| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Studies |
Steno Diabetes Center (J.J., F.P., O.P.K., A.E.K., J.N.), DK-2820 Gentofte, Denmark; and Endocrinology Department, University Hospital (A.P.), 50009 Zaragoza, Spain
Address all correspondence and requests for reprints to: Prof. Jørn Nerup, Steno Diabetes Center, Niels Steensensvej 2, DK-2820 Gentofte, Denmark. E-mail: nerup{at}pres.dk
Abstract
Exposure of human pancreatic islets to a mixture of cytokines induces expression of the inducible nitric oxide synthase (iNOS), impairs ß-cell function, and induces apoptosis. We performed a mutational scanning of all 27 exons of the human NOS2 gene and linkage transmission disequilibrium testing of identified NOS2 polymorphisms in a Danish nationwide type 1 diabetes mellitus (IDDM) family collection. Mutational screening was performed using PCR-amplified exons, followed by single stranded conformation polymorphism and verification of potential polymorphisms by sequencing. The transmission disequilibrium test was performed in an IDDM family material comprising 257 Danish families; 154 families were affected sibling pair families, and 103 families were simplex families. In total, 10 polymorphisms were identified in 8 exons, of which 4 were tested in the family material. A C/T single nucleotide polymorphism in exon 16 resulting in an amino acid substitution, Ser608Leu, showed linkage to IDDM in human leukocyte antigen DR3/4-positive affected offspring (P = 0.008; corrected P = 0.024). No other distorted transmission patterns were found for any other tested single nucleotide polymorphism or constructed haplotypes with the exception of those including data from exon 16. In conclusion, linkage of the human NOS2 gene to IDDM in a subset of patients supports a pathogenic role of nitric oxide in human IDDM.
TYPE 1 DIABETES is the result of loss of insulin production due to selective destruction of the ß-cells in pancreatic islets of Langerhans. The free radical nitric oxide (NO) has been proposed to play an active role in this destructive process, which can be initiated by cytokine exposure of rodent islets of Langerhans (for review, see Refs. 1 and 2). In human islets the involvement of NO in the destructive process is controversial. Cytokine exposure of human islets results in NO production and impaired ß-cell function (3, 4), but inhibitors of NO generation (e.g. aminoguanidine) do not prevent human ß-cell impairment (5), whereas partial ß-cell protection is seen using rodent islets (6). On the other hand, chemical NO donors capable of generating high amounts of NO damage ß-cell function, suggesting a potential deleterious role for NO in human ß-cells (7).
Interleukin-1ß has indirectly been shown to induce superoxide formation due to the interleukin-1ß-mediated induction of heat shock proteins (8) and the superoxide scavenger manganese superoxide dismutase in rat islets (9). Further, we recently demonstrated that overexpression of superoxide scavengers effectively protected cells from the rat insulinoma cell line (RINm5F) from the deleterious effects of cytokines (10). In addition, superoxide may contribute to NO-induced cell dysfunction by participating in the formation of peroxynitrite (11).
The inducible form of NO synthase (iNOS) is the major source of NO production in cytokine-exposed islets. We still consider the human iNOS gene a candidate gene for the development of type 1 diabetes, although its genomic localization at chromosome 17q11.2 (13) has not been highlighted in type 1 diabetes genome scans to date (14, 15, 16).
iNOS production is mainly regulated at the transcriptional level (17), and we have previously tested three promoter polymorphisms of the human iNOS gene without finding any association or linkage to type 1 diabetes (18). The enzyme, however, contains many sites for prosthetic groups and substrate binding (19) and sites affecting dimerization (20), all potentially important for the function of the enzyme. Furthermore, studies of the three different NOS enzymes indicate that even single amino acid changes may have dramatic effects on enzymatic activity (19). We cloned iNOS complementary DNA from islets of Langerhans showing a nearly 100% identity to the iNOS expressed in cytokine-exposed rat hepatocytes and smooth muscle cells (23). The human iNOS gene NOS2 comprises 27 exons, with the transcription start site in exon 2 (E2) and the stop codon in E27 (24).
Here we report the scanning of NOS2 for polymorphisms and testing of the 4 (of a total of 10) most frequent single nucleotide polymorphisms (SNPs) in a Danish family material for linkage to type 1 diabetes by use of the transmission disequilibrium test (TDT).
Materials and Methods
The study population comprised 257 Danish Caucasoid type 1 diabetes families, 1143 individuals in total. One hundred and fifty-four of the families were affected sibling pair families (total of 725 individuals), comprising 455 individuals (317 with type 1 diabetes/138 without) in the offspring generation. The remaining 103 families were simplex families ascertained through the affected child (total of 418 individuals), including 114 unaffected offspring. All diabetic patients were diagnosed according to WHO criteria for type 1 diabetes and were diagnosed before the age of 30 yr (median, 10.7 yr; range, 0–30 yr) (25, 26). The Danish national ethics committee approved the study. A priori, it was decided to stratify the affected offspring for age of onset (below and above the median age of onset of 10.7 yr) and human leukocyte antigen (HLA) risk status (high risk defined here as DR3/4-positive type 1 diabetic individuals) to test whether linkage, if identified, is influenced by HLA status or age of onset.
DNA was prepared by standard procedures from leukocytes
(27). Oligonuclotide sequences for PCR amplification of
the human iNOS gene E1–27 with adjacent intron borders when possible
were designed from genomic human iNOS sequences (see Table 1
). All PCR assays were performed using
20 ng DNA as template, 50 µmol/L deoxy-NTPs, 1.5–2.0 mmol/L
MgCl2, Taq DNA polymerase (0.5–1.0
U/sample), and 1 µmol/L of the primers listed in Table 1, all in
final concentrations in a volume of 20 µL. The cycling conditions
were 95 C for 5 min, followed by 35 cycles of 95 C for 30 s,
annealing at 52–67 C for different primers for 30 s, and
extension at 72 C for 30 s, then a final extension at 72 C for 10
min.
|
For cycle sequencing, either 1)
[
-33P]ATP-kinated primers (Promega Corp., Madison, WI) followed by sequencing according to the
manufacturer’s instructions (Sequenase cycle sequencing kit,
Amersham Pharmacia Biotech), PAGE, and exposure to x-ray
film; or 2) fluorescent deoxy-NTP incorporation (Big Dye Terminator
Chemistry, Perkin-Elmer Corp., Foster City, CA) at the
sequencing reaction, followed by separation of products at the ABI
Prism 310 (Perkin-Elmer Corp.) was used.
Finally, for sequence-verified SSCP polymorphisms, assays were
developed using specific enzymatic digest of the PCR-amplified
polymorphic regions (see Table 1
) using
[33P]GTP incorporation for visualization at
x-ray film after PAGE separation to genotype the entire family
collection. The following enzymes were used: E1, NlaIV
(New England Biolabs, Inc., Beverley, MA); E8,
Fnu4HI (New England Biolabs, Inc.); and E20,
HinfI (Promega Corp.) according to the
manufacturer’s description. For E16, a mutational separation-PCR assay
was established using the forward primer as indicated in Table 1 and
two reverse primers with specific affinity to each allele:
RPWT, GGA GAA AGC TTT ACC TGA ATT TGA AGT TGA GCT
CTT TCA GCA TGA AGC GCG; and RPMT, TTG TTG AGC
TCT TTC AGC ATG AAG AGG A (28). As the different alleles
in the PCR use different primers of different lengths, the alleles
could be distinguished after agarose gel separation.
Statistical analysis
The TDT (29) for linkage and the extended transmission disequilibrium test (ETDT) (30) for linkage with multiallele markers (haplotypes) were carried out. Five percent was chosen as the level of significance. All significant uncorrected P values were corrected for multiple testing by a factor of 3, as we 1) tested the data for each polymorphism unstratified, 2) stratified for HLA, and 3) stratified for age of onset. We did not corrected for the number of tested polymorphisms, as the identified polymorphisms cannot be considered independently of each other.
Results
SSCP scanning results
Of the 29 amplicons, 8 showed SSCP patterns indicative of
polymorphisms, all of which were confirmed by sequencing (see Table 1).
In E20 and E27, 2 polymorphisms were identified. In E20, the two
polymorphisms showed identical patterns within each individual in the
test panel; therefore, a typing assay was only established for one of
the polymorphisms. Furthermore, as the polymorphisms in E7 and E8 and
those in E20 and E22 showed identical patterns within each individual
in the test panel, only the polymorphisms in E8 and E20 were tested in
the complete family material collection.
The frequencies of the mutant variations in E2 and E27 were very
low (<3%) in the test panel, and these polymorphisms were not
analyzed further. Hence, of the 10 identified polymorphisms in 8 exons,
only 4 polymorphisms were tested in the entire type 1 diabetes family
material (see Table 2).
In identifying known sequence variation, the SSCP system in our laboratory has a sensitivity and a specificity of 91% and 92%, respectively.
Linkage transmission disequilibrium analysis
TDT analysis was performed for each of the four polymorphisms. As
shown in Table 2, transmission from heterozygous parents to all
affected offspring was not distorted for any exon polymorphisms tested.
When stratifying for HLA status, however, an increased transmission of
T (PTDT = 0.008; corrected PTDT = 0,024) to
HLA high risk affected individuals was observed for the polymorphism
Ser608Leu in E16. This polymorphism results in an
amino acid change from serine, being polar and having an uncharged R
group, to leucine, having a positively charged R group. No other skewed
transmissions were observed for polymorphisms in other exons.
|
).
The transmission pattern for E1E8 was distorted for some haplotypes,
but this distortion was in a few rare haplotypes (data not shown).
|
Although some differences in the sensitivity to cytokine-mediated ß-cell destruction between human and rodent islets have been described, accumulating evidence supports that NO production may be a pathogenetic factor, and NOS2 might be a susceptibility gene for type 1 diabetes (32, 33).
SSCP screening of the human iNOS gene identified 10 polymorphisms, of which none was previously reported to our knowledge. Four of these, situated in different exons, were examined in a Danish type 1 diabetes family collection. The remaining 6 polymorphisms were only examined within the SSCP screening panel, as they were either in linkage disequilibrium with one of the four tested polymorphisms or had a low frequency. The only distorted transmission in TDT was the increased transmission of the T in E16 to affected individuals (PTDT = 0.008) with the HLA high risk genotype. This transmission pattern was confirmed after E16E20 haplotype construction (PETDT = 0.045). Despite the overall ETDT analysis of E1E8 in some groups with PETDT values below 0.05, these are considered spurious findings, as they are based upon few transmissions of certain haplotypes.
Previous data, e.g. from a recent type 1 diabetes genome scan (15), suggest that it is not possible to predict whether linkage of a putative minor susceptibility gene is present among patients with a high or low risk HLA genotype. On the other hand, currently the use of genome scans to identify minor type 1 diabetes susceptibility loci are complicated, as they have difficulties confirming each other and previously identified loci (14, 15, 16), e.g. the chromosomal localization of the human NOS2 gene has not been highlighted in type 1 genome scans to date. This illustrates that identifying minor susceptibility genes by genome scans either requires a much larger number of informative families or pooling of existing datasets and supports the concept that the use of both a candidate gene approach and genome scans may be needed to identify the minor susceptibility genes in type 1 diabetes.
The amino acid change in E16 may be of functional interest. The iNOS protein is a catalytic enzyme with two domains. The N-terminal oxygenase domain of all NOS-enzymes bind heme, H4B, and L-arginine, receive the electrons from the C-terminal reductase domain binding flavine mononucleotide (FMN) and flavine adenine dinucleotide, and accept electrons from NADPH (19). Functionally, iNOS differs from the constitutive forms of NOS (eNOS and nNOS) in being Ca2+ independent. This is partly explained by the strong binding of calmodulin (CaM) to iNOS, thereby being insensitive to changes in Ca2+ ion concentrations (17). The CaM-binding site lies within the linker region between the N- and C-terminal domains. Regulation of NO synthase activity by CaM has been shown to occur through the control of electron transfer from FMN to heme across the domain-domain interface (34). In addition, the constitutive NOSs contain an additional 40- to 50-amino acid insert in the middle of a conserved region of the FMN-binding subdomain compared with iNOS (35). It has been proposed that this region acts as an autoinhibitory domain, competing with CaM for space on the interdomain linker and thereby increasing the Ca2+ dependency (36). Finally, by removing most of this insert from nNOS, it was shown to be of functional importance, as the deleted mutants retained maximal NO activity at lower concentrations of free Ca2+ compared with the wild-type (37). The identified C/T polymorphism in E16 is located only six amino acids N-terminally from the deletion reported by Daff et al., and specific functional tests of the C/T shift need to be performed to substantiate the putative importance of the Ser608Leu polymorphism.
Taken together, these results suggest an increased risk for type 1 diabetes among HLA DR3/4-positive individuals with a T in position 150 in E16 of the iNOS gene in a Danish type 1 diabetes population. This result suggest an interaction between the iNOS locus and the HLA region and a role for iNOS in the pathogenesis of human type 1 diabetes, but the findings need to be confirmed in other family materials. Finally, the putative functional implication of the amino acid substitution close to the FMN region requires further elucidation.
Acknowledgments
For participating departments in the Danish Study Group of Diabetes in Childhood, see Ref. 25 . For contributing members of the Danish Insulin-Dependent Diabetes Mellitus Epidemiology and Genetics Group, see Ref. 26 . The authors thank Anna-Margrethe Flarup for expert technical assistance.
Footnotes
1 This work was supported by the Danish Diabetes Foundation, The Poul
and Erna Sehested Foundation, the Juvenile Diabetes Foundation
International (DK-96-012), and European Union Grant BMH4-97-2311. 
2 Recipient of research grant for the University of Copenhagen.
3 Supported by Grant 97/5136 from Fondo de Investigacion Santaria,
Spanish Ministry of Health and Consume.
4 Recipient of Juvenile Diabetes Foundation International Fellowship
Grant 3–1999-21.
Received September 13, 2000.
Revised February 21, 2001.
Accepted February 27, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  H. R. Ferguson, C. P. Wild, L. A. Anderson, S. J. Murphy, B. T. Johnston, L. J. Murray, R.G. P. Watson, J. McGuigan, J. V. Reynolds, and L. J. Hardie Cyclooxygenase-2 and Inducible Nitric Oxide Synthase Gene Polymorphisms and Risk of Reflux Esophagitis, Barrett's Esophagus, and Esophageal Adenocarcinoma Cancer Epidemiol. Biomarkers Prev., March 1, 2008; 17(3): 727 - 731. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
