| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Special Articles |
Departments of Pediatrics (M.L.S., D.L.M., R.H.H.), Medicine (D.L., K-Y.N.), and Neuroscience, (X-Y.H., R.H.H.), University of California, San Diego, California 92093; and Department of Molecular and Medical Genetics, Oregon Health Sciences University, (N.G.K.), Portland, Oregon 97201
Address all correspondence and requests for reprints to: Richard H. Haas, Department of Neurosciences, Division of Pediatric Neurology, University of California, San Diego, 9500 Gilman Drive, Department 0935, La Jolla, California 92093-0935.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The diagnosis in patients with the A3243G mutation may be missed initially because of their mild diabetic symptoms, or because they do not fit the severe clinical picture of MELAS. Most patients presenting with diabetes exhibit few of the other symptoms associated with MELAS, such as muscle weakness, opthalmoplegia, or stroke-like episodes. The most common feature seen is nerve deafness, which may occur at an earlier age than symptoms of diabetes (8, 10, 13). There are important genetic implications for other family members when a mitochondrial disease is diagnosed. For these reasons, the demonstration of a specific mitochondrial mutation by DNA testing is particularly important in establishing a definitive diagnosis. Such testing must be performed in a very sensitive manner to reliably detect low levels of this mutation.
The number of mitochondria in each cell ranges from one to several hundred, and mitochondria segregate randomly with each cell division. This results in different percentages of mutant and normal mitochondria in different tissues and a mixture of normal and mutant mtDNA within the same cytoplasm. This is known as heteroplasmy. Because of heteroplasmy, the best test sample is from the affected tissue. In patients with diabetes this is the pancreas, which cannot be biopsied easily. Diagnostic and screening tests are most frequently carried out on cells obtained from blood. The A3243G mutation is present in these cells, but generally at a much lower level than that seen in the muscle tissue of patients with MELAS (14). Several recent surveys of diabetic populations (10, 25, 26) have continued to use nonradioactive methodology for screening despite the inability of such methods to detect mutational loads of 5–10% or lower. The mutation levels in blood samples from many patients manifesting only diabetic symptoms will be in this range (14, 27).
In the case reported here, the mother of a MELAS patient exhibited only isolated IDDM. The standard PCR assay coupled to restriction digestion and ethidium bromide staining did not demonstrate the mtDNA A3243G mutation in her blood or cultured lymphoblasts, but readily demonstrated the mutation in samples from her son who exhibited mild symptoms of MELAS and IDDM. To detect the suspected mutation in the mother’s blood, a radiolabeled PCR assay was necessary. To determine further the conditions necessary for the reliable detection of mutations present at low levels, the variables of the assay were investigated, and the percentage of mutant DNA in various tissues quantitated.
As demonstrated by this family, screening or diagnostic methods used to identify the A3243G mutation or other mtDNA mutations as a cause of diabetes must use highly sensitive methods such as radiolabeled PCR. The methods used in many current reports will have missed cases such as ours.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
KS, the proband, presented at the age of 4 yr with a history of worsening exercise tolerance because of muscle weakness. The birth history, neonatal course, and development in early infancy were normal, apart from a mild gross motor developmental delay. He began walking at 16 months, but he had difficulty with steps and was not able to run. A muscle biopsy showed ragged red fiber myopathy. Electrocardiogram (EKG) showed bifasicular heart block. A brain magnetic resonance imaging scan was normal. Also, at age 4 yr urinary frequency and nocturia lead to a diagnosis of IDDM. Two younger twin siblings were thought to be normal.
When evaluated at age 6–1/2 yr in the General Clinical Research Center at the University of California San Diego, this boy presented a picture of a moderately severe proximal and axial myopathy without central nervous system abnormality. Deep tendon reflexes were 2/4 throughout. The Gower’s sign was positive. Intellectually he was normal with normal language skills. Fundoscopic examination was normal. He had short stature. Weight was 16.8 kg (5th percentile), height 106 cm (<< 5th percentile). Bone age was delayed 18 months. EKG, 2-D cardiac echo, electroencephalogram, and brainstem auditory-evoked responses were normal.
Blood creatine phosphokinase was elevated at 462 IU/L (normal, <175 IU/L). Plasma lactate was 93.1 mg/dL (normal, 6.5–19.3 mg/dL), and simultaneous pyruvate was 1.6 mg/dL (normal, 0.3–0.7 mg/dL) producing a lactate/pyruvate ratio of 58 (normal, <20). Cerebrospinal fluid lactate was elevated at 36 mg/dL (normal, < 20 mg/dL). Plasma bicarbonate was 17 mmol/L. Liver function tests and the complete blood count were normal. Urine organic acid analysis revealed lactic aciduria (6190 mmol/mol creatinine; normal, 0–28) and generalized dicarboxylic aciduria, including two unidentified 3-hydroxydicarboxylic acids, consistent with a defect in fatty acid oxidation. Urine carnitine assay revealed an increased total carnitine excretion with an increase in the esterified fraction (total, 107.8 mmol/L; normal, 26–66) (esters, 75.7 mmol/L; normal, 0–22). A 1.75-g/kg oral glucose load produced a rise in plasma lactate from a baseline level of 49 mg/dL to 83 after 60 min; pyruvate increased from 1.2 to 1.9 mg/dL in the same period. An alanine load (0.4 g/kg) produced a rise in plasma lactate from 56 to 78 mg/dL, with pyruvate increasing from 1.5 to 2.4 mg/dL. An open muscle biopsy under chloral hydrate sedation was carried out allowing muscle electron transport assays and mtDNA studies to be performed.
Treatment was started with coenzyme Q10 20 mg/day, L-carnitine 100 mg/kg daily, ascorbate 750 mg/day, and sodium succinate 1500 mg/day. The sodium succinate was discontinued after 6 months. At age 8 yr 10 months, a glucose load revealed a baseline lactic acid of 15.8 mg/dL, 27.3 mg/dL at 60 min, and 27 mg/dL at 90 min. The creatine phosphokinase was 139, bicarbonate 27 mmol/L, and all other chemistries were normal. An EKG and 2-D cardiac echo were normal. A repeat brain magnetic resonance imaging scan was normal. Currently, at age 10 yr 6 months, exercise tolerance is improved. There has been no clinical progression of his symptoms. His diabetes remains well controlled on a moderate insulin dose of 0.75 U/kg daily. His hemoglobin A1C has remained excellent, between 6.5–7.9 mg/dL (normal nondiabetic is 3.3–6.1 mg/dL).
SS, the mother of the proband, developed IDDM during a pregnancy with twins, 5 yr after the birth of KS. She has remained insulin dependent. Apart from the diabetes, she is asymptomatic with no clinical evidence of muscle weakness or central nervous system disease. She has no family history indicative of mitochondrial disease. A needle biopsy of the quadriceps allowed histochemical and electron microscopy studies as well as mtDNA analysis. The two younger twin brothers of KS have no clinical indication of mitochondrial disease but have not undergone detailed study.
Electron transport assays
Spectrophotometric electron transport complex assays were carried out as previously described (15). Each assay was carried out at least twice on each sample using differing protein concentrations to establish linearity of the assay. Protein was assayed by the method of Lowry et al. (16). Citrate synthetase was assayed in Triton X-100-solubilized mitochondria. Complex I was assayed as NADH-CoQ oxidoreductase. Apparent first-order rate constants were calculated for complex III and IV from the rates determined at multiple time points and at multiple protein concentrations. Complex III (ubiquinone:cytochrome c oxidoreductase) was assayed with dihydroubiquinone-2 as the electron donor (17) by measuring the appearance of reduced cytochrome c. Complex IV was assayed as the oxidation of reduced cytochrome c.
Muscle mitochondrial preparation
Muscle was stored in liquid nitrogen after homogenization in a cryoprotective buffer. From 300–600 mg fresh muscle was collected in saline at 4 C. Within 15 min the sample was diced at 4 C and suspended in 2 vol homogenization buffer (250 mM sucrose, 20% dimethylsulfoxide). Homogenization was carried out with a Thomas’s tissue tearor (Biospec Products Inc., Bartlesville, OK). The homogenate was flash frozen in liquid nitrogen by sealing it in a small plastic bag, flattening the bag to a thin layer, and immersing it in liquid nitrogen. Subsequent mitochondrial isolation was by a modification of the technique of Hatefi et al. (18) using protease (nagarse type XXVII, Sigma P-4789, Sigma Chemical Comp., St. Louis, MO) digestion in 50 mM Tris buffer. Low-speed centrifugation at 800 x g followed by centrifugation of the supernatant at 8000 x g produced a mitochondrial pellet that was washed once in 0.25 M sucrose, recentrifuged, suspended in 0.25 M sucrose, and stored in liquid nitrogen.
Isolation of DNA
As necessary, DNA was quantitated by spectroscopy at 260/280 nm or by a fluorescent assay using Hoechst dye (19).
Blood and lymphoblasts. One to five milliliters of venous blood was collected into acid-citrate dextose anticoagulant and frozen at -70 C for at least 1 h before further processing. Subsequent thawing at room temperature lysed the red cells. This lysate was centrifuged for 10 min at 1500 x g. The supernatant was discarded, and the resulting cell pellet was resuspended with 1 ml fractionation buffer (0.5 M NaCl, 0.025 M EDTA, pH 8.0). This cell suspension was centrifuged for 5 min at 12,000 x g. The supernatant was discarded, and the cell pellet was used for mtDNA extraction using the Wizard Miniprep Kit (Promega, Madison, WI). Epstein Barr virus-transformed lymphoblasts were prepared by infecting peripheral blood lymphocytes with Epstein Barr virus and grown by standard methods (20). Cells were pelleted, and DNA was extracted using the Wizard Miniprep Kit.
Muscle and muscle mitochondria. Muscle tissue, 50–100 mg, snap frozen and stored in liquid nitrogen, was diced with a scalpel blade and homogenized using a sterilized glass/glass microgrinder in 185 µL resuspension buffer (Wizard; 50 mM Tris, pH 7.5; 10 mM EDTA; 100 µg/mL RNase A) with 15 µL Proteinase K (20 mg/mL) and 25 µL 10% SDS. This homogenate was digested at 54 C overnight, after which 200 µL cell lysis solution (Wizard; 0.2 M NaOH, 1% SDS) was added, and the Wizard extraction completed. Purified mitochondria, prepared as described, were pelleted in a microcentrifuge at 12,000 x g and mtDNA prepared by the standard Wizard DNA isolation procedure.
Clones containing normal and mutant mitochondrial sequences
The DNA fragments amplified by PCR were cloned into the PCR II vector (TA Cloning Kit from Invitrogen, San Diego, CA) for both the normal and A3243G mutant sequence. E. coli were transformed, individual colonies picked, and DNA isolated by a miniprep (21). The nucleotide sequence of the clones from KS was confirmed by dideoxy sequencing using a sequencing kit (United States Biochemical Corporation, Cleveland, OH). Samples were labeled with [35S]deoxycytidine triphosphate (dCTP), run on a 6% polyacrylamide/7.0 M urea gel, and detected by autoradiography.
PCR
A PCR protocol was developed allowing identification of the A3243G mutation. This mutation creates a new cleavage site for the restriction enzyme ApaI. Standard methods were used for PCR and detection and resolution of the products by gel electrophoresis (21). A pair of 27 mer primers for the mutation were designed to produce a 501-bp amplified fragment with the new restriction site exactly in the middle. When the mutation was present, a double-strength signal of 250 bp was generated after ApaI digestion. The forward sense primer was from np 2993–3019 (5'-TTGGATCAGGACATCCCGATGGTGCAG); reverse antisense primer was from np 3467–3493 (5'-GTTTTAGGGGCTCTTTGGTGAAGAGTT). PCR amplification was carried out in a final volume of 25 µL including 10 mM Tris pH 8.3, 2.5 mM MgCl, 50 mM KCl, 0.20 mM each dNTP, 2.5 U Taq polymerase (Amplitaq; Perkin-Elmer Corp., Norwalk, CT) and 0.4 pmol of each primer. The PCR reaction was run for 30 cycles as follows: denature 45 sec at 94 C, anneal 30 sec at 60 C, extend 75 sec at 72 C. Amplified products were cut by ApaI (Boerhinger Mannheim, Indianapolis, IN) for 1 h at 37 C, separated by electrophoresis on a 1.5% agarose gel, and stained with 0.5 µg/mL ethidium bromide for 30 min.
For radioactive PCR reactions, the concentration of each dNTP was
reduced to 0.04 mM, and 10 µCi [
32P]dCTP was added to give a final concentration of 2.5
Ci/mmol. This radioactivity was either added at the beginning of the
PCR for continuously labeled PCR, or added for only the last cycle. For
last-cycle labeling, the penultimate denaturation step was extended to
allow the addition of radioactive nucleotide. The last cycle consisted
of annealing and extension with an immediate cooling step. The strands
were not allowed to denature at any point during further analysis.
After restriction digestion, the fragments were separated on a 6%
polyacrylamide gel with 0.5 x TBE buffer: This gel was dried, and
DNA bands were detected by autoradiography. The quantitation of normal
and mutant fragments was by scintillation counting of each band cut
from the gel.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
. Marked deficiencies
in complexes I, III, and IV were found, with elevation of citrate
synthetase, suggesting mitochondrial proliferation. These results were
consistent with the MELAS syndrome, and mtDNA studies were performed.
In Fig. 1 the A3243G mutation, which
creates a new ApaI restriction site, is clearly seen in
mtDNA prepared from muscle, lymphoblasts, and blood from KS. Cloning
and sequencing the entire transfer RNA Leu(UUR) gene from KS muscle DNA
confirmed that this mutation, but no other, was present. However, mtDNA
isolated from the lymphoblasts and blood of his mother (SS) did not
show the A3243G mutation using ethidium bromide staining.
|
|
also shows the sensitivity of the ethidium bromide detection
procedure using mixtures of the cloned mutant and control mtDNA. The
primers were selected so that the mutation was exactly in the center of
the amplified fragment yielding a double-strength band on digestion.
Even so, only a faint band was detected when 12.5% mutant DNA was
present. A very faint band with 6.25% mutant DNA was sometimes visible
immediately after staining.
Given the maternal inheritance of mtDNA mutations and the association
of the A3243G mutation with diabetes, we remained concerned that SS
might also have this mutation. Figure 2
shows the results of a more-sensitive test, a PCR with [
32P]dCTP present throughout the amplification. Samples
were digested with ApaI, separated by gel electrophoresis,
and the DNA visualized by autoradiography. Using this procedure, the
mutation was detected in both lymphoblasts and blood from SS. These
results led to a subsequent needle muscle biopsy of SS. Despite the
lack of clinical symptoms or signs, histochemical studies of muscle
from SS suggested mitochondrial proliferation; this was confirmed by
electron microscopy, which also showed structurally abnormal
pleiomorphic mitochondria, many of which were large and contained
concentric abnormal cristae (Fig. 3).
Following PCR of SS muscle DNA, there was a suggestion of the mutant
band when the gel was viewed immediately after staining with ethidium
bromide.
|
|
shows results
from both continuous and last-cycle radiolabeled PCR assays using DNA
isolated from muscle, blood, and lymphoblasts from both KS and his
mother SS. Last-cycle PCR labeling detects a higher percentage of
mutant DNA and reflects the true proportion of mutant mtDNA, because
all radiolabeled mutant strands are cut by the restriction enzyme.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We were unable to detect the mutation in the mother’s blood or lymphoblasts by standard ethidium bromide staining of ApaI digests. Even in an affected tissue such as SS muscle, which manifested mitochondrial abnormalities, the mutation was barely visible by this method. However, because of the maternal inheritance of mitochondrial disease and reports of relatives of MELAS patients who had only diabetes, we persisted in searching for the A3243G mutation in the mother (SS). Using the sensitive method of radiolabeled PCR, the mutation was detected in her blood and cultured lymphoblasts at low levels. Because she had no other symptoms indicative of mitochondrial disease, the cause of her diabetes would probably have remained unknown, were it not for her son’s presentation. The diagnosis of mtDNA A3243G mutation is made easier if a symptomatic MELAS syndrome patient is present in the family. However, A3243G carriers who have only diabetes may well have no relative with the full MELAS presentation and therefore may be easily missed.
The variable clinical presentation of the A3243G mutation and the
fact that it is heteroplasmic complicates detection. As illustrated by
this family, and previously reported for other heteroplasmic mutations
(22), the formation of heteroduplexes during PCR markedly decreases the
sensitivity of assays that depend on restriction digestion to detect
mutation. At the beginning of the PCR amplification, the concentration
of primers is much greater than the amplified DNA sequences. But in the
later cycles, the concentration of product strands is high, and these
sequences compete with primers during the annealing step. Conditions of
maximum heteroduplex formation occur during the plateau phase of the
PCR with random reannealing of strands and little or no new synthesis.
Then the proportion of homoduplex normal, heteroduplex, and homoduplex
mutant strands will follow the distribution p2, 2pq, and
q2, respectively, where p and q represent the proportion of
normal and mutant strands. In that case, because only mutant homodimers
would be detected by restriction digestion, a mutation level of 0.10
(q) might be detected as only 0.01 (q2) of the total DNA
sequences. The 9-fold greater level of mutant strands present in the
heterodimers (0.09) would remain undetected. Thus the amount of mutant
DNA detected or observed by ethidium bromide staining (Fig. 1) or
continuous PCR labeling (Table 2 column 2) may be much less than the
true level of mutation. The heteroplasmic state of mitochondrial
mutations has also been reported to complicate detection based on
Southern blotting (23) and direct sequencing (24). A radiolabeled assay
with addition of isotope for the last cycle of a PCR amplification
provides the best system for detection of mutant mtDNA when restriction
digestion is used as the detection method (23, 24). This approach
prevents formation of labeled heteroduplexes and allows accurate
quantitation of mutation levels (Table 2 column 3).
If sensitive screening methodologies for mtDNA mutations are not used for testing of patients presenting with only diabetes mellitus, the underlying incidence of this cause of both IDDM and NIDDM will remain unknown. Muscle biopsy is an expensive and invasive procedure; therefore initial screening studies will almost certainly use blood, in which the level of the A3243G heteroplasmic mtDNA mutation will usually be low. Many of these individuals will not be diagnosed if the standard PCR DNA test using ethidium bromide staining is used. This insensitive methodology will underestimate the prevalence of the A3243G mutation, as well as other heteroplasmic mtDNA mutations, denying patients and their relatives who carry these mutations proper diagnosis and genetic counseling. We recommend the use of one of these radioactive PCR methods for sensitive accurate screening for mtDNA mutations in the diabetic population.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received April 25, 1997.
Accepted May 28, 1997.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
G(3243) mutation a clinical and genetic
study. Brain. 118:721–734.This article has been cited by other articles:
 |
|
 |
|
  Y. Jiang, T. A. Hall, S. A. Hofstadler, and R. K. Naviaux Mitochondrial DNA Mutation Detection by Electrospray Mass Spectrometry Clin. Chem., February 1, 2007; 53(2): 195 - 203. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Singh, S. Ellard, A. Hattersley, and L. W. Harries Rapid and Sensitive Real-Time Polymerase Chain Reaction Method for Detection and Quantification of 3243A>G Mitochondrial Point Mutation J. Mol. Diagn., May 1, 2006; 8(2): 225 - 230. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Ohkubo, A. Yamano, M. Nagashima, Y. Mori, K. Anzai, Y. Akehi, R. Nomiyama, T. Asano, A. Urae, and J. Ono Mitochondrial Gene Mutations in the tRNALeu(UUR) Region and Diabetes: Prevalence and Clinical Phenotypes in Japan Clin. Chem., September 1, 2001; 47(9): 1641 - 1648. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Guttman, H.-G. Gao, and R. Haas Rapid Analysis of Mitochondrial DNA Heteroplasmy in Diabetes by Gel-Microchip Electrophoresis Clin. Chem., August 1, 2001; 47(8): 1469 - 1472. [Full Text] [PDF]
|
|
|
|
 |
|
 R. H. HAAS A Comparision of Genetic Mitochondrial Disease and Nucleoside Analogue Toxicity: Does Fetal Nucleoside Toxicity Underlie Reports of Mitochondrial Disease in Infants Born to Women Treated for HIV Infection? Ann. N.Y. Acad. Sci., November 1, 2000; 918(1): 247 - 261. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. D. Graf, J. Marin-Garcia, H.G. Gao, S. Pizzo, R. K. Naviaux, D. Markusic, B. A. Barshop, E. Courchesne, and R. H. Haas Autism Associated With the Mitochondrial DNA G8363A Transfer RNALys Mutation J Child Neurol, June 1, 2000; 15(6): 357 - 361. [Abstract] [PDF]
|
|
|
|
 |
|
 A. Pugliese, E. Kawasaki, M. Zeller, L. Yu, S. Babu, M. Solimena, C. T. Moraes, M. Pietropaolo, R. P. Friday, M. Trucco, et al. Sequence Analysis of the Diabetes-Protective Human Leukocyte Antigen-DQB10602 Allele in Unaffected, Islet Cell Antibody-Positive First Degree Relatives and in Rare Patients with Type 1 Diabetes J. Clin. Endocrinol. Metab., May 1, 1999; 84(5): 1722 - 1728. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
