| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Istituto di Ricovero e Cura a Carattere Scientifico (IRCCS) H. San Raffaele, Unit of Genetics and Molecular Diagnostics and
2
Unit of Protein Engineering, Via Olgettina 58, 20132 Milan, Italy.
3
Section of Chemistry, Faculty of Medicine, University of
Brescia, 25100 Brescia, Italy.
a Address correspondence to this author at: Unit of Protein Engineering, Dibit, HSR, Via Olgettina 58, 20132 Milan, Italy. Fax 39-02-2643 4844; e-mail p.arosio{at}hsr.it.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: The double-gradient denaturing gradient gel electrophoresis (DG-DGGE) method consisted of PCR amplification of the target genomic DNA with GC-clamped oligonucleotides. The sequence encoded the 5' untranslated flanking region of ferritin L-chain mRNA, which includes an IRE stem-loop structure. The product was subjected to DG-DGGE (8.5–15% polyacrylamide and 50–95% denaturant) to separate the homo- and heteroduplexes.
Results: The method clearly identified all eight accessible mutations, including C-G transversions, which are the most difficult to detect. The method was applied to scan DNA samples from 50 healthy subjects and from 230 subjects with serum ferritin >400 µg/L. The new mutation G14C was identified.
Conclusions: The DG-DGGE method detects all the mutations in the L-ferritin IRE sequence, is rapid and economical, and can be applied to scan large populations. The first population study indicated that the mutations are rare and may involve regions of the IRE structure not yet characterized.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Increased serum ferritin concentrations often are indices of excess
iron stores; however, several hyperferritinemic conditions are not
related to iron overload, and they include disorders such as
inflammations, neoplasia (3), and the recently described
metabolic dysfunction-associated liver iron overload syndrome
(4), conditions that are associated to iron
decompartmentalization and characterized by low or normal transferrin
saturation. A distinct and more specific cause of hyperferritinemia is
associated with deregulation of ferritin synthesis, as it occurs in
hereditary hyperferritinemia-cataract syndrome (HHCS; OMIM 600886).
This is a genetic, autosomal dominant disorder caused by heterogeneous
mutations in the IRE of L-ferritin mRNA. The mutations abolish or
reduce the affinity with the IRPs, and this determines a constitutive
up-regulation of ferritin L-chain in serum and all tissues
(5)(6). The 10- to 20-fold higher accumulation
of L-ferritin in serum and tissues has no evident effects on body iron
status, but it often is associated with early bilateral cataract onset,
likely via aggregation of the protein in the lens
(7)(8). Thirteen different mutations have been
described to date in 15–20 families (Table 1
), including one sporadic case (9). The mutations
span over a 30-nucleotide stretch that is part of the IRE structure,
with some affecting the loop that interacts directly with the IRPs,
whereas other mutations affect the stems or the bulge of the structure
and modify the conformation of the IRE, leading to lower affinity
binding (10). The in vitro effects of the mutations on IRE
structures were found to be correlated with clinical symptoms: subjects
with mutations that affect the upper part of the structure, which is
involved in IRP binding, have ferritin in the 1000–1500 µg/L range
and early-onset cataracts, whereas the mutations that affect the distal
parts of the stems are associated with slightly increased serum
ferritin (
500 µg/L) and subclinical cataracts (10).
This finding suggests that mutations may occur that have clinically
undetectable effects on serum ferritin and cataract formation. These
mutations cannot be easily identified with the method used at present,
which consists of direct sequencing of PCR-amplified DNA from patients
with abnormal and unexplained serum ferritin concentrations, either
associated or not associated with early onset of cataracts.
|
For a more complete analysis of the prevalence of mutations in this locus and the identification of possible polymorphisms, a faster and less expensive method than direct DNA sequencing is needed. A suitable approach consists of a fast mutational scanning technique that identifies all samples with mutations to be verified by direct sequencing.
The available scanning techniques consist of single-strand conformation polymorphism (11), denaturing gradient gel electrophoresis (DGGE) (12)(13), chemical cleavage of mismatch (14), RNase cleavage (15), and heteroduplex analysis (16), which identify only a portion of the mutations; among these methods, DGGE is the most effective with a detection rate close to 100% (17). The technique has been further improved by superimposing a porous gradient on the denaturing gradient [double-gradient DGGE (DG-DGGE)], which minimizes band broadening, even in prolonged runs, and permits more accurate band separation (18)(19). The DNA stretch of interest for mutations in HHCS is approximately the optimal size for this type of technology, which allows mutational scanning in a single electrophoretic analysis.
We describe the development of the assay, showing that it is able to identify without ambiguity all of the eight mutations tested. The assay was used to analyze DNA from 50 healthy subjects and 230 subjects with abnormalities in markers of iron metabolism. A new mutation was identified.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
primer design and pcr conditions
The optimal PCR set of primers for GC-clamped DG-DGGE analysis of
the IRE sequence was selected using the MELT87 program, kindly provided
by Dr. L.S. Lerman (MIT, Cambridge, MA) (12).
Primers were designed to allow the entire L-ferritin IRE sequence to
span over a single melting domain with
Tm of 78 °C. The PCR reaction
mixture was as follows: 1 µg of DNA, 200 µM dNTPs 10 mM Tris-HCl
(pH 8.8), 50 mM KCl; 1.5 mM MgCl2, 1 mL/L
Triton X-100, 2.0 U of thermostable DNA polymerase (DynaZyme; Finnzymes
OY, Celbio), and 30 pmol of each primer in a total volume of 100 µL.
PCR cycles were as follows: 1 cycle at 94 °C for 5 min, followed by
30 cycles of 45 s at 94 °C, 45 s at 56 °C, and 45
s at 72 °C. The PCR product was 315 bp in length.
Heteroduplexes were generated at the end of each PCR session by 5 min
of denaturation at 94 °C and 1 h incubation at 56 °C.
optimization of dg-dgge conditions
The amplified DNA samples (15 µL) were subjected to
electrophoresis in a 8.5–15% polyacrylamide linear porosity gradient
gel (acrylamide:bisacrylamide, 39:1 ratio), in Tris-EDTA-acetate buffer
(40 mmol/L Tris, 20 mmol/L NaOH, 1 mmol/L EDTA, titrated to pH 7.6 with
acetic acid). The denaturing gradient was a 50–95% linear gradient of
a mixture of urea and formamide (100% denaturant: 7 mol/L urea
and 400 mL/L formamide). The gel slabs were 0.75 mm thick, 150
mm wide, and 16 cm long. At the end of an 18-h run at 75 V, the gels
were stained in ethidium bromide.
fluorescent sequence analysis
Sequencing was carried out in both directions using fluorescent
dideoxy chain terminator cycle sequencing on a ABI PRISM 377 DNA
sequencer (Perkin-Elmer). Specific primers for sequencing were designed
external to those used for DG-DGGE as follows: forward primer,
5'-TCCTTGCCACCGCAGTTGG-3'; reverse primer, 5'-CTGACGAATCTGGGAGCTCA-3'.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
50 residues; all the mutations
identified to date are between positions 18 and 57 (Table 1
). We
considered that the analysis of a 200-residue stretch from the start of
translation was sufficient to recognize all mutations that may affect
IRE functionality. The region is not interrupted by introns and can be
analyzed by PCR amplification of genomic DNA. To set up the DG-DGGE
assay, we first analyzed the melting profile of the sequence using the
MELT87 program (12) to identify a region covering the whole
IRE sequence that had a single melting domain with a
Tm of 78 °C. GC-clamped
oligonucleotides to amplify this region were synthesized, and PCR
conditions and DG-DGGE runs were optimized with the aim to identify all
mutations in a single electrophoretic analysis. Because of the high G
and C content, which accounts for the high temperature melting profile
of the IRE sequence, we used a narrow highly denaturing gradient
(50–95%) to maximize homo- and heteroduplex separation.
analysis of known ire mutations
To evaluate the scanning efficiency of the method, we analyzed DNA
samples from eight HHCS subjects carrying mutations that had already
been characterized (Table 1
). All of them produced electrophoretic
patterns clearly different from the single sharp band of the controls
(Fig. 1A
). The mutations were characterized by three extra bands, one
corresponding to the mutated homoduplex and two weaker bands
corresponding to heteroduplexes with lower melting points and slower
mobility. The method was also able to reveal G-C transversions (Fig. 1A
, lane 9), which are notoriously difficult to identify.
|
All of the patterns could be interpreted on the basis of the known
mutations. For example, the lack of heteroduplexes in lane 8 (Fig. 1A
)
was expected for the large deletion delC10-A38, as were the
remarkably low melting points of the heteroduplexes in the double
substitution C18T + T22G (Fig. 1A
, lane 6). The A40G mutation is
expected to stabilize double-stranded DNA, and in fact its mutated
homoduplex melting point was higher than that of the wild type (Fig. 1A
, lane 5). The only sample that produced a pattern that could not be
easily interpreted was the one of lane 3 of Fig. 1A
because the G41C
mutation was predicted to produce a homoduplex with a melting point
very close to the wild type, like that of lane 9, and not the one we
observed. Direct sequencing revealed that in fact the DNA contained an
extra mutation in the adjacent residue, A40G (Fig. 2
), which
was not detected before, and explains the lower stability of the
homoduplex.
|
population scanning
To validate the techniques and to evaluate the possible existence
of polymorphisms, we first scanned a control group of 50 DNA samples
from healthy subjects with iron metabolism indices within reference
values. No abnormal electrophoretic patterns were observed,
indicating a prevalence of polymorphisms <1%.
We then analyzed 230 DNA samples from subjects with serum ferritin
concentrations ranging from 420 to 5500 µg/L who underwent
HFE genotyping for diagnosis of hereditary hemochromatosis.
The high serum ferritin concentrations were attributed to iron overload
or to liver disease, and only 12 subjects had transferrin saturation
<30%. All of them showed single-band normal electrophoretic patterns
in the DG-DGGE assay, except one (Fig. 1B
, lane 1), which had
an extra, close-running heteroduplex. DNA sequencing revealed the
mutation C14G (Fig. 1C
), the location of which in shown in Fig. 2
. The subject had high serum ferritin (1200 µg/L) and high
transferrin saturation (96%), whereas Perl’s staining of the liver
biopsy showed no evident iron accumulation. The abnormal indices of
iron status were attributed to the presence of hepatitis, indicated by
the positivity to hepatitis B virus. This patient showed no signs of
cataracts. Family studies could not be performed.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
).
Other mutations that we had no access to were C36A (21),
deletion A38-C39, and deletion U42-G57 (22), which can be
predicted to produce easily separable homoduplexes with lower melting
points. More important is the evidence that the method clearly
identified C-G conservative transversions, which do not alter overall
base composition and have a minor effect on DNA
Tms (Fig. 1A
, lane 9). These
transversions are known to be the most difficult mutations to detect in
scanning techniques because the two homoduplexes often are not
separated and recognition is based on identification of the weak
heteroduplex bands, which can be ambiguous. The finding that the method
is able to recognize all tested mutations in the region made us
confident that it can distinguish all possible mutations in the DNA
stretch of the L-ferritin DNA that encodes for the complete IRE
structure. An additional advantage of the method includes the
deductions that can be made based on the electrophoresis patterns,
which complement the results of DNA sequencing. This was useful at
least in one case and allowed us to correct an error by demonstrating
that the described point mutation G41C (23) is really a
double mutation (G41C + A40G; Fig. 1A
, lane 3). Reanalysis of DNA from
the family members confirmed that all of the affected subjects carry
the double mutation (not shown).
The DG-DGGE technique is designed for the scanning of large samples,
and we first applied it to the analysis of 100 alleles from 50 healthy
controls. No abnormal patterns were observed, and with the assumption
that the method has 100% sensitivity, a prevalence of mutations or
polymorphism <1% was deduced. The analysis was then extended to 230
samples from subjects with some abnormalities in iron metabolism who
underwent HFE genotyping for diagnosis of hemochromatosis.
In principle, this is not a population at risk for HHCS because the
abnormal serum ferritin concentrations are attributed to iron overload.
However, we decided to analyze the group because serum ferritin is an
important marker for the diagnosis of hemochromatosis, and it has
previously been reported that subjects with HHCS had been diagnosed
with hemochromatosis (23). The DG-DGGE assay identified one
abnormal electrophoretic pattern (Fig. 1B
), and direct DNA sequencing
revealed the new mutation, C14G (Fig. 1C
).
The mutation C14G abolishes a base pairing in the lower part of the
stem, and it is expected to affect the conformation of the IRE
structure. In fact, structural modeling predicts a folding stability of
-25 kcal/mol, compared with -29.7 kcal/mol for the wild type
(24). A direct experimental in vitro study of the effects of
the mutation on IRP binding would be complex because this region of the
IRE structure has not been explored and electrophoretic mobility shift
assays for IRE activity use RNA fragments from residue 26 to residue
55, which excludes this portion (25). The subject that
carries the mutation had high serum ferritin, which was partially
explained by iron overload, as indicated by the high percentage of
transferrin saturation. Ophthalmologic analysis excluded cataracts, and
genetic analysis did not identify HFE mutations. We could
not perform family studies; thus, we do not have direct evidence that
the mutation has phenotypic expression. The mutation structurally
closest to this one (C18U + U22G) was shown to have mild clinical
effects, with serum ferritin concentrations 500 µg/L and
subclinical cataracts (5); by analogy the C14G mutation may
have a minor effect on serum ferritin or be the first polymorphism of
the region identified to date.
In conclusion, we demonstrate that the DG-DGGE technique we developed is adequate for the scanning of large populations for L-ferritin IRE mutations and that mutations and polymorphisms in the region are rare. The method allows more accurate definition of the prevalence of the mutations in populations with juvenile, bilateral, and congenital cataracts and in populations with hyperferritinemia. We speculate that it is particularly suitable for the study of translational disorders in which the DNA segment exposed to mutation is sufficiently short to be scanned in a single electrophoretic analysis. HHCS is the major example of these types of diseases, but IRE structures are present on the mRNA for H-ferritin, mitochondrial aconitase, ALAS1, IREG1, DMT1, and also transferrin receptor-1. Their mutations have not been explored, although they may have serious pathological consequences (26). In addition, hereditary thrombocythemia is a translational disorder characterized by mutation in the regulatory 5' untranslated region of the TPO gene, a region that can be efficiently scanned with DG-DGGE (26).
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  J. E. Craig, J. B. Clark, J. L. McLeod, M. A. Kirkland, G. Grant, J. E. Elder, M. G. Toohey, L. Kowal, H. F. Savoia, C. Chen, et al. Hereditary Hyperferritinemia-Cataract Syndrome: Prevalence, Lens Morphology, Spectrum of Mutations, and Clinical Presentations Arch Ophthalmol, December 1, 2003; 121(12): 1753 - 1761. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. M. Torti and S. V. Torti Regulation of ferritin genes and protein Blood, May 15, 2002; 99(10): 3505 - 3516. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
10-38, and
G51C, respectively. (B), pattern of the new mutation,
C14G (lane 1), compared with wild-type control
(lane 2). (C), sequence analysis of DNA
from a control (top) and from the subject with the C14G
mutation (bottom). The arrow indicates
the mutated residue.