| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
* Department of Medical Genetics,
Ophthalmology and Vision Sciences,
§ Department of Epidemiology and Public Health, Queen's University, Belfast, U.K.;
|| Eye Unit, Royal Liverpool Hospital, Liverpool, U.K.; and
The Wolfson Institute for Biomedical Research, The Rayne Institute, University College, London, London, U.K.
1Correspondence: Ophthalmology and Vision Sciences, Institute of Clinical Sciences, Grosvenor Road, Belfast BT12 6BA, U.K.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: microvascular • transcription • NOS2A promoter • pentanucleotide repeat
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. This scenario is
incompatible with normal vision. Nitric oxide
(NO)2
, a potent biological second messenger, is involved in the regulation of
vascular tone, wound repair, defense mechanisms, and other processes.
It is synthesized by the action of nitric oxide synthases (NOS) through
the 5-electron oxidation of the terminal guanidino-N2 of the amino acid
L-arginine (2)
. A principal action of NO is its effect on
vascular smooth muscle. Its synthesis by the vascular endothelium,
which lies close to the vascular smooth muscle, is recognized as the
pathway through which the endothelin-mediated control of blood flow is
exercised.
Endothelium-mediated vasodilatory responses in blood vessels are
aberrant in response to agonist stimulation in diabetic subjects
(3
4
5)
. In vitro and in vivo studies
have shown that the synthesis and release of vasoconstrictors by the
vascular endothelium are increased in the diabetic state (6
, 7)
. Evidence suggests that endothelium-mediated vasodilation is
defective and reduced in diabetes along with an increase in
vasoconstrictor activity. It is now recognized that aberrations in
retinal blood flow in early diabetes are also linked to vascular
endothelial dysfunction (8
, 9)
. The retinal circulation,
which is devoid of any extrinsic innervation, is dependent entirely on
endothelium-mediated autoregulation (10)
; thus,
endothelial dysfunction in diabetes is likely to have a major effect on
the circulation within the retina.
In addition to environmental factors (11)
, results of the
Diabetes Control and Complications Trials indicated that genetic
factors may also affect the development of onset of diabetic
retinopathy (12)
. The risk of severe diabetic retinopathy
in the siblings of affected individuals is increased, with a strong
tendency of familial clustering of this complication (12)
.
Studies of twins (13)
and various ethnic populations
(14)
also demonstrate the genetic influence in diabetic
retinopathy.
Three members of the nitric oxide synthase gene family have been
identified: neuronal (NOS1), inducible (NOS2A), and endothelial (NOS3),
all of which could play a role in the diabetic retina. Under normal
conditions, NOS2A is not expressed in the retinal vasculature. Exposure
to high ambient glucose may influence NO release via increased NOS2A
expression and reduced constitutive endothelial NOS gene (NOS3)
expression in cultured retinal vascular endothelial cells
(15)
.
We have examined a sample of patients with Type 1 and Type 2 diabetes
for associations with the individual repeat alleles of a polymorphic
pentanucleotide repeat microsatellite marker located in the 5' promoter
region of the NOS2A gene (16)
. The individual repeats were
also examined for their ability to induce transcription of the NOS2A
gene under normal (5 mM) and high (25 mM) glucose conditions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Patients with intermediate grades of retinopathy
(levels 20–50) and/or maculopathy alone were excluded from the study
to allow clear separation of the two groups selected for study. The following clinical data were collected: age and sex; type, duration and family history of diabetes; presence of vascular disease (angina, hypertension, peripheral vascular disease); renal function (dipstick microalbuminuria, albumin creatinine ratio); and glycemic control (two Hba1c readings obtained per year over the last 3 years for the majority of patients, i.e., 85%). Type 1 diabetes was defined as age 30 years or less at diagnosis or evidence of absolute insulin dependence (e.g., episode of ketoacidosis); all other cases were deemed to be Type 2 diabetes. Data shown here represent 211 diabetic patients from Northern Ireland and 127 diabetic patients from Liverpool. A sample population of 97 individuals obtained from the Northern Ireland Regional Genetics Laboratory was used as a control.
DNA extraction and PCR
Each patient donated a 10 ml blood sample. Total DNA was
extracted from the white blood cell pellet (18)
. DNA from
each patient was used as a template for polymerase chain reaction (PCR)
amplification (19)
for the detection of polymorphism in
the NOS2A gene. Alleles were scored from autoradiographs and the
information entered into a database (MS Excel spreadsheet program)
along with the clinical data.
PCR primers and PCR conditions have been described previously for the
NOS2A marker (16)
. The marker is a pentanucleotide
(CCTTT)n repeat located in an S1 hypersensitive
region ~2.5 kb upstream of the human NOS2A gene transcription start
site. There are two reasons for choosing this repeat for analysis.
First, the repeat is highly polymorphic and is informative in the
general Caucasian population (16)
. Second,
polypyrimidine/polypurine repeats in the promoter can possibly affect
transcription; one way this can occur is by forming the unusual
structure of triplex DNA (20
, 21)
. This NOS2A marker has
11 reported alleles (175–225 bp, 8–18 repeats) with heterozygosity of
0.80.
Construction of NOS2A luciferase promoter-reporter plasmids
containing different numbers of (CCTTT)n repeats
PCR was used to obtain a native 1.2 kb fragment immediately
upstream of the transcription start site of the human NOS2A gene (EMBL
accession number X97821) from a human NOS2A cosmid clone, pCOS4
(22)
. The forward primer
5'-CAAAGTGTTGGTACCGTGAGATCA-3' is located -1183 bp
from the transcription start site and the reverse primer,
5'-CTTCGGGACTCTCGAGAACTGCCCAG-3', is located +122 bp, exon
1 (KpnI and XhoI restriction sites are
underlined). The PCR product was cut with these restriction enzymes and
cloned into a vector, pGL3-enhancer (Promega, Madison, Wis.), which
contains the promoterless firefly luciferase reporter and SV40 enhancer
elements. The construct (Fig. 1
) includes consensus promoter elements including a TATA box, NF
B, and
NFIL-6 binding sites in domain region I from -45 to -249 bp (73%
sequence identity with the mouse region I, -44 to -206). Promoter
elements from region II include NFB, ISRE [interferon (IFN)
-stimulated response element], and GAS (IFN-r-activated
sequence) from -1058 to -1183 (65% sequence identity with the mouse
region II, -901 to -1029). Different deletions of this promoter
element were then made by using the Erase-a-Base System (Promega) to
form a series of deletion constructs from the original pNOS1200
plasmid. The deletion construct pNOS200 spans a region from -200 to
+122 bp of the human NOS2A promoter, whereas pNOS400 spans a region
from -400 to +122 bp.
|
The (CCTTT)n pentanucleotide repeat region was cloned into the pNOS1200 construct using a pair of primers, 5'-ACCCCTGGTACCCTACAACTGCAT-3' and 5'-GCCACGGTACCCTAGCCTGTCTCA-3', by directional PCR from human genomic DNA obtained from individuals with different repeats. The PCR primers were designed to incorporate a restriction site, Acc651. Different PCR fragments were cut with the restriction enzyme Acc651 and cloned into the Acc651 site at the 5'-end of the 1.2 kb pNOS1200 promoter fragment. The resulting constructs were named pNOS2-A9, pNOS2-A12, pNOS2-A14, pNOS2-A15, and pNOS2-A17, containing 9, 12, 14, 15, and 17 repeats, respectively. All the constructs were also sequenced using the flanking sequencing primers (pGL3 forward and reverse) in both orientations to confirm the authenticity of the PCR product. Plasmid DNA was purified using Qiagen columns.
Transient transfections, cell induction, and luciferase assays
The human colon carcinoma cell line, DLD-1 (American Type
Culture Collection, Rockville, Md.) was maintained in RPMI 1640
supplemented with 10% fetal bovine serum in a humidified atmosphere of
5% CO2, 95% air. Cells at 70% confluence were
treated with Superfect (Qiagen, U.K.) containing 1–2 µg of the
various constructs described above and between 0.2 and 0.4 µg of
pRL-TK, a Renilla luciferase gene under control of a TK promoter as an
internal control. After transfection for 36–48 h, cells were treated
with IL-1ß (1.5 ng/ml) alone or a cytokine mixture: IFN
(100
units/ml), IL-1ß (0.5 ng/ml), IL-6 (1 ng/ml), and tumor necrosis
factor 
(TNF-; 2 ng/ml). The rationale for selecting IL-1ß as
the inducer was based on the evidence for involvement of this cytokine
in diabetic microvascular disorders (23
, 24)
. The
conditions for using IL-1ß were based on its efficiency as an inducer
in previous experiments on human NOS2A (25
26
27
28)
. All
cytokines were purchased from Genzyme (Boston, Mass.). Cell extracts
were prepared 4 h after induction, and 40 µl of lysate was used
for the determination of luciferase activity. A dual luciferase
reporter assay system (Promega) was used to quantitate the luminescent
signal from firefly luciferase produced from the series of promoter
constructs. The luciferase activity of the internal control Renilla
luciferase plasmid pRL-TK was measured in the same test tube by adding
Stop & Glo Reagent, using the manufacturers recommended procedure
(Promega). The instrumentation used was an LKB Designs Model 1250
Luminometer; luciferase activity was expressed as resonance light
units. The concentration of glucose in transfection experiments was 5
mM in all experiments, unless specifically stated, and represents
normal glucose concentrations in mammalian systems.
To mimic the diabetic state, the concentration of glucose was raised to
25 mM (15)
in certain experiments. Although some
biochemical changes are seen immediately, alterations in the
transcription of specific genes are generally reported to occur 72 h after cellular exposure to hyperglycemia. Thus, cells were cultured
in 25 mM glucose for 4 h or 96 h and then subjected to
IL-1ß induction for 4 h while remaining in high glucose media
(15
, 29)
. To mimic the fluctuation of glucose levels seen
in vivo, cells were incubated for 96 h in high glucose
(25 mM), then transferred back to normal glucose levels for 4 h,
after which the luciferase assays were carried out. Transfections were
performed in sets with triplicate plates for each construct. Promoter
activity for each transfection is reported after normalization against
the control activity from the Renilla luciferase. The reported results
are from three separate transfection experiments.
Statistical methods
In Northern Ireland, a study size of 100 diabetic patients with
retinopathy and 100 patients with no retinopathy was originally chosen
to detect a minimum 2.5-fold increase in risk of retinopathy associated
with an allele whose frequency was 20% in the nonretinopathic
patients, with 80% power at the 0.25% significance level (i.e., the
conventional 5% level with a Bonferroni correction for 20 multiple
comparisons). Allele frequencies were compared using Pearson's
2 test. The 5% level of significance was
used, but P values for comparisons between the retinopathic
and nonretinopathic subgroups were adjusted for the number of
comparisons performed by using the Bonferroni correction. The relative
risk of retinopathy associated with carriership of an allele was
estimated by the odds ratio, and 95% confidence limits were obtained
by Cornfield's approximation.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Patients from Northern Ireland in the `no retinopathy' group and the
`retinopathy' group were well-matched for age at recruitment and
duration of diabetes. The majority of the patients recruited from
Northern Ireland belonged to the Type 1 category. Analysis showed that
Hba1c, presence of renal complications and
hypertension were significantly different between the `retinopathy'
group and the `no retinopathy' group.
|
Unlike the Northern Ireland sample, the Liverpool sample consisted mostly of individuals with Type 2 diabetes. Analysis of the risk factors confirmed that renal complications and Hba1c showed significant differences between the Liverpool `retinopathy' and `no retinopathy' groups. However, hypertension was not significantly different between the two groups, which may be explained by the predominance of Type 2 patients in the Liverpool sample.
There is an association with NOS2A and retinopathy status
The 14-repeat allele is significantly associated with the `no
retinopathy' group (Table 2
). The allele frequencies of the `no retinopathy' group were similar
to the random samples from Northern Ireland (Table 2)
. Five of the
NOS2A alleles were considered to have potential for comparison because
their frequency in the combined group exceeded 5%.
|
The 14-repeat allele was found in both Type 1 and Type 2 diabetes
sufferers. The prevalence of hypertension and renal complications was
compared in carriers and noncarriers of the 14-repeat allele. Patients
with the 14-repeat allele had a lower frequency of renal complications
(2=5.32, df=1; P=0.02) and the
prevalence of hypertension was reduced but not statistically
significant (2=2.89, df=1; P=0.09).
However, stratification of the data by retinopathy status showed that
these associations with the 14-repeat allele could largely be
attributed to retinopathy status itself (data not shown). The risk of
retinopathy in an individual with diabetes carrying the 14-repeat
allele is estimated to be 0.21 times the risk of retinopathy in an
individual with diabetes who does not carry this allele (95% CI,
0.06–0.60) in Northern Ireland (Table 3
).
|
Similar trends were observed for the Liverpool patients. Almost twice as many individuals carry the 14-repeat allele in the `no retinopathy' group as in the retinopathy group, but this result did not achieve significance. Carriers of the 14-repeat allele in the Liverpool sample also have a lower incidence of renal complications and hypertension (data not shown).
IL-1ß inducibility of NOS2A is affected by repeat length and high
glucose concentrations
The effects of varying numbers of (CCTTT)n
repeats on NOS2A gene transcription are shown in Fig. 1C
.
The CCTTT repeat sequences enhance the minimal NOS2A promoter induction
activity in response to IL-1ß. The parent clone pINOS-1200 shows a
twofold increase in luciferase activity on IL-1ß induction. The 9,
12, and 14 repeats gave progressive increases in luciferase activity in
response to IL-1ß induction (2.3-fold, 3.4-fold, and 4.0-fold,
respectively). The magnitude of increase in luciferase activity in
these experiments was further compared. The construct containing the 12
repeats produced a significantly greater induction of luciferase as
compared to the construct with 9 repeats, whereas the 14-repeat
construct produced significantly greater luciferase activity than the
12-repeat construct (P<0.001; Student's t
test). The luciferase levels attained by the 9-repeat construct were
not significantly different from those of the parent clone pNOS2–1200.
A fourfold induction over background levels was found with the
15-repeat construct, indicating it has an effectiveness similar to the
14-repeat allele (the differences between these two constructs was not
significant). Similarly, the 17-repeat construct increased induction to
3.9-fold over basal levels (data not shown). In separate experiments, a
cytokine mixture comprised of IFN (100 units/ml), IL-1ß (0.5
ng/ml), IL-6 (1 ng/ml), and TNF (2 ng/ml) did not enhance measurable
inducibility as compared to IL-1ß alone (data not shown).
To test the effects of hyperglycemia on the activity of the panel of
promoter constructs, the transfected cells were cultured in 25 mM
glucose and incubated with IL-1ß. Culture in 25 mM glucose for 4 h did not modify IL-1ß induction (data not shown). Culture in 25 mM
glucose for 96 h followed by 4 h of 5 mM glucose prior to
induction, which was also in 5 mM glucose, resulted in a reduction in
transcription in transfectants containing the 9, 12, 14, and 15 repeats
(Fig. 2
). The 14 repeat was the least affected, with a reduction in
transcription from 4.0-fold to 2.6-fold. The net increase in
transcription (2.6-fold) was still highly significant when compared to
noninduced cultures (P<0.01). Constructs containing the 12
and 15 repeats demonstrated greater reduction in transcription after
culture in high glucose (3.0-fold to 1.5-fold and 4.0-fold to 1.4-fold,
respectively; Fig. 2
, Set 1) than that measured for constructs
containing 14 repeats. However, the transcription levels of the 12-,
14-, and 15-repeat constructs were significantly different from
noninduced cultures. Constructs containing 9 repeats exhibited an
induction profile similar to the construct containing 12 repeats (data
not shown). When cells were cultured in 25 mM glucose and maintained in
that medium during IL-1ß induction, transcription by the 12-repeat
construct was completely abolished. Under these same experimental
conditions, measured levels of transcription by the 14- and 15-repeat
constructs were not different from that reported when culture was
carried out in 25 mM glucose and returned to normoglycemic conditions
prior to induction.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
Furthermore, Murohara et al. (31)
have demonstrated that
NO can modulate the process of angiogenesis. In addition to its key
role in regulating vasodilator tone, NO has several other important
roles in the maintenance of the integrity of the vasculature. NO has
antithrombogenic and antiplatelet regulatory activities, can act both
as a free radial scavenger and as a free-radical itself, and has growth
regulatory effects on the vascular smooth muscle (2)
.
Therefore, a lack of NO has the potential to result in a procoagulant,
vasospastic, and pro-oxidant environment leading to closure of the
microvessels of the retina, a state of affairs seen with monotonous
regularity in diabetic retinopathy. NOS2A was thus considered a prime
candidate gene for the investigation of genetic susceptibility to
diabetic retinopathy. In the present study, individuals carrying the 14-repeat allele of the NOS2A gene generally displayed no retinopathy, had a reduced prevalence of renal complications, and, to a lesser degree, a reduced prevalence of hypertension. The calculated odds ratio showed that a person with diabetes carrying this allele has 0.21-fold chance of developing retinopathy as compared to those not carrying the allele. As this trend was found in both Type 1 and Type 2 individuals, carriage of the 14-repeat allele was not considered a feature of diabetes itself, but specifically of diabetic retinopathy. It is unclear from this study whether carriage of the 14 repeat would have an affect on the length of the disease free interval. If the 14-repeat allele confers protection from severe diabetic retinopathy in patients, one would expect an excess of this allele in the `no retinopathy' sample compared to the controls (nondiabetic control sample). We do observe this trend, but it does not reach significance. Our samples were selected to reflect the extreme phenotypic range (thus excluding many patients) in order to have clear separation of the clinical phenotypes, and this may help explain the lack of a larger difference in frequencies of the 14-repeat allele between the `no retinopathy' and control populations.
Inducible NOS protein expression studies (26
, 32)
and
NOS2A promoter-luciferase assays (33
, 34)
have confirmed
that IL-1ß is a strong inducer of human NOS2A; furthermore, this
cytokine has been implicated in the pathogenesis of diabetic
microvascular complications (23
, 24)
. The different repeat
alleles appear to have diverse effects on the ability of the 5'
upstream promoter region to act as an effective transcription
regulatory element. This effect was not solely dependent on the length
of the repeat because the larger repeats tested (15 and 17 repeats)
were no more effective than the 14 repeat in mediating IL-1ß
induction of NOS2A.
Upon mimicking a diabetic situation by culturing the transfectants in 25 mM glucose for 96 h, differences were observed on IL-1ß -mediated inducibility of NOS2A. Transfection of DLD-1 cells with the 12-repeat construct and growth in high glucose for 96 h, followed by 4 h of exposure to normal glucose, showed a substantial reduction in transcription (3.5-fold to 1.5-fold). In the same construct, continuing high glucose conditions during induction by IL-1ß resulted in total inhibition of transcription of NOS2A. This complete abrogation of expression did not occur in cells transfected with constructs containing the 14 and 15 repeat plasmids, although inducibility in both these cases was reduced significantly. However, NOS2A expression after IL-1ß induction was least affected by culture in high ambient glucose when the transfectants contained the 14 repeat.
The level of expression of NOS3 (constitutive NOS) is reduced in the
retinal vascular endothelial cells in vivo (6)
and in vitro (15)
in a diabetic milieu. Thus,
it is possible that biological feedback systems come into play to
permit induction of NOS2A in an attempt to achieve homeostasis. This is
supported by the work of Graier et al. (35)
, who found
that cyclic GMP is up-regulated in endothelial cells after exposure to
44 mM glucose for 24 h. They postulated that this result was due
to EDRF/NO formation. The isoform of NOS involved was not specified
but, due to the time course of induction, NOS2A was the gene likely to
be involved. That endothelial cells can synthesize NO through a pathway
inducible by TNF- and other cytokines has been recognized and
characterized (36)
. Therefore, inducibility of NOS2A may
be crucial in preventing or delaying pathological alterations in the
microcirculation in diabetes.
The role of `genetic influences' in diabetic retinopathy has
been difficult to define due to differences in patient recruitment
methods, patient selection criteria, risk factors, variation in
ethnicity, and clinical differences in evaluating retinopathy status.
There have been a number of association studies of diabetic
retinopathy, including those implicating involvement of the major
histocompatibility complex (MHC) (37
, 38)
. We are
analyzing various markers from the NO regulation pathway and MHC region
of chromosome 6. Although our studies show some association between
retinopathy status and alleles of genes involved in the NO pathway, no
associations have been observed for the MHC loci examined thus far
(data in preparation). The signal transduction components of the
NO-regulating pathway in the vasculature may provide a starting point
for understanding the complex genetic background of the pathological
processes in diabetic retinopathy.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication October 2, 1998. Revised for publication April 23, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. C. Looker, R. G. Nelson, E. Chew, R. Klein, B. E.K. Klein, W. C. Knowler, and R. L. Hanson Genome-Wide Linkage Analyses to Identify Loci for Diabetic Retinopathy Diabetes, April 1, 2007; 56(4): 1160 - 1166. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Batra, T. Pratap Singh, U. Mabalirajan, A. Sinha, R. Prasad, and B. Ghosh Association of inducible nitric oxide synthase with asthma severity, total serum immunoglobulin E and blood eosinophil levels Thorax, January 1, 2007; 62(1): 16 - 22. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Gonzalez-Gay, J. Llorca, E. Sanchez, M. A. Lopez-Nevot, M. M. Amoli, C. Garcia-Porrua, W. E. R. Ollier, and J. Martin Inducible but not endothelial nitric oxide synthase polymorphism is associated with susceptibility to rheumatoid arthritis in northwest Spain Rheumatology, September 1, 2004; 43(9): 1182 - 1185. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. S. BOUTLIS, M. R. HOBBS, R. L. MARSH, M. A. MISUKONIS, A. N. TKACHUK, M. LAGOG, J. BOOTH, D. L. GRANGER, M. J. BOCKARIE, C. S. MGONE, et al. INDUCIBLE NITRIC OXIDE SYNTHASE (NOS2) PROMOTER CCTTT REPEAT POLYMORPHISM: RELATIONSHIP TO IN VIVO NITRIC OXIDE PRODUCTION/NOS ACTIVITY IN AN ASYMPTOMATIC MALARIA-ENDEMIC POPULATION Am J Trop Med Hyg, December 1, 2003; 69(6): 569 - 573. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Lopez-Nevot, L. Ramal, J. Jimenez-Alonso, and J. Martin The inducible nitric oxide synthase promoter polymorphism does not confer susceptibility to systemic lupus erythematosus Rheumatology, January 1, 2003; 42(1): 113 - 116. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
