Christian Boitard

Clinical Spectrum Associated with Hepatocyte Nuclear Factor-1β Mutations

Context Maturity-onset diabetes of the young type 5 (MODY5) is an uncommon variant of a dominantly inherited disease associated with mutations in the hepatocyte nuclear factor-1 (HNF-1) gene. We know little about its spectrum except that it may include urogenital abnormalities. Contribution This multicenter study describes 8 probands with MODY5 and known nondiabetic kidney disease and 5 offspring. Each proband had a different mutation in the HNF-1 gene. Probands and offspring had various renal abnormalities, including dysplastic kidneys and renal cysts. Some also had genital tract abnormalities, pancreatic atrophy, and abnormal liver enzyme levels. Implications Maturity-onset diabetes of the young type 5 is genetically and phenotypically heterogeneous. The Editors The prevalence of diabetes mellitus is estimated to be 6% to 7% in western industrialized countries; type 2 diabetes mellitus accounts for 90% of the cases (1). Maturity-onset diabetes of the young (MODY) is defined by the occurrence, typically before age 25 years, of nonketotic diabetes mellitus due to a primary defect of insulin secretion, along with autosomal dominant inheritance. With the exception of MODY2, which is due to mutations in the glucokinase gene, the identified MODY subtypes are related to mutations of transcription factor genes: hepatocyte nuclear factor-4 (HNF-4) for MODY1, HNF-1 for MODY3, insulin promoter factor-1 for MODY4, HNF-1 for MODY5, and neurogenic differentiation factor-1 for MODY6. The subtypes of MODY may account for 2% to 5% cases of type 2 diabetes, with types 3 and 2 representing 65% and 15% of these cases, respectively. The other subtypes of MODY are thought to be rare (2). Maturity-onset diabetes of the young type 5 was initially described in a Japanese family as the association of early-onset diabetes and nephropathy (3). Heterogeneous phenotypes, including diabetes, renal abnormalities, and genital malformations, were subsequently described through isolated case reports in white and Japanese populations (3-16). Currently, all but 2 families (6, 16) with MODY5 demonstrated autosomal dominant inheritance of private mutations. In our report, using a standardized evaluation, we describe clinical and genetic findings in 13 patients from 8 unrelated families with novel HNF-1 mutations. Detailed phenotypic analysis underlines the systemic spectrum of the disease and its wide variability, leading to different modes of presentation. We also confirm that MODY5 may occur because of de novo mutation in HNF-1. These findings may help to define criteria for HNF-1 gene testing. Methods Patients Participants were recruited from the patients of 2 departments of diabetes, 1 internal medicine department, and 1 nephrology department, on the basis of the following clinical characteristics: 1) diabetes (fasting plasma glucose level 7.0 mmol/L [126 mg/dL]), suggesting a MODY phenotype if diabetes occurred before 40 years of age, if the patient was not obese (body mass index < 30 kg/m2), and the patient did not have islet-cell antibodies and glutamic acid decarboxylase autoantibodies; 2) impaired renal function (creatinine clearance < 1.34 mL/s [<80 mL/min]) without diabetic retinopathy and albumin excretion greater than 0.5 g/d (17); and 3) kidney structural abnormalities (reduced kidney size, presence of cysts, or both). A family history of diabetes was not a criterion. A total of 20 unrelated white patients were identified from clinic lists and screened for HNF-1 mutation: 13 men and 7 women with a mean (sd) age of 26 8 years, a mean body mass index of 23.4 2.7 kg/m2 at onset of diabetes, and functional and structural renal abnormalities at presentation. The Ethics Committee of Necker Hospital, Paris, France, approved the study, and all participants gave written informed consent. Clinical Evaluation Clinical history of diabetes mellitus and renal involvement were recorded by using a standardized examination of the patients and their relatives. Creatinine clearance was calculated according to the Cockcroft-Gault formula (18). Imaging studies of the kidneys, consisting of ultrasonography and intravenous urography or computed tomography, and renal biopsy specimens were reviewed by a nephrologist and a renal pathologist. Genital tract abnormalities were assessed by ultrasonography. Endogenous insulin secretion was assessed by measuring C-peptide plasma concentration (Bio-Rad Specific C-Peptide, Bio-Rad, Marne la Coquette, France) before and 6, 10, and 15 minutes after intravenous injection of 1 mg of glucagon (19). Pancreas structure was assessed by computed tomography. Pancreas exocrine function was evaluated by measuring fecal fat excretion and elastase concentration (Schebo-Biotech, Guiessen, Germany). Liver biopsy was performed in 3 patients whose liver test results were persistently abnormal. Mutation Analysis of the HNF-1 Gene Genomic DNA was extracted from peripheral blood samples by standard procedures. The minimal promoter, the coding region of the 9 exons, and exon-intron boundaries of the HNF-1 gene were screened for mutations by direct sequencing as previously described (3). Direct diagnosis of the mutations identified in probands was offered to first-degree relatives (parents, siblings, and offspring) regardless of their clinical status. Results Molecular Analysis A mutation of HNF-1 was found in 8 of the 20 unrelated patients. Among 19 of 35 first-degree relatives from 6 unrelated families tested, 5 of 8 offspring had inherited a mutation. None of the 8 mutations have been reported to date. All are located in the DNA-binding domain (20). Five are point mutations resulting in amino acid substitutions (also called missense mutations) that affect residues conserved in HNF-1 human, pig, mouse, rat, xenopus, and salmon sequences. Two other mutations are nonsense mutations resulting in a truncated protein lacking the 3-part of the DNA-binding domain and the C-terminal transactivation domain. The last mutation is localized in the splice donor site of intron 2 at the highly conserved +1 position. Another base change (GA) at the same position was described in a Japanese family with MODY5 (8). No nucleotide variant was detected in 212 control chromosomes of unrelated nondiabetic white participants and in 170 chromosomes of patients with classic type 2 diabetes. In 4 families (families 1, 5, 6, and 8), cosegregation of a HNF-1 mutation and MODY5 phenotype was consistent with dominant inheritance because 5 affected offspring inherited the mutation detected in the proband. A de novo mutation was demonstrated in 2 probands (patient II-1, family 6, and patient II-1, family 7). In each case, the probands parents were not carriers of the mutation. Parental relationships were confirmed by using microsatellite markers (data not shown). Two relatives in family 3 (patients I-2 and II-1) and 2 relatives in family 6 (patients I-1 and I-2) had diabetes mellitus but did not exhibit the corresponding mutation of HNF-1. Clinical Phenotype Diabetes was present in all probands and in 2 offspring (patient III-1, family 1, and patient II-1, family 8) with a mutation. Five patients presented with clinical symptoms, including severe metabolic decompensation in 2 patients, at age 1 year (patient II-1, family 8) and 13 (patient II-1, family 6) years, respectively. In the 5 other patients, diabetes was found by routine plasma glucose measurement. Clinical characteristics were consistent with a defect of insulin secretion: All patients were lean at diagnosis, ketosis was observed in 2 patients, and progression of the disease was observed in 3 patients who ultimately required insulin therapy. Endogenous insulin response to intravenous glucagon, assessed in 7 of the 8 probands, disclosed heterogeneous degrees of impairment, but residual insulin secretion was detectable in all probands. In patient II-1, family 6, endogeneous insulin secretion was measurable 22 years after diabetes was revealed by severe ketoacidosis. No patient had diabetic retinopathy or neuropathy. Pancreas atrophy was found in 5 of the 6 patients assessed by abdominal computed tomography (Table 1 and Figure). No pancreas calcification or cyst was noticed. No patient exhibited symptoms of overt pancreatic exocrine deficiency. However, of the 7 probands tested, 6 had a subclinical defect of pancreas exocrine function (Table 1 and Figure). Table 1. Characteristics of Mutations, Diabetes, Pancreas Exocrine Function, and Liver Tests in 8 Probands with Hepatocyte Nuclear Factor-1 Mutations Figure. Pancreas and kidney abnormalities in patients with hepatocyte nuclear factor-1 ( HNF-1 ) mutation. A B C. A arrow arrowheads B C. arrowheads arrows D. E F. HNF-1 E. F. arrow * Renal involvement was found in all probands (Table 2) and 4 offspring. In the probands, age at recognition of kidney disease ranged from 18 to 41 years. All exhibited structural kidney abnormalities, including decreased kidney size, ranging from 85 to 105 mm in height with global cortical loss, and a cystic pattern (Figure). Pelvicaliceal abnormalities were observed in 6 of the 7 patients who were tested and consisted of clubbing or tiny diverticulae of the calices or mild pelvic dilatation without obstructive uropathy (Figure). All probands also had mild to moderate renal failure with creatinine clearance ranging from 0.48 to 1.28 mL/s (29 to 77 mL/min). On follow-up, the mean annual decrease of creatinine clearance ranged from 0 to 0.03 mL/s (0 to 2 mL/min). It averaged 0.02 mL/s per year (1.1 mL/min per year) in the 4 patients who were followed for 14 to 16 years (Table 2). Since the clinical, biological, and radiologic characteristics of renal disease made the diagnosis of diabetic glomerulopathy very unlikely, a kidney biopsy was performed in 6 probands. No proband exhibited diabetic glomerulosclerosis. Interstitial fibrosis was observed in all biopsy specimens. Enlarged glomeruli were observed in 4 cases, associated with glom

Prevention of diabetes in NOD mice treated with antibody to murine IFNγ

The NOD mouse is studied as an animal model of human insulin-dependent diabetes mellitus (IDDM). To evaluate the role of IFN gamma in the pathogenesis of the disease, we have studied the effect of anti-IFN gamma mAb on the expression of insulitis and clinical diabetes. Treatment of mice with anti-IFN gamma mAb prevented the induction of early IDDM by cyclophosphamide as well as the adoptive transfer of diabetes by spleen cells from diabetic NOD mice. The protection against induction of diabetes by cyclophosphamide was observed in animals treated with the anti-IFN gamma mAb within 24 h following the first cyclophosphamide injection but not in animals in which mAb treatment was started 7 days later. Transfer of disease was prevented both in adult irradiated and in newborn recipients. The absence of clinical signs in these mice was corroborated by a significant reduction of both the extent and severity of insulitis. Over-expression of Ia antigen on endothelial cells lining the islets was also considerably reduced in mice treated with mAb. These data strongly suggest a role for IFN gamma during the autoimmune process leading to beta cell destruction in diabetes and prompt further investigation of the use of such antibodies in the immunoprevention of IDDM.

POEMS syndrome presenting as systemic sclerosis: Clinical and pathologic study of a case with microangiopathic glomerular lesions

A rare form of plasma cell dyscrasia characterized by the various association of polyneuropathy, organomegaly, endocrinopathy, monoclonal protein, and skin changes has been termed POEMS syndrome. The pathogenesis of the multisystemic features of this syndrome remains unclear. Herein is reported a case of POEMS syndrome with striking clinical similarities with scleroderma, and microangiopathic glomerular lesions, as well as diffuse perivascular non-amyloid deposits, which could explain certain features of the syndrome, including peripheral nerve demyelination. It is proposed that a pathogenic role might be played by a non-immunoglobulin vasculotoxic component.

Evidence of CD4+ regulatory T cells in the non-obese diabetic male mouse

SummaryThe NOD mouse, which shows many features of human IDDM, is extensively used to evaluate the role of T lymphocytes in the pathogenesis of autoimmune diabetes. The development of diabetes in this model appears to be controlled by a finely tuned immunoregulatory balance between autoaggressive T cells and regulatory immune phenomena, the disruption of which may result in destruction of insulin-secreting cells. The absolute requirement of sublethal irradiation to permit transfer of the disease to non-diabetic adult syngeneic mice provides indirect evidence for the presence of regulatory T cells in non-diabetic NOD mice. We have previously reported that the reconstitution of irradiated recipients by CD4 + T cells from nondiabetic female NOD mice blocks the transfer of diabetes by spleen cells from diabetic donors. We now report evidence that anti-CD4 monoclonal antibodies can substitute for irradiation in rendering adult NOD male mice susceptible to diabetes transfer by diabetogenic spleen cells. Efficient diabetes transfer can be achieved in non-irradiated adult NOD recipients provided they are thymectomized and CD4 + T-cell depleted prior to the transfer. The role of thymectomy is to limit T cell regeneration after anti-T cell monoclonal antibody challenge. Our data confirm that regulatory CD4 + T-cells, which efficiently counterbalance diabetogenic cells, are present in adult NOD male animals.

Family study of linkage disequilibrium between TAP2 transporter and HLA class II genes. Absence of TAP2 contribution to association with insulin-dependent diabetes mellitus.

The polymorphic TAP1 and TAP2 genes encode a transporter protein required for delivery of cytosolic peptides to class I molecules in the endoplasmic reticulum. Associations have been observed between TAP2 alleles and predisposition to autoimmune diseases such as IDDM but their interpretation has been complicated by the existence of LD between TAP2 and HLA class II loci, and conclusions are still contradictory. In order to precisely define LD on class II haplotypes, we performed an extensive familial analysis. A total of 466 individuals from 55 normal families and 49 IDDM multiplex families was studied, providing information on 420 independent haplotypes. The IDDM-predisposing DRB1*03 and DRB1*04 alleles were in strong negative LD with TAP2-B (delta = -0.035 and -0.034, respectively), and positive LD with TAP2-A (delta = + 0.055 and + 0.012). Positive LD was also found between TAP2-B and DRB1*01 and TAP2-C and DRB1*11 alleles. We then addressed the question of whether TAP2 is an independent additional IDDM-protective or predisposing genetic factor. No TAP2 effect was evidenced when considering DRB1*03 and/or 04 patients. A decreased TAP2-B phenotype frequency was observed in DRB1*03- and DRB1*04-negative IDDM patients compared with DRB1*03- and DRB1*04-negative normal controls (38.6% vs 63%, pc < 0.05), but was probably related to a combination of different weak LD between DRB1 and TAP2 alleles. It thus appears that there is no primary association between TAP2 alleles and IDDM. However, TAP polymorphism may allow us to define particular extended HLA haplotypes involved in susceptibility to autoimmune diseases.

Origin of the beta cells of the islets of Langerhans is further questioned by the expression of neuronal intermediate filament proteins, peripherin and NF-L, in the rat insulinoma RIN5F cell line

Intermediate filament proteins of the rat insulinoma RIN5F cell line were characterized. Two-dimensional gel analysis followed by immunostaining of proteins demonstrated that these cells express both peripherin and the low-molecular-mass neurofilament protein (NF-L); this was confirmed for peripherin by immunohistochemistry, peptide analysis and Northern blot. No expression of these proteins could be detected with these same methods either in the adult pancreas or in the tumor at the origin of the cell line, although such expression was apparent on sections of rat pancreas at embryonal day 16. These results were compared to those obtained on the rat pheochromocytoma PC12 cell line: expression in the adrenal medulla of the embryo, no expression either in the adult tissue or in the tumor, but solely in the derived cell line. The expression of neuronal intermediate filament proteins in the rat insulinoma RIN5F cell line is discussed in relation to its similarity in the rat pheochromocytoma PC12 cell line, and its meaning as to the developmental cell lineage; an ectodermal origin is suggested for the pancreatic islet cells.

Extensive study of DRB, DQA, and DQB gene polymorphism in 23 DR2-positive, insulin-dependent diabetes mellitus patients

To gain insight into the HLA subregions involved in protection against insulin-dependent diabetes mellitus (IDDM) we investigated the polymorphism of HLA-DR and -DQ genes in 23 DR2 IDDM patients. Results show the following. (1) Fourteen patients (61%) possess the DRB1, DRB5, and DQB1 alleles found in DRw16/DQw5 healthy people. These data contrast with the 5% of DRw16 normally found in DR2 populations and are in agreement with former observations supporting that the DRw16 haplotype is not protective. (2) Nine DR2 patients, i.e., 39% versus 95% in published DR2 controls, possess the DRB alleles found in DRw15 unaffected people. Among them, six patients have also DQA1 and DQB1 alleles identical to those found in DRw15/DQw6 healthy individuals. These data confirm that the DRw15/DQw6 haplotype is protective but indicate that none of the DR or DQ alleles, alone or in association, confers an absolute protection. (3) Our most striking results concern the very high frequency of recombinant haplotypes among the DRw15 patients: 3 of 9. In these three patients recombinations led to the elimination of both DQB1 and DQA1 alleles usually associated with DRw15. This strongly suggests that the occurrence of IDDM in these DRw15 patients is due to the absence of the usual DQ product and thus reinforces the assumption that DQ rather than DR region is involved in the protection conferred by the DRw15/DQw6 haplotype. Finally, analysis of the non-DRw15 haplotypes in heterozygous patients showed that IDDM can occur in the absence of any DQ alpha beta heterodimer of susceptibility.

Islet cell antibody heterogeneity among Type 1 (insulin-dependent) diabetic patients

SummaryConventional detection of islet cell antibodies is based on indirect immunofluorescence performed on frozen human pancreas sections. The number and nature of epitopes recognized by antibodies detected by such techniques are unknown. To determine the existence of heterogeneous fluorescence patterns of islet cell antibodies on pancreatic sections, we selected two sera showing a distinctive granular fluorescence. We then tested random sera from patients with Type 1 (insulin-dependent) diabetes mellitus for their ability to block ultimate binding of fluorescein isothiocyanate-labelled immunoglobulins purified from these two sera with a characteristic granular pattern. Among 102 subjects with recent-onset Type 1 diabetes, 79 had detectable anti-islet cell antibodies; 21 showed complete blockade of the binding to islets of granular fluorescein isothiocyanate-labelled immunoglobulins. The majority of these 21 patients were women carrying a DR3 non-DR4 DQB1*0201 allele, with under-representation of DRB1*0402 and 0405. Discrimination between islet cell antigenic specificities may help in identifying islet cell autoantibodies in autoimmune Type 1 diabetes.

Nonneoplastic circulating sézary-like cells in cutaneous T-cell lymphoma. Ultrastructural, immunologic, and T-cell receptor gene-rearrangement studies

Approximately 67% of peripheral blood lymphocytes in a case of nonmycosis fungoides‐type cutaneous T‐cell lymphoma, exhibited Sézary‐like changes. Immunotyping showed a helper phenotype with an abnormally faint expression of the T3 membrane antigen, while quantitative electron microscopic study was suggestive of neoplasia. However, T‐cell receptor gene study did not show any DNA rearrangement in Ficoll‐separated blood lymphocytes, whereas clonal T‐cell proliferation was simultaneously evidenced in cutaneous (non‐Sézary‐like) tumor cells. It is concluded that reactive nonneoplastic Sézary‐like cells may be present not only in various benign conditions, as previously known, but also in T‐cell lymphomas. This study provides further evidence for the nonspecificity of Sézary‐like changes, and stresses the utility of T‐cell receptor gene rearrangement study for diagnostic and prognostic procedures during the course of T‐cell lymphoproliferative disorders.

Normal renal function 8 to 13 years after Cyclosporin A therapy in 285 diabetic patients

Cyclosporin A (CyA) may induce acute nephrotoxicity. The question has been raised of the possible long‐term unfavorable course of CyA‐induced lesions. Advantage was taken of a large cohort of diabetic patients treated for several months using moderate CyA dosage to evaluate the long‐term evolution of renal function in such patients.