide 12,310 in tRNALeu(CUN). The fifth mutation was a 5′-AAT to 5′-AAC transition at nucleotide 16,172 in the displacement loop (D-loop). The sixth mutation we identified by RFLP analysis was a heteroplasmic 5′-AAAT to 5′-AAAA mutation at nucleotide 5,628 in tRNAAla of skin fibroblast, peripheral blood leukocyte, and tissue mtDNA from six patients (Table 3). T5628C was identified as a pathogenic mutation in a family with maternally inherited progressive external ophthalmoplegia (17).
mtDNA point mutations in MNGIE patients screened by RFLP analysis
Overall, of the nine mutations screened by RFLP analyses, seven (nucleotides 3,386, 4,370, 5,628, 5,814, 10,221, 15,956, and 16,172) were identified in at least half of the patients’ skin fibroblasts, and three (nucleotides 5,814, 10,221, and 15,956) were present in all six cell lines, with heteroplasmic levels ranging from below 2% to 81% (Table 3). In addition, five mutations (nucleotides 3,386, 5,814, 10,221, 15,956, and 16,172) were detected by RFLP analysis in leukocytes and in autopsy tissues from patients (MN2-1, MN3-1, MN4-2, and MN7-2), with levels of heteroplasmy at or below 63% (Table 3 and Table 4). Remarkably, all tissue and cell samples from the 13 patients tested contained the nucleotide 5,814 mutation. T5814C was identified as a pathogenic mutation (A5814G in the coding heavy strand) in patients with mitochondrial encephalopathy (18, 19) or mitochondrial myopathy and cardiomyopathy (20).
mtDNA point mutations in autopsy tissues of MNGIE patients screened by RFLP analysis
All of the mutations were absent in 33 control samples from 22 individuals (six cultured skin fibroblast lines, 11 brains, four small intestines, six kidneys, and six livers) and in unaffected carriers of TP mutations (23 peripheral blood leukocyte samples) (data not shown).
To determine whether analogous mutations were present in nuclear DNA from MNGIE patients, we performed RFLP screening analyses of seven genomic 5′-AAAT sites, one in the TP gene itself, three in the 18S ribosomal RNA gene (21), and three in the human β-actin gene (22), using total DNA from six skin fibroblast lines, 13 peripheral blood leukocyte samples, and 32 tissues obtained from 13 MNGIE patients. No mutations were detected (data not shown).
mtDNA point mutations in MNGIE patient HVSs detected by sequencing of PCR clones. To screen for mutations at non–5′-AAAT sites, we sequenced HVS 1 and HVS 2 in the mtDNA D-loop, which are considered to be hot spots for mutations (13–15). We subcloned PCR-amplified HVSs of cultured skin fibroblast mtDNA from two patients (MN4-3 and MN7-1) and two controls, and sequenced between 84 and 92 clones of each PCR fragment. We identified four additional 5′-AAAT to 5′-AAAC heteroplasmic mutations at nucleotides 291, 292, 16,043, and 16,352 (Table 5). In addition, we identified four other mutations that were not 5′-AAAT to 5′-AAAC: 5′-GGAGT to 5′-GGAGC at nucleotide 16,247, 5′-AAAT to 5′-AAAA at nucleotide 291, 5′-AAATT to 5′-AAAAA at nucleotides 291–290, and deletion of a T at 5′-AAATTTTTT (Table 5). Four 5′-AAAT sites (at nucleotides 16,094–16,091, 16,137–16,140, 16,325–16,322, and 159–156) in HVS 1 and HVS 2 did not show mutations. DNA from two control skin fibroblast lines did not show any mutations in the D-loop.
mtDNA mutations in MNGIE patients’ HVSs