Glucose transporters of rat peripheral nerve. Differential expression of GLUT1 gene by Schwann cells and perineural cells in vivo and in vitro.

Expression of GLUTs in rat peripheral nerve was first studied at the mRNA level with Northern transfer analysis with cDNAs specific for GLUT1, GLUT2, GLUT3, and GLUT4. GLUT1 mRNA was the only GLUT mRNA detectable in rat sciatic nerve. In situ hybridization localized this mRNA to the perineurium and to some endo- and epineurial capillaries. Indirect immunofluorescence stainings demonstrated that GLUT1 protein epitopes were concentrated primarily in the perineurium and endoneurial capillaries. Also, some Schwann cells, a few epineurial capillaries, and medium-sized blood vessels showed a faintly positive immunoreaction. All cell types present in primary cultures initiated from rat sciatic nerve (perineurial cells, Schwann cells, and fibroblasts) expressed GLUT1 protein in vitro. Thus, Schwann cells, which expressed GLUT1 only occasionally at a low level in vivo, have the potential to express GLUT1 at a markedly higher level under cell culture conditions. Incubation of the cultures in 25 mM D-glucose for 7 days caused a 39% reduction in the amount of immunodetectable GLUT1 protein, and a marked (34%) decrease of GLUT1 mRNA compared with cultures incubated in 5.5 mM D-glucose. Interestingly, the reduction of [3H]-2-DG uptake in the same cultures exceeded 70%, suggesting that the reduced amount of GLUT1 protein alone did not explain the marked reduction in glucose uptake in these cultures. Immunostaining of the cell cultures suggested that perineurial cells were the main target for the glucose-induced decrease of GLUT1 protein.

Increased matrix gene expression by glucose in rat neural connective tissue cells in culture.

The effects of different glucose concentrations on the expression of extracellular matrix genes were examined in primary cell cultures initiated from sciatic nerves of nondiabetic Sprague-Dawley rats. The cells were incubated in medium containing 5.5, 15, or 25 mM D-glucose, and the expression of type I and IV collagens, laminin, and fibronectin genes was examined at mRNA steady-state levels by Northern hybridizations. Incubation of cell cultures, consisting of Schwann cells, perineurial cells, and fibroblasts, in high glucose concentrations (15 or 25 mM D-glucose) resulted in elevation of pro-alpha 1(IV), pro-alpha 2(IV), and pro-alpha 1(I) collagen chain and fibronectin mRNAs after only 3 days of incubation, whereas laminin B2 chain mRNA levels appeared unaltered. These observations suggest that hyperglycemia may partially contribute to basement membrane thickening in peripheral nerves of diabetic individuals by increasing the expression of genes coding for basement membrane components, e.g., type IV collagen.

Cow's milk formula feeding induces primary immunization to insulin in infants at genetic risk for type 1 diabetes.

Insulin autoantibodies (IAAs) often appear as the first sign of islet cell autoimmunity in prediabetic children. Because cow's milk contains bovine insulin, we followed the development of insulin-binding antibodies in children fed with cow's milk formula. Bovine insulin- and human insulin-binding antibodies by enzyme immunoassay and IAA by radioimmunoassay were analyzed in 200 infants carrying HLA-DQB1*0302 but no protective alleles who participated in a Finnish population-based birth-cohort study. Based on the prospectively registered information, the first 100 infants enrolled in the study who were exposed to cow's milk formula before age 12 weeks and the first 100 infants enrolled in the study who were exclusively breast-fed for longer than their first 12 weeks of life were selected for the present study. Also, 11 children from the birth cohort who developed at least two diabetes-associated autoantibodies, 98 children with newly diagnosed type 1 diabetes, and 92 healthy children were studied. We found that the amount of IgG-antibodies binding to bovine insulin was higher at age 3 months in infants who were exposed to cow's milk formula than in infants who were exclusively breast-fed at that age (median 0.521 vs. 0.190; P < 0.0001). The antibodies binding to bovine insulin cross-reacted with human insulin. None of these infants tested positive for IAA. The levels of bovine insulin-binding antibodies declined in both groups at ages 12 and 18 months, whereas in the 11 children with at least two diabetes-associated autoantibodies the levels increased during the follow-up period (P < 0.0001). IgG antibodies correlated with IgG2 antibodies binding to bovine insulin (r = 0.43, P = 0.004) and IAA (r = 0.27, P = 0.02) in diabetic children, but not in healthy children. Cow's milk feeding is an environmental trigger of immunity to insulin in infancy that may explain the epidemiological link between the risk of type 1 diabetes and early exposure to cow's milk formulas. This immune response to insulin may later be diverted into autoaggressive immunity against beta-cells in some individuals, as indicated by our findings in children with diabetes-associated autoantibodies.

Enterovirus infection as a risk factor for beta-cell autoimmunity in a prospectively observed birth cohort: the Finnish Diabetes Prediction and Prevention Study.

Previous studies suggest that enterovirus infections may initiate and accelerate beta-cell damage years before the clinical manifestation of type 1 diabetes. We have now analyzed the role of enterovirus infections in the initiation of autoimmunity in children who have tested positive for diabetes-associated autoantibodies in a prospective study starting at birth (the Finnish Diabetes Prediction and Prevention Study). The frequency of enterovirus infections was studied using both serology and testing for the presence of enterovirus RNA in the sera of 21 children who developed and retained autoantibodies and in 104 control subjects chosen from the same study cohort and matched for the time of birth, sex, and HLA alleles determining genetic diabetes susceptibility. Sample intervals were taken as basic units of follow-up, to which the observed number of infections was adjusted. Enterovirus infections were detected in 26% of sample intervals in the case subjects and in 18% of the sample intervals in the control children (P = 0.03). A temporal relationship between enterovirus infections and the induction of autoimmunity was found; enterovirus infections were detected in 57% of the case subjects during a 6-month follow-up period preceding the first appearance of autoantibodies compared with 31% of the matched control children in the same age-group (odds ratio 3.7, 95% CI 1.2-11.4). The frequency of adenovirus infections did not differ between the patient and control groups. Our data imply that enterovirus infections are associated with the development of beta-cell autoimmunity and provide evidence for the role of enteroviruses in the initiation of beta-cell destruction.

A fibroblast cell line cultured from a hypertrophic scar displays selective downregulation of collagen gene expression: barely detectable messenger RNA levels of the pro alpha 1(III) chain of type III collagen.

The present study was designed to investigate the expression of type I, III and VI collagens by a fibroblast cell line initiated from a hypertrophic scar. The same tissue has previously been demonstrated to display markedly elevated expression of type I and III collagen mRNAs in vivo. Unexpectedly, slot-blot and Northern hybridizations revealed a barely detectable steady-state level of pro alpha 1(III) collagen chain mRNA in cultured hypertrophic scar fibroblasts. The levels of pro alpha 1(I) and alpha 2(VI) collagen chain mRNAs were essentially the same in fibroblasts cultured from hypertrophic scar and in fibroblasts cultured from normal skin. However, Northern blot analyses indicated that the ratio of 5.8 kb to 4.8 kb species of pro alpha 1(I) collagen mRNA was slightly reduced in fibroblasts originating from the hypertrophic scar compared to that in normal fibroblasts. When normal fibroblasts were incubated in conditioned medium from hypertrophic scar cultures, the expression of pro alpha 1(III) collagen chain mRNA decreased to a markedly lower level. Our studies suggest that collagen synthesis by fibroblasts in hypertrophic scars is stimulated by humoral factors which are active only in vivo. Furthermore, the results suggest that fibroblasts cultured from hypertrophic scar display a selective downregulation of different collagen genes and that this downregulation is exerted through an autocrine mechanism.

Connective tissue metabolism in diabetic peripheral nerves.

Diabetes mellitus is associated with multiple connective tissue changes, such as generalized thickening of basement membranes. These alterations are suspected of contributing to the development of diabetic long-term complications encountered in many organs, including kidney, eye and peripheral nerves. The latter tissue, however, has gained relatively little attention with respect to connective tissue changes associated with diabetes. The morphological alterations of connective tissue in the diabetic peripheral nerve include thickening of basement membranes, increased diameter of endoneurial collagen fibrils, and accumulation of microfibrillar material. Recent studies have further elucidated the changes in the extracellular matrix of diabetic nerves and the molecular mechanisms underlying these alterations. For instance, elevated glucose concentrations modulate the expression of several proteins of the extracellular matrix in cultured nerve-derived connective tissue cells. In this article, we review the recent progress in the field of connective tissue alterations in diabetes and particularly in the diabetic peripheral nerve.

Expression of glucose transporter 1 in adult and developing human peripheral nerve.

Northern hybridization of total RNA isolated from adult human sciatic nerve demonstrated a readily detectable hybridization signal for glucose transporter 1 (GLUT 1) mRNA. Western blot analysis demonstrated that GLUT 1 proteins extracted from adult human and from mature rat sciatic nerves had different electrophoretical mobilities. The migration positions of human and rat GLUT 1 proteins corresponded to 60-70 kDa and 55-60 kDa, respectively, as estimated by markers with known molecular masses. Indirect immunofluorescence staining localized GLUT 1 to the perineurium in the adult human sciatic nerve. Only a few endoneurial capillaries of human adult nerve stained positively for GLUT 1, which was in contrast to rat peripheral nerve where most endoneurial capillaries were positive for GLUT 1. In developing human nerves, the staining pattern for GLUT 1 was markedly different from that of the adult nerves: at 14 weeks, the perineurial cells were entirely negative for GLUT 1. Between 22 and 26 weeks of gestation, the staining reaction for GLUT 1 in the perineurium became markedly more prominent, and by 35 weeks the intense perineurial staining resembled that observed in the adult human nerves. In contrast to adult nerves, both endo- and epineurial blood vessels stained intensely for GLUT 1 in the fetal samples. Thus, the immunoreactivity for GLUT 1 in the perineurium seems to increase concomitant with the maturation of barrier properties of perineurium, whereas the transient expression of GLUT 1 in the vasculature of developing nerve may have a specific role in the proliferating endothelial cells.

Diabetes induces the formation of large diameter collagen fibrils in the sciatic nerves of BB rats.

Diabetes mellitus (DM) is known to be associated with widespread connective tissue changes. However, the connective tissue of peripheral nerves in diabetes has gained little attention. Thickening of the basement membranes of Schwann, perineurial and endothelial cells suggests a perturbation in the metabolism of type IV collagen. We studied the ultrastructure of endoneurial collagen fibrils in the sciatic nerves of spontaneously diabetic BB rats and found that the fibrils in the diabetic rats were significantly thicker than in the age and sex matched non-diabetic BB rats, both in the proximal (52.6 vs. 46.1 nm) and the distal part (52.4 vs. 45.5 nm) of the nerve, respectively. The mechanism for the thickening of endoneurial collagen fibrils remains unknown, but conceivably involves changes in the biochemical composition of the nerve connective tissue matrix: e.g. altered metabolism of type I and/or type III collagen, increased glycosylation of the collagen molecules, or changes in the glycosaminoglycan content of the ground substance.

Type VI collagen gene expression in experimental liver fibrosis: quantitation and spatial distribution of mRNAs, and immunodetection of the protein.

Type VI collagen is a minor but essential matrix component in the liver. In this study, we utilized an acute and a chronic injury model to clarify the process of liver fibrosis in rats by administration of carbon tetrachloride. Collagen gene expression, with particular emphasis on type VI collagen, was studied by molecular hybridization techniques. The alpha 2(VI) collagen mRNA levels were markedly elevated on day 3 of acute injury and were approximately at the same high level at 7 and 14 weeks of chronic injury, as determined by Northern hybridizations and slot-blot analyses. Marked enhancement of type I collagen gene expression was similarly noted at these time points. The activation of collagen gene expression in acute injury, as determined by in situ hybridization, was particularly prominent in the vicinity of the central veins. Indirect immunofluorescence demonstrated marked accumulation of type VI collagen protein as early as day 3 of acute injury, and the reaction appeared to be initiated in the proximity of central veins. These results indicate that type VI collagen gene expression, together with other connective tissue components, including type I collagen, is activated in the early stages of the fibrotic process. Type VI collagen accumulation may contribute to the distorted architecture and functional impairment of the liver in hepatic fibrosis.

Extracellular matrix of peripheral nerves in diabetes.

Peripheral nerves are susceptible to develop multiple changes in their morphology and biochemical composition as consequences of diabetes mellitus. This review focuses on diabetes-induced alterations of the extracellular matrix of the peripheral nerves, and on the potential molecular mechanisms causing these changes. The interest towards the extracellular matrix of peripheral nerves of diabetic patients is highlighted by the fact that the extracellular matrix does not only mechanically support the cells which it surrounds, but it also regulates their behavior through specific interactions mediated via molecules on the cell surface, such as integrin receptors and cell surface proteoglycans. Thus, changes in the structure and composition of the extracellular matrix may alter cellular functions in multiple ways. At the ultrastructural level, these changes include e.g. thickening of vascular, perineurial and Schwann cell associated basement membranes; accumulation of microfibrillar material in the vicinity of perineurial cells; and increased diameter of endoneurial collagen fibrils. At the molecular level, the changes may be associated with altered metabolism of various collagen types, such as type I, III, IV and VI collagens.

Normal and hypertrophic scars: quantification and localization of messenger RNAs for type I, III and VI collagens.

The expression of type I, III and VI collagens was studied in nine normal and two hypertrophic scars using slot-blot and in situ hybridization techniques. Slot-blot hybridization indicated that the steady-state levels of pro alpha 1(I) and pro alpha 1(III) collagen chain mRNAs were moderately elevated in two of the nine normal scars, whereas the two hypertrophic scars analysed displayed markedly elevated mRNA levels when compared with normal skin. The mRNA levels of alpha 2(VI) collagen chain were only slightly elevated in both types of scars studied. In situ hybridization was most informative when applied to hypertrophic scars. These lesions were characterized by the presence of intense hybridization signals for type I and III collagen mRNAs, and a moderate signal for type VI collagen mRNA, in nodules which were located in the upper dermis on each side of the original wound. This may explain, in part, why hypertrophic scars rise above the level of the surrounding skin. The results of the present study are in marked contrast to our previous findings on collagen gene expression in keloids and neurofibromas, in which the steady-state levels of type VI and I collagen mRNAs in particular were shown to be elevated. Thus, our results emphasize that distinct molecular mechanisms are operative in the development of clinically different dermal fibrotic conditions, such as normal and hypertrophic scars, keloids and neurofibromas.

Presence of type I and VI collagen mRNAs in endothelial cells in cutaneous neurofibromas.

In situ hybridization and peroxidase anti-peroxidase immunodetection were used in the same tissue sections to elucidate the spatial distribution of collagen gene expression in cutaneous neurofibromas, particularly in relation to blood vessels; the latter structures were identified by the presence of factor VIII-related antigen. The data indicate a clear relationship between the vascular structures and sites of locally elevated expression of type I and VI collagen genes. Specifically, some, but not all, blood vessels were surrounded by stromal cells highly active in expressing pro alpha 1(I) and alpha 2(VI) collagen genes. Furthermore, these genes were expressed by a subpopulation of endothelial cells within the walls of blood vessels traversing the lesion. To quantitate the overall expression of five genetically distinct collagen genes in cutaneous neurofibromas, we performed Northern analyses and slot blot hybridizations with pro alpha 1 (I), pro alpha 2 (I), pro alpha 1 (III), pro alpha 1 (IV) and alpha 2(VI) collagen cDNAs. Although the mRNA levels for all five genes were slightly increased in neurofibroma tissue, only the abundance of alpha 2(VI) collagen mRNAs was significantly elevated, as compared with normal skin. We conclude that endothelial cell populations with different levels of collagen gene expression exist within cutaneous neurofibromas: some are actively expressing type I and VI collagen genes, whereas in other the expression of these genes is effectively down-regulated. The markedly elevated steady-state levels of type VI collagen mRNAs suggest that synthesis of type VI collagen may contribute to the growth and architecture of cutaneous neurofibromas.

Expression of type I, III, and VI collagen mRNAs in experimentally injured porcine intervertebral disc.

Young domestic pigs were incised with a scalpel blade into the anterior part of annulus fibrosus of lumbar discs in order to study the reparative processes in the annulus fibrosus and the secondary reactions in the nucleus pulposus. Northern and slot-blot hybridizations were used to investigate type I, III, and VI collagen gene expression in the disc tissue. For this purpose a method for RNA isolation was modified so as to be applicable to the intervertebral disc, which has a low cell density and a high proteoglycan content in its extracellular matrix. The amount of total RNA was found to be very low, particularly in the nucleus pulposus. Intact RNA could be isolated from most parts of the injured discs, but only from the outer annulus of the control discs. Hybridizations showed that healing of the wound in the annulus fibrosus involves an increase in the synthesis of particularly type I and type III collagens. However, no changes in the collagen gene expression were detectable in the nucleus pulposus two weeks after the injury.

Segmental neurofibromatosis: immunocytochemical analysis of cutaneous lesions.

Cutaneous lesions from three patients with segmental neurofibromatosis were evaluated. Routine histologic studies revealed the presence of redimentary neural structures within an abundant collagenous matrix. The majority of the cells in all three cases expressed S-100 protein, suggesting their identity as Schwann cells. The stromal component stained positively for fibronectin and type IV collagen; the latter indicated the presence of basement membrane material. Embedded in the tumor mass were glandular epithelial structures that stained with epithelial membrane antigen antibody. Staining for factor VIII-related antigen revealed vascular endothelium and multiple scattered mast cells. In one case strands of cells stained with antibodies to desmin, suggesting muscle cell differentiation. This case may represent a distinct subset of neurofibromas.

Feasibility of genetic and immunological prediction of type I diabetes in a population-based birth cohort.

AIMS/HYPOTHESIS: Population-wide genetic screening of susceptibility to multifactorial diseases will become relevant as knowledge of the pathogenesis of these diseases increases and preventive interventions are identified. METHODS: Feasibility and acceptance of neonatal genetic screening for Type I (insulin-dependent) diabetes mellitus susceptibility and adherence of the at-risk children to frequent autoantibody follow-up were studied. Screening was offered to all families. The infants with HLA-DQB1 genotypes *02/*0302 and *0302/x (x not equal to *02, *0301, *0602) were invited to autoantibody follow-up. The children who developed signs of beta-cell autoimmunity were invited to a separate prevention trial. RESULTS: The parents of 31,526 babies born between November 1994 and April 1999 (94.4% of those eligible) agreed to genetic screening. We found that 4651 infants (14.8%) had increased genetic risk (2.5 to 15 times that of the general population) for Type I (insulin-dependent) diabetes mellitus, and 80% of them joined the autoantibody surveillance. At the age of 1, 2, 3 and 4 years, 74, 69, 68 and 76% of the at-risk children, respectively, attended the follow-up. A total of 17 of the 22 children (77%) who were born during the study period and have developed diabetes carry the risk genotypes we currently use for screening. CONCLUSIONS/INTERPRETATION: Population-based screening of genetic susceptibility for Type I diabetes, linked with a possibility to participate later in a prevention trial, is highly accepted in Finland and identifies about 75% of those developing diabetes at an early age. Families adhere well to the frequent measurement of signs of beta-cell autoimmunity in the children at-risk.

Hyperglycemic glucose concentrations up-regulate the expression of type VI collagen in vitro. Relevance to alterations of peripheral nerves in diabetes mellitus.

Electron microscopy of peripheral nerves obtained from two diabetic patients revealed large deposits of microfibrils and the presence of Luse bodies in the vicinity of perineurial cells. Microfibrils were found to accumulate also in the sciatic nerves of diabetic BB rats; these microfibrillar deposits were shown to contain type VI collagen by immunoelectron microscopy. Connective tissue cells cultured from rat sciatic nerves were exposed to high glucose concentrations. High glucose concentrations up-regulated the mRNA steady-state levels of alpha 1(VI), alpha 2(VI), and alpha 3(VI) chains of type VI collagen and caused accumulation of type VI collagen-containing fibrils in the cultures. Immunostaining and in situ hybridizations demonstrated that perineurial cells, Schwann cells, and fibroblasts expressed type VI collagen at the mRNA and protein levels. The results suggest that the turnover and supramolecular assembly of type VI collagen are perturbed in diabetic nerves and that glucose per se increases the expression of type VI collagen in vitro.

Effects of glucose on collagen mRNA levels and collagen secretion in EAhy 926 endothelial cell line.

Diabetes mellitus (DM) is a complex metabolic disease associated with increased accumulation of extracellular matrix by endothelial cells and contributing to vascular complications of long-standing diabetes. On the other hand, DM is also associated with decreased accumulation of extracellular matrix in granulation tissue, which is suggested to be a consequence of impaired angiogenesis. The role of hyperglycemia in these situations is not fully understood. We examined the effects of high glucose concentrations on the gene expression and secretion of various collagens in cultured EAhy 926 endothelial cells. EAhy 926 endothelial cells expressed alpha1(I) collagen mRNA at a low level and small amount of the corresponding peptide was secreted from the cells; mRNA was not affected but peptide secretion was increased by elevated glucose concentration. mRNAs for type III and VI collagens were not detected in the endothelial cells. Furthermore, high glucose concentration in long term had no morphological effects on cultured endothelial cells but increased the expression of type IV collagen, which could rather be beneficial for angiogenesis in a healing wound. Our results suggest that high glucose concentration per se may contribute to increased accumulation of extracellular matrix in blood vessels but probably is not responsible for decreased angiogenesis and granulation tissue formation in diabetic patients.

Effects of glucose on collagen mRNA levels and collagen secretion in EAhy 926 endothelial cell line.

Diabetes mellitus (DM) is a complex metabolic disease associated with increased accumulation of extracellular matrix by endothelial cells and contributing to vascular complications of long-standing diabetes. On the other hand, DM is also associated with decreased accumulation of extracellular matrix in granulation tissue, which is suggested to be a consequence of impaired angiogenesis. The role of hyperglycemia in these situations is not fully understood. We examined the effects of high glucose concentrations on the gene expression and secretion of various collagens in cultured EAhy 926 endothelial cells. EAhy 926 endothelial cells expressed alpha 1(I) collagen mRNA at a low level and small amount of the corresponding peptide was secreted from the cells; mRNA was not affected but peptide secretion was increased by elevated glucose concentration. mRNAs for type III and VI collagens were not detected in the endothelial cells Furthermore, high glucose concentration in long term had no morphological effects on cultured endothelial cells but increased the expression of type IV collagen, which could rather be beneficial for angiogenesis in a healing wound. Our results suggest that high glucose concentration per se may contribute to increased accumulation of extracellular matrix in blood vessels but probably is not responsible for decreased angiogenesis and granulation tissue formation in diabetic patients.