insulin4

Progress in Retinal and Eye Research 22 (2003) 545–562

Functions of insulin and insulin receptor signaling in retina: possible implications for diabetic retinopathy Chad E.N. Reitera,c, Thomas W. Gardnera,b,c,* a

Department of Cellular and Molecular Physiology, The Ulerich Ophthalmology Research Center and The JDRF Diabetic Retinopathy Center at The Penn State College of Medicine, M.S. Hershey Medical Center, 500 University Drive, Hershey, PA 17033, USA b Department of Ophthalmology, The Ulerich Ophthalmology Research Center and The JDRF Diabetic Retinopathy Center at The Penn State College of Medicine, M.S. Hershey Medical Center, 500 University Drive, Hershey, PA 17033, USA c The Penn State Retina Research Group, The Ulerich Ophthalmology Research Center and The JDRF Diabetic Retinopathy Center at The Penn State College of Medicine, M.S. Hershey Medical Center, 500 University Drive, Hershey, PA 17033, USA

Abstract Insulin action regulates the metabolic functions of the classically insulin-responsive tissues: liver, adipose, and skeletal muscle.

Evidence also suggests that insulin acts on neural tissue and can modulate neural metabolism, synapse activity, and feeding behaviors. Insulin receptors are expressed on both the vasculature and neurons of the retina, but their functions are not completely defined. Insulin action stimulates neuronal development, differentiation, growth, and survival, rather than stimulating nutrient metabolism, e.g., glucose uptake as in skeletal muscle. Insulin receptors from retinal neurons and blood vessels share many similar properties with insulin receptors from other peripheral tissues, and retinal neurons express numerous proteins that are attributed to the insulin signaling cascade as in other tissues. However, undefined neuron-specific signals downstream of the insulin receptor are likely to also exist. This review compares retinal insulin action to that of peripheral tissues, and demonstrates that the retina is an insulin-sensitive tissue. The review also addresses the hypothesis that dysfunctional insulin receptor signaling in the retina contributes to cell dysfunction and death in retinal diseases. r 2003 Elsevier Science Ltd. All rights reserved.

Keywords: Insulin; Insulin receptor; Retina; Diabetic retinopathy

Contents 1.

Introduction to the insulin signaling pathway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

546

2.

Properties of retinal insulin receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

548

3.

Insulin in the retina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

550

4.

3.1. Insulin and the blood–retina barrier . . . . . .

3.2. Retinal insulin production . . . . . . . . . . .

3.3. Insulin signaling in Caenorhabditis elegans and Retinal insulin signaling . . . . . . . . . . . . . . .

. . . .

550 551 552 554

5.

Insulin and insulin receptors in retinal diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

555

6.

Conclusions and future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

558

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

558

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

558

. . . . . . . . . . . . . . . . . . . . . . . . . . . .

Drosophila melanogaster . . . . . . . . . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

*Corresponding author: Department of Ophthalmology, Penn State University, College of Medicine, 500 University Drive, H097, Hershey, PA 17033, USA. Tel.: +1-717-531-8783; fax: +1-717-531-7667.

E-mail address: tgardner@psu.edu (T.W. Gardner).

1350-9462/03/$ - see front matter r 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S1350-9462(03)00035-1

546

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

1. Introduction to the insulin signaling pathway Insulin is a potent anabolic hormone that stimulates the uptake and storage of carbohydrates, fatty acids, and amino acids into glycogen, fat, and protein, respectively. The signal transduction pathway (Fig. 1) by which insulin action affects various cells to accomplish their biosynthetic roles has been extensively investigated and has increased our understanding of metabolism, physiology, and pathophysiology of disease. It is also the subject of numerous excellent reviews for further details (Nystrom and Quon, 1999; Saltiel and Pessin, 2002; Virkamaki et al., 1999). Insulin action begins with the synthesis of the mature peptide in the b cells of the pancreas. The b cells are constituents of a pancreatic ‘‘organelle’’ termed the islets of Langerhans which also consist of a; d; and PP cells. These cells, B2% of the pancreatic mass, secrete the hormones glucagon, somatostatin, and pancreatic polypeptide, respectively, and are organized into ovoid structures dispersed throughout the pancreas. Thus, the islets are organized to produce the hormones that have opposing actions on blood glucose levels.

The single insulin gene in humans encodes a 110 amino acid peptide, preproinsulin. The ‘‘pre-’’ portion is a 24 amino acid signal sequence that directs synthesis of

the peptide into the endoplasmic reticulum (ER), where it is cleaved. This remaining peptide is termed proinsulin. In the ER, proinsulin is folded and three disulfide bonds stabilize the insulin A and B chains of 21 and 30 amino acids, respectively. Proinsulin is then stored in secretory granules within the b cell until the b cell receives signals to release the contents into circulation.

While in the secretory granule, the C (connecting) chain (35 amino acids) that links the A and B chains together is removed to form the mature insulin molecule. Insulin, C peptide, and uncleaved proinsulin are all secreted into circulation via the portal vein (Davis and Granner, 1996; Ganong, 1997). In the liver, insulin works reciprocally with glucagon to control the balance of glucose and lipid metabolism.

Insulin action at the cellular level begins with the localization of the insulin receptor (IR) into plasma membranes. The IR is a heterotetramer comprised of two extracellular a subunits and two transcellular b subunits linked by disulfide bonds. The diverse effects of insulin commence when the receptor binds ligand on the a subunits, which induces a conformational change and activates the intrinsic tyrosine kinase activity in the b subunits in the cytoplasm. The IR autophosphorylates its b subunit on tyrosine residues within its tyrosine kinase domain and initiates a cascade of phosphorylation,

Fig. 1. Comparison of the conserved insulin-like signaling pathway among flies, worms, and mammals. Insulin-like peptides signal through structurally similar receptor tyrosine kinases, and the signal is perpetuated down a conserved pathway. IRS proteins (termed CHICO in Drosophila and unidentified in C. elegans) recruit PI3K and lead to Akt activation in response to insulin-like hormones. Not discussed in the text is PTEN, a lipid phosphatase which negatively regulates PI3K activity by reducing phosphatidylinositol 30 -OH phosphate (PIP3 ) content in cell membranes, and mutations of have been described in tumors. Conservation of this central growth-promoting pathway suggests numerous effectors have evolved around it to add specificity between different cell types within one organism. Dashed lines indicate hypothesized interactions. Adapted from Garofalo RS, Trends in Endocrinology and Metabolism, 13, 156–162.

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

dephosphorylation, and translocation events within the cell. Immediate protein substrates of the IR kinase include, among others, the insulin receptor substrates (IRS 1-4), Grb-2 associated binder (Gab-1), Shc, and Cbl, which in turn recruit other adapter and enzymatic proteins to propagate the signal and produce a biological effect.

Specifically, the IR-IRS- phosphatidylinositol 30 OH kinase ðPI3KÞ-Akt pathway has received much attention as it is well conserved (Fig. 1) and plays a role in many cellular processes including glucose uptake, glycogen synthesis, protein synthesis, and cellular survival.

The IRS proteins interact with the IR b subunit and are phosphorylated on numerous residues with insulin stimulation (White, 2002a; Yenush and White, 1997).

Four IRS proteins derived from four independent genes, and a similar protein, Gab-1, have been described. These molecules all contain similar domains and numerous sites for both tyrosine and serine phosphorylation. IRS and Gab proteins possess no catalytic activity before or after phosphorylation, but serve as docking proteins to concentrate other signaling molecules at the plasma membrane by the binding of its pleckstrin homology (PH) domain to the inner leaflet. At various phosphorylated tyrosine residues, proteins that contain Src homology 2 (SH2) domains bind and are recruited to the plasma membrane. The tissue distribution and presumed function of the IRS proteins are not fully redundant despite their similarities. This is illustrated in genetic ablation studies in which IRS-1 = mice show significantly reduced growth, while in IRS-2 = mice, growth is only diminished by 10%, but they develop overt diabetes leading to death (Araki et al., 1994; Tamemoto et al., 1994; Withers et al., 1998). It is currently believed that IRS-2 may play a more prominent role in brain function which may also be mediated by other growth factors that signal through IRS-2 (White, 2002a).

The regulatory subunit of class IA PI3K contains two SH2 domains, which flanks their p110 interaction domain, and binds to IRS proteins following insulin stimulation. Like IRS proteins, these proteins also act as docking molecules, and five proteins are in this class: p85a; p50a; AS53 ðp55a), p85b; and p55g: These proteins recruit the catalytic subunits of PI3K, p110a; p110b; or p110d (Fruman et al., 1998; Myers et al., 1992). The primary function of p110 proteins is to phosphorylate the 30 -OH position of inositol rings of the plasma membrane to form phosphatidylinositol 3-phosphate ðPIP3 Þ; but they also contain serine phosphorylation activity. Membrane regions rich in PIP3 aid in the recruitment of PH-domain containing proteins.

Further along in the signaling sequence, Akt-1, -2, and -3 (or PKB-a; -b; and -g; respectively), via their PH domains, are recruited from the cytosol to plasma

547

membranes enriched in 30 phosphorylated inositol rings where the kinase becomes maximally active by phosphorylation on conserved threonine and serine residues.

The Akt kinases were also termed RAC kinase for Related to A and C kinase because of structural % % similarities among the% three (Kandel and Hay, 1999).

A consensus amino acid sequence has been identified for efficient phosphorylation of substrates by Akt, but there is significant overlap among other kinases as well such as structural, localization, and tissue-specific considerations in determining whether a protein is a substrate of the Akt isoforms. Akts are involved in numerous insulin-induced cellular events listed above. For example, the glucose transporter 4 (GLUT4) is the insulinresponsive transporter in skeletal muscle and adipose tissue; with insulin, intracellular vesicles containing GLUT4 translocate and fuse with the plasma membrane. A dominant negative form of Akt, when overexpressed in adipocytes, blocks insulin-stimulated GLUT4 translocation, and overexpression of a dominant active form of Akt is sufficient for GLUT4 exocytosis in the absence of insulin (Cong et al., 1997; Kohn et al., 1996). The targets of Akt which promote GLUT4 translocation have yet to be identified. Akt also phosphorylates and inhibits glycogen synthase kinase 3b ðGSK-3bÞ; which allows glycogen storage to occur in liver and muscle as glycogen synthase is not phosphorylated by GSK-3b and, therefore, remains active (Cross et al., 1995). It has been postulated that Akt regulates protein synthesis. Various proteins of the protein synthesis initiation factor complex are phosphorylated

by Akt, as well as mTOR/FRAP and p70 S6K. Together, mTOR/FRAP and p70 S6K further regulate the initiation rates of protein synthesis by modulating the activity of the mRNA cap-binding complex and S6 ribosomal protein phosphorylation (Schmelzle and Hall, 2000).

Akt has also been shown to exert a strong antiapoptotic effect on cells. Indeed the viral form of Akt, vakt, has transforming effects on cells, suggesting an overabundance of Akt signaling may override apoptotic signals within cells, contributing to an oncogenic phenotype (Bellacosa et al., 1991; Testa and Bellacosa, 2001). Numerous targets of Akt signaling have been identified that contribute to the prosurvival effect of insulin. Caspases (cysteine–aspartic acid proteases) are responsible for cleavage of cellular proteins resulting in apoptosis of the cell. Caspases are cleaved and activated by numerous cellular signals and other caspases in a cascading manner (Earnshaw et al., 1999). The phosphorylation of caspase 9 by Akt inhibits this activation to promote cellular survival (Cardone et al., 1998). The Bcl family member protein, Bad, has also been implicated in regulating apoptosis in an Akt-dependent manner (Datta et al., 1997). When unphosphorylated, Bad antagonizes pro-survival Bcl proteins such as Bcl-2

548

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

and Bcl-XL in mitochondria, but when phosphorylated

by Akt, Bad is inactivated. This step is believed to maintain the integrity and functionality of the mitochondria for the production of ATP. Bcl-2 and Bcl-XL dimerize and function to maintain mitochondrial integrity, and following phosphorylation by Akt, Bad is sequestered by chaperone 14-3-3 proteins in the cytosol to promote cell survival. It is also hypothesized that the tumor supressor, p53, is also a target of Akt activity. Rat hippocampal neurons that were exposed to hypoxic conditions or nitric oxide toxicity to induce apoptosis were rescued by IGF-1 treatment and Akt activation, which was also dependent on p53 expression (Yamaguchi et al., 2001). Clearly other factors, such as hypoxia inducable factor (HIF) proteins, can also mediate a protective effect against hypoxia, but this observation is consistent with others that Akt translocates to the nucleus where it can regulate other proteins (Andjelkovic et al., 1997). Forkhead family transcription factors are also targeted by Akt in the nucleus (Brunet et al., 1999). Upon phosphorylation by Akt, Forkhead is transported out of the nucleus where it is sequestered in the cytoplasm by 14-3-3 proteins in an inactive state. Forkhead family proteins regulate expression of genes that induce apoptosis such as FasL; conversely, Akt may regulate pro-survival genes by influencing other transcription factors such as NF-kB and CREB (Brunet et al., 2001). This function of Akt may be a key contrast between the classically insulinresponsive tissues and retina, or neurons in general, in which Akt plays a greater role in promoting cellular survival (Dudek et al., 1997) and insulin-mediated synapse integrity (Puro and Agardh, 1984) than nutrient storage. Further specificity may involve specific Akt isoform expression in the retina, which has not yet been fully elucidated.

In general, the current understanding of insulin action is derived from studies involving the classically ‘‘insulinresponsive’’ tissues or cells and cell lines derived from them. The isoforms of the IR, IRS proteins, PI3K subunits, and Akt vary by tissue and cell type. These expression characteristics probably confers locally specific responses. It has been demonstrated that the IR is expressed abundantly in brain, retina, and other neurons, and its function probably differs from previously studied systems (Heidenreich et al., 1988; Reiter et al., 2003). Thus, further study is required to understand the similarities and differences of IR signaling in the central nervous system and retina compared to other insulin-sensitive tissues.

2. Properties of retinal insulin receptors Insulin binding characteristics have been described in the mammalian brain and other regions of the central

nervous system (CNS), including the retina (Havrankova et al., 1978a). However, our understanding of insulin action in the retina and CNS has lagged considerably compared to other insulin-responsive tissues in which insulin signaling is perturbed by diseases such as diabetes and polycystic ovarian syndrome (Gambineri et al., 2002; Virkamaki et al., 1999). Comprehensive studies by Waldbillig and Rodrigues et al. answered many questions about basic mammalian retinal IR characteristics, which are summarized within Table 1 (Rodrigues et al., 1988; Waldbillig et al., 1991, 1987a, b, 1988). They demonstrated specific insulin binding in purified bovine rod outer segments, which was specifically competed with excess insulin as opposed to 100fold greater amounts of proinsulin. Similar to brain and liver, 1 nanomolar insulin stimulated phosphate incorporation into the IRb subunit (Waldbillig et al., 1987a).

Furthermore, the retinal IR, when stimulated with insulin, possesses tyrosine kinase activity towards an exogenous substrate (Waldbillig et al., 1987b). These similarities among IRs in vitro suggest that retinal IRs may function in a cell-specific manner in vivo.

Our lab has recently demonstrated that the rate of retinal IR activity in vivo also resembles brain IR activity (Reiter et al., 2003). In contrast, the tyrosine phosphorylation and kinetic activity of the IR from liver and skeletal muscle fluctuate with circulating insulin levels during periods of fasting and feeding, which in turn affects the physiology of downstream mediators of insulin signaling. Within 90 min of re-feeding fasted rats, IRb and IRS-1 tyrosine phosphorylation and p85 association with IRS-1 are elevated concomitant with plasma insulin (Ito et al., 1997). Likewise, PI3K activity associated with phosphotyrosine-containing proteins or Shc is reduced with fasting in chicken liver and restored upon refeeding (Dupont et al., 1998). Our observations are consistent with those on brain IR kinetic activity (Simon et al., 1986) in which retinal IR tyrosine phosphorylation is unaltered between fasted and fed rats. Likewise, the kinetic activity of the IR in retina mirrors brain IR activity and is resistant to changes associated with fasting, which diminishes liver IR activity. Furthermore, in fasted rats, this tonic state of activity was greater than liver IR activity (Reiter et al., 2003). Taken together with the above studies by Waldbillig and Rodrigues, the IR in retina behaves similarly to liver IR in vitro, while retinal IR activity is more similar to brain IR activity in vivo where the IR is maintained in a tonic state of activity. This difference may result from how circulating insulin is delivered to tissues of the body, and suggests that the blood-retinal barrier (BRB) stabilizes insulin access to the retina.

One major structural difference between IRs expressed in retina/neurons and IRs expressed in liver, skeletal muscle, or adipose is the extent of glycosylation.

Both a and b subunits are glycosylated, but in general,

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

549

Table 1 A summary comparison of b cell, brain, and retina insulin production and IR properties among peripheral/vascular tissues, brain, and retina from mammalian systems

Insulin

b Cell

Brain

Retina

Synthesized and packaged into secretory granules for secretion as an endocrine hormone in response to circulating nutrients

mRNA and immunoreactive insulin (IRI) detected

mRNA and IRI detected Not in granules

Not in granules Secreted in vitro Mode of secretion unknown; may act as an autocrine/ paracrine effector

May act as autocrine/ paracrine effector

Insulin receptor

Liver/muscle

Brain

Retina

Molecular weight

IRaB135 kDa and IRbB97 kDa Greater sialic acid content Variable exon 11 expression Reduces kinase activity Rapid tyrosine phosphorylation in skeletal muscle and liver ðo90 sÞ

IRaB125 kDa and IRbB95 kDa Less extensive glycosylation Mostly exon 11 No change in kinase activity Increased tyrosine phosphorylation after 3 min in cerebellum, but no change in forebrain

IRaB125 kDa and IRbB95 kDa Less extensive glycosylation Exon 11 expression undetermined No change in kinase activity Increased tyrosine phosphorylation after 30 min

Effects of fasting Effects of systemic insulin injection

the extent of glycosylation of IRs expressed on neuronal cells is less than IRs expressed in liver and skeletal muscle, and can be variable even among different neurons (Hedo et al., 1981; Heidenreich et al., 1983; Ota et al., 1988). 125 I-insulin cross-linking studies by Waldbillig et al. (1987a) demonstrated further similarities between brain and retinal IRa in which the ligand-receptor complex migrates B10 kDa further on SDS-PAGE than liver IRa: Terminally-linked sialic acid residues comprise most of the IR glycosylation motifs in adipocytes, as shown by neuramidase sensitivity, and the glycosylation is absent in brain IRs; however, the core peptide remains identical following removal of all N-linked glycans from 14 potential a subunit sites (Heidenreich and Brandenburg, 1986).

Four potential N-linked sites are on the extracellular portion of the IRb; which also contains O-linked oligosaccharides (Collier and Gorden, 1991). Like the IRa; the IRb subunit in retina and brain also migrates slightly faster on SDS-PAGE compared to liver IR (Reiter et al., 2003). Enzymatically catalyzed IR glycosylation is important for normal IR function as demonstrated by site-directed mutagenesis of the four N-glycosylation sites on the IRb (Leconte et al., 1992).

The mutant IRb subunit is able to undergo autophosphorylation in response to insulin in vitro, but the IR lost all kinase activity towards an exogenous substrate.

This alteration prevented any effect of insulin in cells that overexpressed the mutant IR on either glucose uptake or thymidine incorporation into DNA. Although glycosylation is required for IR functionality, an explanation for variability of the glycosylated IR among

tissues currently remains elusive. However, one might speculate that the decreased glycosylation of neural IRs influences tonic IR activity. Further analysis of IR glycosylation by mass spectroscopy may help to clarify structure–function relationships.

The a and b subunits of the IR are transcribed and translated as a single peptide from one gene on human chromosome 19 before assembly into the mature a2 b2 form. A second difference involves exon 11 expression on the a subunit, which is the only sequence variance in the mature peptide (Ebina et al., 1985; Ullrich et al., 1985). This splice variant is differentially expressed in various tissues. Brain exclusively expresses the exon 11 (type A) form of the IR, but the majority of liver IRs are the þexon 11 (type B) form (Seino and Bell, 1989). The expression of the splice variants in retina has not been evaluated, nor has the tissue-specific mechanism of its expression, making the overall function of IR structural heterogeneity still unclear.

Consistent with the ‘‘housekeeping’’ nature of the IR promoter, with a GC-rich promoter and the lack of a TATA box, (Araki et al., 1987), the IR has been shown to be expressed constitutively and broadly throughout the retina on neuronal, endothelial, and retinal pigmented epithelial (RPE) cells. Using B10 anti-sera, a human anti-IRa sera from a patient with severe insulin resistance, Rodrigues et al. (1988) localized the IR to the outer segments and retinal nuclear layers of bovine, monkey, and human retina. Interestingly, only rod outer segment disks in monkey, and human Muller . glial cells were also immunoreactive for the IR with this antibody.

IRa structural heterogeneity may have contributed to

550

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

the variable results with this IRa antibody. Therefore, Naeser (1997) and Gosbell et al. (2002) extended those observations by detecting IR immunoreactivity in all retinal layers of human and bovine retina, respectively, using polyclonal antibodies against intracellular portions of the IRb that are not differentially glycosylated.

Retinal endothelial cells and pericytes both display specific binding for insulin. Treatment of these cells in vitro rapidly stimulated thymidine incorporation into DNA suggesting a strong mitogenic effect (King et al., 1985). Interestingly, it was also observed that the retinal vascular cells were more sensitive to the growth promoting effect of insulin than aortic endothelial and smooth muscle cells. The IR in RPE cells has also been investigated, and it was found that its form is most similar to the peripheral form and not the neural form.

Compared to liver, it also exhibits neuramidase sensitivity, and overall specific binding was on the same order as liver (Waldbillig et al., 1991). Furthermore, the IR in RPE cells localized exclusively to the basolateral surface of the cell, suggesting a possible role in unidirectional insulin transport from the choroidal circulation to the photoreceptors (Sugasawa et al., 1994).

Lastly, it has been reported that hybrid receptors exist, and their expression appears to be widely distributed and to include neurons (Bailyes et al., 1997; Garcia-de Lacoba et al., 1999). That is, an ab segment of the IR is crosslinked to an ab segment of the homologous IGF-1R, and this heterodimer has the potential to bind either insulin or IGF-1. The function of these receptors is not completely understood, but their existence may be a mechanism to downregulate a signal from one receptor, or to add greater specificity of signaling by either insulin or IGF-1. Hybrid IR/IGF-1R expression may also change in disease states as in those with hyperinsulinemia (Federici et al., 1998). Our lab has detected this complex in immunoprecipitates of the IR followed by immunoblot analysis for the IGF-1R, and vice versa, in retina (unpublished observations).

Further investigation is required to determine the functions of IR/IGF-1R hybrids in retina and other tissues alike.

In summary, because the retina, and other regions of the CNS, expresses the IR at levels similar to other classically insulin-responsive tissues, it must be considered as a major target of insulin action.

3. Insulin in the retina 3.1. Insulin and the blood–retina barrier The blood–retina barrier (BRB) exquisitely regulates the exchange of material between the plasma and the neural retina. This partition gives the retina immune privilege as in other regions of the body such as the

brain, testes, and anterior chamber of the eye (Ferguson and Griffith, 1997). The BRB selectivity functions in stark contrast to the endothelium of other tissues which contain numerous fenestrations, and larger gaps allow for the free diffusion of molecules between the interstitial space and endothelial lumen. This is the case for liver, skeletal muscle, or adipose tissue which must rapidly metabolize nutrients as they are absorbed from the gut, and insulin is released into the body. Therefore, the relatively rapid passage of insulin from the blood to some tissues is necessary, and it is reasonable to hypothesize that insulin must be transported across the BRB for insulin to act on the neural retina. Some comparisons can be drawn from studies on insulin transport across the blood–brain barrier (BBB). In the brain, accumulation of exogenous insulin begins after 60 min of insulin infusion and is thought to occur by a saturable, receptor-mediated transport process (Schwartz et al., 1992, 1990). After 60 min; insulin accumulation in brain proceeds in a linear fashion despite constant levels of insulin in circulation. This results in relatively constant levels of insulin in the cerebrospinal fluid. The mechanism of transport across the BBB is also thought to be more efficient at physiological levels of insulin, as rates of insulin accumulation in brain decreases with infusion of pharmacological levels of insulin (Schwartz et al., 1992).

As stated above, the IR is expressed throughout the retinal endothelial network. The retinal vasculature expresses IRs which resemble those expressed in liver, rather than the smaller neural form of the IR, as assessed by insulin cross-linking and SDS-PAGE autoradiography (Haskell et al., 1984; Im et al., 1986).

Investigations lead by King et al. (King and Johnson, 1985) spearheaded our understanding of endothelial cell transport of insulin. Using an aortic endothelial cell monolayer separating two compartments as their model, they demonstrated that insulin is carried unidirectionally across the monolayer. This transport of insulin was mediated specifically by the IR, as pre-treatment with anti-IR antibodies blocked the transport, and B80% of the transported insulin remained intact. This finding strongly suggests that endothelial cells do not degrade significant amounts of insulin, as compared to other growth factors, which may be reflected in a lower expression of insulin degrading enzymes in retina and vascular cells (Varandani et al., 1982). These data suggest that a major function of IRs on endothelial cells is to regulate transport of the hormone to a specific tissue. The specific pathway that endothelial cells utilize to transport insulin has not been fully described.

Incubating endothelial cells with leupeptin, a lysosomal protease inhibitor, had no effect on insulin transport suggesting those cellular compartments are not involved. Monensin, a proton ionophore which raises endocytic vesicle pH, reduced insulin transport. This

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

result suggests that a cellular compartment requires an acidic environment to dissociate insulin from the IR for complete insulin transport (Hachiya et al., 1988).

Using electron microscopy to detect gold-labeled insulin, transcellular flux of insulin was also described in cultures of retinal vascular cells (Stitt et al., 1994).

Within 2 min of insulin binding, IRs with bound insulin were clustered into clatherin-coated pits—the beginning of the well-described receptor-mediated endocytosis pathway (Brodsky et al., 2001). After 10 min; the insulin/IR/clathrin-coated pit complex had invaginated and localized to the interior of vacuole-like compartments, which the authors identified as endosomes.

Following 20 min of insulin stimulation, the labeled insulin was associated with what the investigators termed a multivesicular body, and the labeled insulin again was localized to the interior of this electron-dense, highly membranous structure. After 60 min; the majority of labeled insulin resided in secondary lysosomal structures, while only B23% of the labeled insulin had been transported to the basal plasma membrane or released to the extracellular space. These experiments in vitro demonstrate an IR-mediated transport of insulin across retinal endothelial cells, and that circulating insulin has the potential to reach the neural retina.

Similar transport results have been reported in postmortem bovine eyes that were reperfused with bovine blood and 125 I-insulin (James and Cotlier, 1983). In their time course experiments lasting up to 60 min; insulin was localized to the basal surface of endothelial cells and also associated with pericytes. However, no labeled insulin was observed associated with the neural retina, which may suggest either more time is required for insulin transport in bovine eyes, or the vascular, neural, and glial functions were compromised in eyes of exsanguinated animals contributing to reductions in insulin transport rates. By contrast, Shires et al. (1992) showed that 125 I-insulin injected into the superior vena cava of anesthetized rats accumulated in the vitreous of normal rats within 2 h: This accumulation is also thought to be due to a saturable transport process.

The tight junctions that form in retinal endothelial cells promote a greater resistance across the barrier and may restrict retinal insulin permeability compared to other tissues. When cultures of aortic endothelial cells were directly compared to retinal endothelial cell cultures, the electrical resistance across the monolayer was B4-fold greater in retinal endothelial cells (Gillies et al., 1995). The permeability of insulin in the retinal endothelial cell cultures was significantly less than the aortic endothelial cell cultures, indicating that expression of specific barrier proteins contribute to the permeability characteristics of retinal cells.

The results presented above and observations from our lab illustrate the substantial differences between the tighter BRB and the endothelium of other tissues.

551

Administering superphysiological amounts of insulin intraportally to fasted rats induces significant phosphorylation of the IRb; IRS, and IRS-1 associated PI3K activity in muscle within 90 s (Folli et al., 1992). By contrast, we found no effect on retinal IR phosphorylation until 30 min post injection using a similar model (Reiter et al., 2003). These results are consistent with other studies for brain and retina insulin uptake (Niswender et al., 2003; Schwartz et al., 1990; Shires et al., 1992). Collectively, these studies confirm the difference between the endothelium of retina and different insulin-sensitive tissues, and that the transport of insulin across the BRB is significantly slower than other vascular beds. This suggests the neural retina is not likely a major target for immediate nutrient metabolism as is skeletal muscle, but the steady-state transport of insulin to the neural retina may provide a stable trophic signal as opposed to insulin’s better known rapid metabolic functions.

Insulin signaling characteristics in endothelial cells has been investigated and described as in muscle, liver, adipose, and IR-overexpressing cells, but some of the final biological outputs of endothelial cells in response to insulin are unique. As demonstrated by King and others (Jialal et al., 1985; King et al., 1985; King and Johnson, 1985), retinal vascular endothelium contains functional IRs that promote cell division, and they are required for unidirectional insulin transport. In general, the IR signaling characteristics in endothelial cells resembles what has been described in other systems, and the Akt pathway is activated in response to insulin.

In human endothelial cell cultures induced to die with TNF-a treatment, insulin restores Akt phosphorylation and reduces rates of apoptosis (Hermann et al., 2000).

This mechanism also appeared to be dependent on Akt’s ability to inhibit caspase 9 activity, while Bad may not be a target of Akt in endothelial cells. Insulin’s ability to stimulate production of other autocrine growth factors, such as VEGF, may also play a role in endothelial cell survival. Insulin and VEGF may have combined effects in promoting the formation of endothelial tubes in vitro (Yamagishi et al., 1999). Glucose uptake in retinal endothelial cells is also not affected acutely by insulin, consistent with a lack of expression of the GLUT4 transporter in endothelial cells (Betz et al., 1983; Harik et al., 1990; Kumagai, 1999). Together, insulin promotes both mitogenic and survival effects on endothelial cells.

In terms of diabetic retinopathy, these growth promoting effects may produce temporary detrimental consequences on the retinal vasculature as discussed below.

3.2. Retinal insulin production An interesting observation by Havrankova et al. added considerable controversy concerning the role of insulin in neuronal cells (Havrankova et al., 1978b).

552

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

They detected insulin immunoreactivity in brain tissue, and they put forth the hypothesis that insulin may be synthesized de novo by other mammalian tissues than pancreatic b cells. It was discovered that insulin concentrations were at least 10-fold greater in brain than in plasma, but others have disputed those results (Baskin et al., 1983). Therefore, brain may either have mechanisms in which to concentrate insulin or synthesize the peptide. Devaskar et al. (1994) furthered those observations by investigating whole brain and cultured primary neurons and glial cells from rabbits. The insulin transcript was detected using sensitive techniques, such as nuclear run-on assays and reverse-transcription PCR, but not by Northern blot analysis suggesting insulin mRNA in these tissues is a rare transcript. When neonatal rabbit neurons and glial cells were cultured, it was demonstrated that insulin continued to accumulate in the culture media over 6 days in culture, supporting the hypothesis that neurons synthesize insulin de novo.

Further in vivo evidence of brain-derived insulin came from mRNA in situ hybridization which showed significant binding to neuronal cells over various regions of the brain. However, insulin antisera was unable to detect specific insulin immunoreactivity as it had in previous studies, which may be due to the inability of insulin antibodies to react across different species. This evidence for de novo synthesis was strengthened with electron microscopic analysis of the rat CNS demonstrating the presence of both insulin I and II mRNAs and insulin protein in the ER and Golgi apparatus (Schechter et al., 1996). Although the amount of insulin in brain is probably considerably lower than that expressed in b cells, it is intriguing to speculate that neuronal insulin may provide a local trophic or developmental signal, act as a neurotransmitter, or regulate nutrient metabolism in neurons. In fact, a lower concentration would be predicted on the basis that b cells concentrate insulin in granules, but the brain does not (see Table 1).

The evidence for neuronal insulin production has been extended to retina tissue. Using insulin antisera, insulin immunoreactivity was detected in human retinal tissue and in the optic nerve of mice (Das et al., 1984).

Insulin was reported to be located in the ganglion cell layer, the inner nuclear layer, and the inner and outer plexiform layers, but it was completely absent from the photoreceptors. Glial cells from the optic nerve also stained positive for insulin. A human retinoblastoma cell line, Y79, also was positive for insulin immunoreactivity, mRNA, and binding sites, although this may be an artifact of the transformed nature of the cells, as insulin and IR expression may be altered (Das et al., 1991; Pansky et al., 1986). Using cultured primary rat Muller . glial cells, the Das group showed both insulin protein and mRNA within this specific subset of retinal cells (Das et al., 1987). Given the supportive nature of

glial cells in general, this suggests the Muller . cells may secrete the insulin for other neurons as a paracrine hormone. Furthermore, intact retina tissue was shown to express the mRNA for preproinsulin in rats (Budd et al., 1993). In embryonic chickens, the amount of insulin expression in retina was considerably less than pancreatic insulin expression, as a more sensitive analysis using RT-PCR was required for insulin mRNA detection (de la Rosa et al., 1994).

Given the extensive evidence that retina and numerous other extra-pancreatic tissues express insulin-like molecules (LeRoith et al., 1988), we must now recognize the potential of insulin to modulate neuronal functions.

The observations in chick retina, as observed in other species, suggests that retinal (or neuronal) expression of insulin is not purely a mammalian phenomenon, and that insulin, or other insulin-related peptides in other species, may have an evolutionarily conserved role in general neuronal development. Retinal development is sculpted by coordinated differentiation and apoptosis, and Diaz et al. have established that retinal cell survival in chicks depends on insulin produced by the retina (Diaz et al., 2000). Using antibodies to inhibit insulin action in developing chick retina, they observed a significant elevation in apoptosis; ganglion cells were particularly sensitive to the induction of apoptosis by blocking insulin signaling. Lastly, consistent with our observations in mammalian retina, chick retina responds to insulin in vitro with significant elevations in Akt phosphorylation, while the Ras/Raf/MEK/ERK pathway is relatively insensitive to insulin stimulation.

This supports the hypothesis that retinal insulin serves as a trophic factor in either an autocrine or paracrine manner. However, developing chick retina stimulated with insulin exhibits increased 3 H-thymidine-DNA incorporation and protein synthesis, suggesting insulin may play a role in mitogenesis and differentiation during embryonic stages (Hernandez-Sanchez et al., 1995). These results suggest insulin signaling maybe important in the adult retina as well, and it also plays a major role in development. In fact, insulin binding in the retina is greater in fetal versus adult chick retina (Waldbillig et al., 1991).

3.3. Insulin signaling in Caenorhabditis elegans and Drosophila melanogaster Recent studies of the insulin-like signaling in worms and flies have provided important insights into insulin physiology of mice and men. A particularly interesting line of study in the nematode worm, Caenorhabditis elegans, has revealed a neuron-specific insulin-like signaling system that controls the worm’s life span (Kimura et al., 1997). This signaling system closely mirrors what has been established in mammalian cells and described above. The IR homolog in C. elegans,

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

DAF-2, signals to AGE-1 (homolog to mammalian PI3K), AKT-1 and -2, and DAF-16 (homolog to mammalian Forkhead transcription factors, Fig. 1) (Paradis and Ruvkun, 1998). In wild-type C. elegans, when nutrients are abundant, the worms’ life span is B12 days, but in the absence of nutrients, their growth becomes arrested and they enter a dauer stage in response to a dauer-inducing pheromone. This is a form of ‘‘hibernation’’ that allows the worm to prolong its life span until its environment becomes more favorable for reproduction. Experiments by Ruvkun and co-workers (Paradis and Ruvkun, 1998) have established the IR-like signaling of DAF-2 in C. elegans to be intimately linked to regulation of life span.

Genetic sequencing of the DAF-2 gene revealed that it is an IR homolog in C. elegans (Kimura et al., 1997). It is 36% identical to the human IR within its ligand binding domain (35% identical to the IGF-1R), and 50% identical with 70% amino acid similarity within the tyrosine kinase domain. The genetic organization is also quite similar as it has a proform, a proteolytic site, and a transmembrane domain. C. elegans carrying a mutation in DAF-2 displayed arrested growth and a rise in fat accumulation that is associated with dauer formation, despite a favorable environment for survival and reproduction. This evidence suggests a strong evolutionary link in IR function in which the insulin signaling pathway regulates metabolism.

Since that seminal report established a significant role of insulin-like signaling in C. elegans, an entire cascade homolog has been described (Paradis and Ruvkun, 1998). Comprehensive genetic studies of C. elegans mutants have established a pathway from DAF-2 to DAF-16, the Forkhead homolog, that regulates life span in the worm. For example, essentially all viable worms with DAF-2 mutations alone enter the dauer stage and have a prolonged life span, as discussed above. A complementary mutation in DAF-16 will reestablish a normal phenotype (Ogg et al., 1997). Therefore, activity of the Forkhead transcription factor, DAF-16, is required for dauer formation, and signaling from DAF-2 directly antagonizes such activity, as the IR affects forkhead activity in mammalian cells. Using that strategy, other mediators of the DAF-2 signaling pathway were linked. C. elegans with AGE-1 mutations, like DAF-2 mutations, enter the dauer stage, and normal life span is restored when an activating mutant of AKT-1 is also expressed. Furthermore, the gene products of AKT-1 and AKT-2 in C. elegans directly antagonize DAF-16 as is the case in mammalian cells (Paradis and Ruvkun, 1998).

Now the question becomes, again as in mammalian systems, what is the role of the IR-like DAF-2 signaling system in neurons? Using tissue specific promoters, Wolkow et al. (2000) elegantly demonstrated in C. elegans that restoring the insulin-like signaling system to

553

DAF-2 and AGE-1 mutants in neuronal cells regulates longevity. This was in contrast to restoring DAF-2 signaling to muscle and gut cells of C. elegans, which did not restore normal life span. However, a functional divergence for insulin-like signaling emerged from their study. It was discovered that restoring DAF-2 to the muscle and intestinal cells reduced fat accumulation and restored normal metabolic processes, but did not alter life span. This signaling dichotomy in lower organisms suggests it may also be conserved in mammals.

Furthermore, 37 insulin peptide-like genes have been identified in C. elegans, and deletion of one of them does not effect life span, so there is probably functional redundancy (Pierce et al., 2001). Interestingly, the INS genes were expressed by neurons suggesting a possible neuroendocrine role for insulin conserved throughout evolution.

The fruit fly, Drosophila melanogaster, also possesses an insulin-like signaling cascade with many similarities to C. elegans and mammalian systems with mutations that effect fly growth. Comparisons among the conserved insulin signaling pathways can be seen in Fig. 1 (Garofalo, 2002). In D. melanogaster, evidence suggests that the primary role for the insulin-like signaling is to control growth by regulating cell size and number, and mutations within the Dinr gene result in abnormal development (Brogiolo et al., 2001; Fernandez et al., 1995). Like C. elegans, D. melanogaster express multiple insulin-like peptides, three of which localize to neurosecretory cells in the brain. When one of these peptides is overexpressed, it results in increases in both cell size and number. When the Dinr gene is also overexpressed, the eye (comprised mostly of photoreceptor cells) displays an overgrowth, while Dinr mutations result in reduced body weight and fewer ommatidia, the fly photoreceptor units. Genetic analysis of these mutants suggest that the insulin-like peptides are ligands for the DINR as the eye phenotypes are corrected in insulin-like peptide overexpressing mutants with Dinr mutations (Brogiolo et al., 2001). The Dinr peptide is unique in that it expresses a B60 kDa C-terminal extension that is proteolytically cleaved in some cells, and this extension is hypothesized to give the receptor different functions in different cells.

A similar function of the IR may occur in mammalian cells; that is, some functions of the IR may depend on the cell type expressing the IR rather than a proteolytic event which modifies the receptor. Given the conserved nature of the IR throughout evolution, the function of the IR remains very similar in promoting anabolic effects, yet other modifications or effector molecules remain to be discovered. Glycosylation and exon 11 expression are two other modifications of the IR that also may be contributing to differential IR activity and regulation in mammalian cells. Taken together, the conservation of insulin-like peptide expression in neurons or neurosecretory cells from lower species to

554

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

mammals suggest a specialized function of insulin and the IR, and that neuronal cells have evolved to utilize insulin signaling differently. In terms of the retina, this function may be in place to regulate proper embryonic or adult development and cell growth, maintain cellular survival, metabolism, or regulate phototransduction, all of which require further investigation. Indeed, an insulin-secreting, gut-associated b cell does not arise until after neurons in evolution, so the b cells may be modified neural cells. In fact, they express proteins previously described as only neuron-specific, such as neurofilament (Escurat et al., 1991). Such properties of cells blur the line between a pure neuron and a hormone-secreting endocrine cell in complex mammalian systems. Therefore, some of the properties attributed solely to pancreatic b cells may be in common with retinal and other neurons, and vice versa.

4. Retinal insulin signaling The physiological outcomes of IR activation in retina and other neurons are beginning to be elucidated. It has been established that IRs are expressed on different neurons and glia, and co-localization studies have determined that the IR, IRS-1, PI3K, and phosphotyrosine motifs topographically overlap in neurons (Baskin et al., 1993; Folli et al., 1994). Hence, an insulin signaling pathway could be functional in those cells since it resembles IR signaling in other systems, but the physiological role has received little attention.

Our observations extended previous observations and have begun to establish a possible function of insulin signaling in retina (Reiter et al., 2003). As stated above, we observed slow in vivo phosphorylation of the IR and Akt in retina with systemic insulin administration. In vitro, we saw no significant change in IRS-1 tyrosine phosphorylation, unlike IRS-2, which was rapidly phosphorylated and dephosphorylated. Retina expresses mRNA for all three Akt isoforms, but insulin stimulates only Akt-1 activity. This suggests tissue specific signaling in the retina tissue. Consistent with the findings of Diaz et al. (2000), the ERK 1/2 pathway stimulation was relatively unaffected compared to the Akt pathway with insulin. It cannot be ruled out that other proteins mediate the insulin signal in retina, or that subsets of the IR expressed on different parts of the retina may signal differently. In brain, the IR is a component of the synapse, and signals arising there may be mediated by p58/53, a substrate of the IR which is expressed in brain and absent in liver and skeletal muscle (Abbott et al., 1999; Yeh et al., 1996). The IR is expressed broadly on single neurons, while IR and p58/53 co-localize in the synapse, suggesting a specialized function for insulin signaling within the same neuron. Indeed, novel mediators of insulin signaling may exist in retina and

other neurons that have not been described in other cell types. Our hypothesis in retina is that the IR-Akt pathway in retina is tonically active and does not fluctuate in response to the organism’s nutritional state, and that this signal is required for proper development and cell survival.

Similar experiments analyzing systemic insulin in rats have produced interesting results suggesting different parts of brain may be more insulin sensitive. Within 3 min of systemic injection, cerebellar IRs were phosphorylated, but there was no effect on the frontal cortex (de L.A. Fernandes et al., 1999). Therefore, regions of the brain with low permeability to insulin behave similarly to retina in which transport is slower, resulting in delayed activation of the IR. A recent report by Niswender et al. (2003) characterized signaling in hippocampus in response to intraperitoneally administered insulin. In this model, PI3K activity was not significantly elevated until 15 min post injection. These results suggest that, like retina, the brain can respond to circulating insulin, and it does so by signaling to known mediators of the signaling pathway. However, the complete physiologic outcome of neuronal insulin signaling in vivo remains to be seen.

We have also identified some of the molecular mechanisms of how insulin may reduce apoptosis by studying retinal neurons in culture (Barber et al., 2001).

Serum starvation-induced apoptosis of retinal neurons was partially inhibited with insulin treatment, and inhibitors of the PI3K pathway inhibited insulin’s effects. Furthermore, insulin reduced cleavage and activation of caspase 3, suggesting it may also be a target of Akt. Interestingly, the mitotic rates of these retinal neurons were unaffected with insulin treatment, in direct contrast to retinal endothelial cells which proliferate in response to insulin, as reported by King et al. (1985). Together with our results in vivo, these data strongly point to the Akt-mediated survival pathway as a mediator of insulin-induced retinal cell survival. This, however, does not rule out effects from other growth factors that are present in the retinal milieu. For instance, to promote retinal ganglion cell survival in vitro, combinations of growth factors are required in addition to membrane depolarization, cell– cell contact, and the second-messenger cAMP (Goldberg et al., 2002; Meyer-Franke et al., 1995).

Heidenreich et al. previously showed reduced rates of glucose uptake in response to insulin in neurons in vitro (Heidenreich et al., 1988), and insulin can stimulate amino acid uptake in brain neurons (Yang and Fellows, 1980). The question of insulin-stimulated substrate uptake has also been addressed in human Y79 retinoblastoma cells (Yorek et al., 1987). Of the amino acids assayed, only glycine uptake was stimulated by both insulin and IGF-1. In this way, insulin could potentially regulate neurotransmission as glycine is a

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

putative neurotransmitter. However, uptake of other neurotransmitters such as GABA and dopamine were unaltered in this cell line.

Other aspects of retinal function have been investigated in response to insulin. Insulin has been reported to mediate differentiation and synaptogenesis (Puro and Agardh, 1984), and to regulate acetyltransferase proteins (Holdengreber et al., 1998; Kyriakis et al., 1987).

Embryonic chick retinal cell in vitro increased their cholinergic acetyltransferase activity over time in an insulin dose-dependent manner. Holdengreber et al. also showed an insulin-mediated increase in choline acetyltransferase protein with short-term exposure to insulin, and that this increase was found in the ganglion and inner nuclear cell layers. Insulin, known to modulate the activity of various transcription factors, may regulate choline acetyltransferase protein by regulating jun (Ren et al., 1997), but other transcription factors may also be involved. Insulin may also synergize with other trophic factors in retina to promote differentiation. In the postnatal chicken retina, insulin and FGF2 were required to induce waves of proliferation and expression of progenitor markers, such as Pax6, in Muller . glia cells (Fischer et al., 2002). Together, insulin appears to promote some aspects of neurogenesis in retina (Hernandez-Sanchez et al., 1995).

Acute effects of insulin on retinal electrophysiology have also been studied. For instance, the a- and b-waves of the electroretinogram (ERG) were reduced when insulin was applied to an in vitro system (Gosbell et al., 1996). Such an effect on the photoreceptors is consistent with our and other’s observations of IR expression on photoreceptors (Gosbell et al., 2002; Naeser, 1997; Rodrigues et al., 1988). A mechanism by which insulin may alter the ERG was provided by Stella et al. (2001) when they showed an insulin-dependent reduction of Ca2þ influx in photoreceptors. However, these results are in contrast to those of Lansel and Niemeyer (1997) in cat retina where, under normoglycemic conditions, the b-wave amplitude of the ERG was unaffected. Only when the eye was perfused with hypoglycemic solutions, which reduced the b-wave, did insulin recover retinal nerve ERG activity. The physiological relevance of these results is questionable because under hypoglycemic conditions, circulating insulin is also usually reduced, but their results also suggest insulin crosses the BRB. So, if the BRB were more permeable to insulin, one may expect vision to change briefly with fluctuations in circulating insulin. Therefore, IRb phosphorylation in retina may be maintained stably for proper phototransduction and synapse function.

Since the retina is the only neural tissue exposed to light, some questions concerning light-mediated IR activity are beginning to be answered. Rajala et al., found that the IRb interacts with the regulatory subunit of PI3K, p85, in bovine rod outer segments. In pull

555

down assays, the interaction between the IR and PI3K was dependent on tyrosine phosphorylation of the IRb; and treatment of the tissue with insulin in vitro increased PI3K association and activity associated with the IRb (Rajala and Anderson, 2001). This suggests some aspects of insulin signaling in retina may occur independent of the IRS proteins in activating PI3K, or this IRb-PI3K interaction may be specific to the rod outer segments. In a second report, the same group (Rajala et al., 2002) showed that light increases IRb phosphorylation and PI3K activity in rod outer segments associated with the IR in mice and rats. They propose that this phenomenon may be a mechanism to protect the retina from light induced damage. This suggests Akt would also be activated by light since it is immediately downstream of PI3K. Our observations in retina treated in vitro with insulin suggest this would be the case as all nuclear layers of the retina, including photoreceptors, had increases in immunoreactivity for the active, phosphorylated form of Akt, and Akt-1 and 3 mRNA localized to this region of the retina (Reiter et al., 2003). Even more interesting is their suggestion that IR activation could be modulated by other ligands and that the intracellular portion of the IRb may be modulated by novel ligand-independent mechanisms to increase its phosphorylation, or that the IR may play some role in phototransduction.

5. Insulin and insulin receptors in retinal diseases Theories about the pathogenesis of diabetic retinopathy have been plentiful and include increased polyol pathway flux, nonenzymatic glycosylation and receptor for advanced glycation endproducts (RAGE) activation, increased oxidative stress, protein kinase C activation, altered growth factor expression, and hypoxia (Aiello and Gardner, 2000; Brownlee et al., 1988; King et al., 1996; Kowluru et al., 2001). As of this writing no comprehensive mechanism has yet been established that provides a causal link between the fundamental metabolic abnormalities of diabetes and the initiation and progression of retinopathy that includes an explanation to account for loss of vision. During this period there have been two implicit assumptions underlying most diabetes complications research: first, that retinopathy is solely or primarily a blood vessel disorder; and second, that excess glucose is necessary and sufficient to account for the phenotype (Gardner et al., 2002). Until recently these appeared to be reasonable operative assumptions. The clinical signs of retinopathy are determined by the appearance of hemorrhages, microaneurysms, cotton–wool spots, macular edema, lipid exudates, and neovascularization. These features are visible with ophthalmoscopy because they contain pigment (hemoglobin) or reduce retinal transparency (cotton–wool

556

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

spots and lipids). However, at least 95% of the retinal volume and mass is comprised of neurons and glial cells that lack pigment and are thus indistinguishable by biomicroscopy, fluorescein angiography, or ultrasound, and only marginally detectable by current optical coherence tomography. Recent work by several investigators now provides evidence that challenges this first assumption. There is now clear evidence for loss of retinal neurons by apoptosis (Barber et al., 1998; Zeng et al., 2000), activation of glial cells (Barber et al., 2000; Rungger-Brandle et al., 2000), and microglial cells (Rungger-Brandle et al., 2000; Zeng et al., 2000) by altered expression of glial fibrillary acidic protein (GFAP). These changes develop early in the course of experimental diabetes and precede detectable structural microvascular lesions as assessed by fluorescein angiography or trypsin digest preparations. In fact, the loss of retinal neurons was first described 40 years ago (Bloodworth, 1962; Wolter, 1961), and close histologic examination of eyes with macular edema reveals extensive loss of ganglion cells and inner nuclear layer neurons (Yanoff and Fine, 1989). These anatomic changes support the well-recognized functional changes in the ERG, color vision, and contrast sensitivity that precede vascular changes (Ghirlanda et al., 1997).

Collectively these data strongly support a major if not direct effect of diabetes on retinal neural function that may or not depend on alterations in vascular function (Lieth et al., 2000a).

The second assumption regarding the singular causative role of glucose appears to be based on the observation that insulin does not increase glucose uptake in retinal endothelium via the GLUT1 glucose transporter. Moreover, the results of the Diabetes Control and Complications Trial (DCCT) (DCCT Research Group, 1993) have been misinterpreted as ‘‘proving’’ that hyperglycemia alone accounts for retinopathy and other complications (Davis and Granner, 1996).

The major indices of metabolic control in the DCCT were hemoglobin A1c and blood glucose values; these tests reflect the net carbohydrate metabolism (intake minus utilization) and the degree of insulin action.

Deficient insulin action (qualitative or quantitative)— rather than hyperglycemia—is the cardinal metabolic feature of diabetes, and the improved control achieved in the DCCT was accomplished via significantly higher insulin doses (DCCT, 1995). Thus, the major question relevant to retinopathy research remains: what is the mechanism of the effects observed in the DCCT? Is it from obvious changes such as lowering the glucose burden or higher insulin levels and greater insulin action, or from effects on amino acid or lipid metabolism that are also central to diabetes? At this point the answer remains elusive but it is clearly premature to exclude factors other than glucose. In

fact, no single parameter may account for vascular and neural lesions.

With these points in mind the question may be asked whether retinal insulin signaling changes in diabetes and if these changes are relevant to the genesis of retinopathy. The published work in this area is remarkably limited. Zetterstrom et al. (1992) found that retinas from rats with 4 weeks of experimental diabetes expressed higher IR levels than controls, and this increase reflected a doubling of the neuronal and a 20% decrease in the peripheral (vascular) IR subtypes. Neuronal insulin receptors in wheat-germ agglutinin-purified cortical synaptosomal membranes were also increased, and the receptors exhibited normal autophosphorylation properties. The authors concluded that retinal and cortical IR are sensitive to circulating insulin and glucose levels.

Wheat-germ agglutinin binds to sialic acid residues so the method employed to isolate IRs in this study selected for vascular over non-vascular IR. Studies of retinal IR kinase activity are now ongoing in our lab.

It is not possible to investigate retinal IR in intact human eyes, and the only one study has examined vitreous insulin levels in patients (Feman et al., 1978).

However, details of the subjects’ medical history, medications, diabetes type, or duration were not included, making interpretation of the data difficult.

Type I diabetic patients may have reduced vitreous insulin, whereas hyperinsulinemic Type II patients may have normal or elevated vitreous insulin concentrations.

Insulin levels have been reported to be reduced by approximately half in the vitreous of rats with 4 weeks of streptozotocin-induced diabetes (Shires et al., 1992).

Thus, it appears that circulating insulin readily gains access to the vitreous cavity in normal and diabetic rats, and that at least part of the vitreous insulin content derives from the pancreas. Unfortunately it is difficult to investigate this point in humans.

While long-term intensive insulin therapy clearly reduces the risk of retinopathy progression, a minority of patients may develop exacerbation of their retinopathy after institution of tight control. In the DCCT, 13% of intensively treated patients versus 7.6% of conventionally treated patients who had retinopathy at baseline developed worsening within 6 months of treatment (Anonymous, 1998). However, only two of 1441 patients developed proliferative retinopathy and 3 developed clinically significant macular edema requiring laser treatment. Thus, ‘‘early worsening’’ is uncommon and seldom associated with vision loss. Recently, Poulaki et al. (2002) found that subcutaneous insulin implants delivering low-dose (2 units) of insulin daily increased blood–retinal barrier permeability and VEGF expression in rats with short-term streptozotocin diabetes. These short term responses may reflect impaired ability of diabetic rat retinas to respond to insulin therapy, but are in contrast to our observations that 3

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

days of insulin treatment substantially improves tight junction protein expression and glial reactivity and glutamine synthetase activity without complete restoration of euglycemia (Barber et al., 2000; Lieth et al., 2000b), or that chronic insulin therapy reduces apoptosis (Barber et al., 1998). The long-term benefits of intensive insulin therapy are unequivocal, and one might reasonably interpret the DCCT data to represent an insulin dose–response trial (Fig. 2). If so, systemically administered insulin may be a direct pharmacologic intervention for the retina and other complicationsprone tissues, including the kidneys and peripheral nerves. Indeed, a direct role for insulin on retinal cell viability is strongly suggested by the observations that in developing chicken retinas, insulin reduces apoptotic cell death (Diaz et al., 1999), and insulin antibodies induce ganglion cell apoptosis (Diaz et al., 2000). These observations have yet to be confirmed in human retinas but the high degree of homology in insulin signaling and retinal development across species suggests an important role for insulin in the maintenance of retinal cell survival as the insulin-signaling pathway is a key determinant of neuronal survival in species from C. elegans (Lin et al., 2001), Drosophila melanogaster (Brogiolo et al., 2001; Chen et al., 1996), and mice (White, 2002b). We have also found that insulin is a potent inhibitor of cell death in rat retinal neurons in culture in response to serum deprivation (Barber et al., 2001). Thus, the cumulative data support a direct and beneficial role of insulin on retinal cell function and survival during embryologic development and stressful conditions such as nutritional deprivation and hypoxia.

In clinical medicine, exceptional cases often provide opportunities to gain new insights into the specific etiologic factors in disease pathogenesis. Although

557

hyperglycemia is most often considered to be a strong predictor of diabetes complications, more than 15 welldocumented cases of diabetic retinopathy and nephropathy have been described in patients without overt hyperglycemia or glucose intolerance (Barnes et al., 1985; Chan et al., 1985; Collens et al., 1959; Harrower and Clark, 1976; Johansen, 1969). One of us (TWG) has also evaluated two obese adult males with classical proliferative diabetic retinopathy who did not have blood glucose values greater than 120 mg=dl (unpublished data). Together, these cases suggest that hyperglycemia may not be essential for the development or progression of retinopathy. How could impaired insulin action, and thus diabetes, not alter glucose levels and yet result in retinopathy? Insulin has protean consequences from tissue to tissue and cell to cell that are modified by substrate levels and interactions with other hormones and growth factors. The signaling pathways for insulin are highly complex and divergent, and many probably remain undiscovered. It is conceivable that some patients may have selective impairments in insulin action that affect glucose uptake and metabolism less than other aspects of diabetes such as lipid metabolism as suggested by Chaturvedi (2002). That is, specific mutations in the complex signaling pathways of insulin could lead to a phenotype in which lipid or amino acid metabolism is defective with minimal changes in glucose metabolism, so patients may be ‘‘diabetic’’ in terms of overall insulin action without frank hyperglycemia. This concept requires additional investigation.

Further evidence for a role of insulin action derives from observations that insulin sensitivity is an important determinant for the progression of retinopathy in Types I and II diabetes. The European Diabetes Prospective Diabetes Complications Study (EURODIAB-PCS)

Fig. 2. In individuals with IDDM, the maintainance of retinal function depends on insulin. Adapted from the DCCT trial (N. Engl. J. Med. 329, 977–986), the percentage of patients developing new retinopathy (y-axis), is significantly reduced in individuals receiving greater amounts of insulin per day (intensive insulin therapy). Similar results were also found in patients with pre-existing mild diabetic retinopathy. Therefore, insulin per se may have therapeutic implications in preventing diabetic complications.

558

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

found that insulin resistance (as measured by body mass index and triglyceride levels) is nearly as an important predictor of retinopathy as is HBA1c (Chaturvedi et al., 2001). Similar results have also been found in the DCCT cohort (Zhang et al., 2001), and these findings confirm earlier studies showing that acute insulin sensitivity is a strong predictor of retinopathy and other vascular complications in Type I diabetes (Martin and Hopper, 1987), and that patients with retinopathy are more insulin resistant than those without retinopathy (Maneschi et al., 1983). Insulin sensitivity was also found to be a major risk factor for the presence of retinopathy at the time of Type II diabetes diagnosis in the United Kingdom Prospective Diabetes Study (UKPDS) (Kohner et al., 1998).

Taken together, these laboratory, epidemiologic, and clinical trial data strongly suggest that altered insulin action, at least at the systemic level, plays an important role in the pathogenesis of diabetic retinopathy. It is now important to determine if retina-specific changes in insulin action has causal relationship to cell death, vascular permeability, and vision loss. Better understanding of the importance of insulin sensitivity and action will provide the opportunity to develop new drugs that enhance the general or specific actions of insulin and the insulin receptor.

6. Conclusions and future directions Genes and their protein, lipid, nucleic acid, and carbohydrate products are assigned names based on their initial context of tissue origin or predicted function. Inevitably the scope of understanding of the expression and function of genes expands over time. For example, presumed gastric hormones such as gastrin and cholecystokinin, are also expressed in the CNS. Thus, classical concepts of gene function and disease pathogenesis is often limited by assumptions. Three such instances are been pointed out in this review. First, that insulin is solely of pancreatic origin and functions only to control glucose levels; second, that diabetic retinopathy is purely a vascular disease; and third, that diabetic retinopathy development is related purely to plasma glucose levels. By examining these assumptions it is possible to develop new concepts about the role of insulin in normal retinal physiology and in disease pathogenesis.

Several additional questions can now be asked based on these concepts: 1. Does retinal insulin content change in diabetes?

2. Does retinal insulin serve a metabolic or trophic role in adult mammalian retinas, and, if so, what is the mechanism?

3. Does retinal insulin receptor signaling change in diabetes or other disease states, and is it causative for the pathology?

Insulin undoubtedly functions as but one factor in a complex system to regulate retinal processes. However, as the primary anabolic peptide hormone it undoubtedly makes a major contribution to normal vision. Further studies will be required to answer key questions that will provide the information needed to better prevent and treat retinal diseases.

Acknowledgements This work is supported by the ADA, JDRF, NIH EY12021, The Pennsylvania Lions Sight Conservation & Eye Research Foundation, Fight for Sight, and Mr. and Mrs. Jack Turner, Athens, GA.

References Abbott, M.A., Wells, D.G., Fallon, J.R., 1999. The insulin receptor tyrosine kinase substrate p58/53 and the insulin receptor are components of CNS synapses. J. Neurosci. 19, 7300–7308.

Aiello, L.P., Gardner, T.W., 2000. Proliferative diabetic retinopathy.

The growth factor paradigm. In: Flynn, H.F., Smiddy, W. (Eds.), Diabetic Retinopathy. American Academy of Ophthalmology, San Francisco, pp. 239–273.

Andjelkovic, M., Alessi, D.R., Meier, R., Fernandez, A., Lamb, N.J., Frech, M., Cron, P., Cohen, P., Lucocq, J.M., Hemmings, B.A., 1997. Role of translocation in the activation and function of protein kinase B. J. Biol. Chem. 272, 31515–31524.

Anonymous, 1998. Early worsening of diabetic retinopathy in the diabetes control and complications trial. Arch. Ophthalmol. 116, 874–886.

Araki, E., Shimada, F., Uzawa, H., Mori, M., Ebina, Y., 1987.

Characterization of the promoter region of the human insulin receptor gene. Evidence for promoter activity. J. Biol. Chem. 262, 16186–16191.

Araki, E., Lipes, M.A., Patti, M.E., Bruning, J.C., Haag 3rd, B., Johnson, R.S., Kahn, C.R., 1994. Alternative pathway of insulin signalling in mice with targeted disruption of the IRS-1 gene.

Nature 372, 186–190.

Bailyes, E.M., Nave, B.T., Soos, M.A., Orr, S.R., Hayward, A.C., Siddle, K., 1997. Insulin receptor/IGF-I receptor hybrids are widely distributed in mammalian tissues: quantification of individual receptor species by selective immunoprecipitation and immunoblotting. Biochem. J. 327, 209–215.

Barber, A.J., Lieth, E., Khin, S.A., Antonetti, D.A., Buchanan, A.G., Gardner, T.W., 1998. Neural apoptosis in the retina during experimental and human diabetes. Early onset and effect of insulin. J. Clin. Invest. 102, 783–791.

Barber, A.J., Antonetti, D.A., Gardner, T.W., 2000. Altered expression of retinal occludin and glial fibrillary acidic protein in experimental diabetes. The Penn State Retina Research Group.

Invest. Ophthalmol. Vis. Sci. 41, 3561–3568.

Barber, A.J., Nakamura, M., Wolpert, E.B., Reiter, C.E.N., Seigel, G.M., Antonetti, D.A., Gardner, T.W., 2001. Insulin rescues retinal neurons from apoptosis by a phosphatidylinositol 3-kinase/

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562 Akt-mediated mechanism that reduces the activation of caspase-3.

J. Biol. Chem. 276, 32814–32821.

Barnes, A.J., Kohner, E.M., Johnston, D.G., Alberti, K.G.M.M., 1985. Severe retinopathy and mild carbohydrate intolerance: possible role of insulin deficiency and elevated circulating growth hormone. Lancet, 1465–1468.

Baskin, D.G., Porte Jr., D., Guest, K., Dorsa, D.M., 1983. Regional concentrations of insulin in the rat brain. Endocrinology 112, 898–903.

Baskin, D.G., Sipols, A.J., Schwartz, M.W., White, M.F., 1993.

Immunocytochemical detection of insulin receptor substrate-1 (IRS-1) in rat brain: colocalization with phosphotyrosine. Reg. Peptides 48, 257–266.

Bellacosa, A., Testa, J.R., Staal, S.P., Tsichlis, P.N., 1991. A retroviral oncogene, akt, encoding a serine-threonine kinase containing an SH2-like region. Science 254, 274–277.

Betz, A.L., Bowman, P.D., Goldstein, G.W., 1983. Hexose transport in microvascular endothelial cells cultured from bovine retina. Exp. Eye Res. 36, 269–277.

Bloodworth, J.M.B., 1962. Diabetic retinopathy. Diabetes 2, 1–22.

Brodsky, F.M., Chen, C.Y., Knuehl, C., Towler, M.C., Wakeham, D.E., 2001. Biological basket weaving: formation and function of clathrin-coated vesicles. Annu. Rev. Cell Dev. Biol. 17, 517–568.

Brogiolo, W., Stocker, H., Ikeya, T., Rintelen, F., Fernandez, R., Hafen, E., 2001. An evolutionarily conserved function of the Drosophila insulin receptor and insulin-like peptides in growth control. Curr. Biol. 11, 213–221.

Brownlee, M., Cerami, A., Vlassara, H., 1988. Advanced glycosylation end products in tissue and the biochemical basis of diabetic complications. N. Engl. J. Med. 318, 1315–1321.

Brunet, A., Bonni, A., Zigmond, M.J., Lin, M.Z., Juo, P., Hu, L.S., Anderson, M.J., Arden, K.C., Blenis, J., Greenberg, M.E., 1999.

Akt promotes cell survival by phosphorylating and inhibiting a Forkhead transcription factor. Cell 96, 857–868.

Brunet, A., Datta, S.R., Greenberg, M.E., 2001. Transcriptiondependent and -independent control of neuronal survival by the PI3K-Akt signaling pathway. Curr. Opin. Neurobiol. 11, 297–305.

Budd, G.C., Pansky, B., Glatzer, L., 1993. Preproinsulin mRNA in the rat eye. Invest. Ophthalmol. Vis. Sci. 34, 463–469.

Cardone, M.H., Roy, N., Stennicke, H.R., Salvesen, G.S., Franke, T.F., Stanbridge, E., Frisch, S., Reed, J.C., 1998. Regulation of cell death protease caspase-9 by phosphorylation. Science 282, 1318–1321.

Chan, J.Y., Cole, E., Hanna, A.K., 1985. Diabetic nephropathy and proliferative retinopathy with normal glucose tolerance. Diabetes Care 8, 385–390.

Chaturvedi, N., 2002. Differing aspects of insulin resistance in diabetic complications: the shape of things to come. RD Lawrence lecture 2000. Diabetic Med. 19, 973–977.

Chaturvedi, N., Sjoelie, A.K., Porta, M., Aldington, S.J., Fuller, J.H., Songini, M., Kohner, E.M., Study, T.E.P.C., 2001. Markers of insulin resistance are strong risk factors for retinopathy incidence in type 1 diabetes. Diabetes Care 24, 284–289.

Chen, C., Jack, J., Garofalo, R.S., 1996. The Drosophila insulin receptor is required for normal growth. Endocrinology 137, 846–856.

Collens, W.S., Silverstein, J.N., Dobkin, G.B., 1959. Case of a diabetic with a Kimmelstiel-Wilson syndrome and a normal glucose tolerance. Ann. Intern. Med. 50, 1282–1288.

Collier, E., Gorden, P., 1991. O-linked oligosaccharides on insulin receptor. Diabetes 40, 197–203.

Cong, L.N., Chen, H., Li, Y., Zhou, L., McGibbon, M.A., Taylor, S.I., Quon, M.J., 1997. Physiological role of Akt in insulinstimulated translocation of GLUT4 in transfected rat adipose cells.

Mol. Endocrinol. 11, 1881–1890.

559

Cross, D.A., Alessi, D.R., Cohen, P., Andjelkovich, M., Hemmings, B.A., 1995. Inhibition of glycogen synthase kinase-3 by insulin mediated by protein kinase B. Nature 378, 785–789.

Das, A., Pansky, B., Budd, G.C., Kollarits, C.R., 1984. Immunocytochemistry of mouse and human retina with antisera to insulin and S-100 protein. Curr. Eye Res. 3, 1397–1403.

Das, A., Pansky, B., Budd, G.C., 1987. Demonstration of insulinspecific mRNA in cultured rat retinal glial cells. Invest. Ophthalmol. Vis. Sci. 28, 1800–1810.

Das, A., Pansky, B., Budd, G.C., Reid, T.W., 1991. Human retinoblastoma Y79 cells contain both insulin-specific mRNA and insulin-binding sites. Neurosci. Lett. 121, 231–233.

Datta, S.R., Dudek, H., Tao, X., Masters, S., Fu, H., Gotoh, Y., Greenberg, M.E., 1997. Akt phosphorylation of BAD couples survival signals to the cell-intrinsic death machinery. Cell 91, 231–241.

Davis, S.N., Granner, D.K., 1996. Insulin, oral hypoglycemic agents, and the pharmacology of the endocrine pancreas. In: Molinoff, P.B., Ruddon, R.W. (Eds.), Goodman & Gilman’s The Pharmacological Basis of Therapeutics. McGraw-Hill, New York, pp. 1487–1517.

DCCT Research Group, 1993. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. The Diabetes Control and Complications Trial Research Group. N. Engl. J. Med. 329, 977–986.

DCCT Research Group, 1995. Implementation of treatment protocols in the diabetes control and complications trial. Diabetes Care 18, 361–376. de L.A. Fernandes, M.L., Saad, M.J., Velloso, L.A., 1999. Insulin induces tyrosine phosphorylation of the insulin receptor and SHC, and SHC/GRB2 association in cerebellum but not in forebrain cortex of rats. Brain Res. 826, 74–82. de la Rosa, E.J., Bondy, C.A., Hernandez-Sanchez, C., Wu, X., Zhou, J., Lopez-Carranza, A., Scavo, L.M., de Pablo, F., 1994. Insulin and insulin-like growth factor system components gene expression in the chicken retina from early neurogenesis until late development and their effect on neuroepithelial cells. Eur. J. Neurosci. 6, 1801–1810.

Devaskar, S.U., Giddings, S.J., Rajakumar, P.A., Carnaghi, L.R., Menon, R.K., Zahm, D.S., 1994. Insulin gene expression and insulin synthesis in mammalian neuronal cells. J. Biol. Chem. 269, 8445–8454.

Diaz, B., Pimentel, B., de Pablo, F., de La Rosa, E.J., 1999. Apoptotic cell death of proliferating neuroepithelial cells in the embryonic retina is prevented by insulin. Eur. J. Neurosci. 11, 1624–1632.

Diaz, B., Serna, J., De Pablo, F., de la Rosa, E.J., 2000. In vivo regulation of cell death by embryonic (pro)insulin and the insulin receptor during early retinal neurogenesis. Dev. Suppl.

127, 1641–1649.

Dudek, H., Datta, S.R., Franke, T.F., Birnbaum, M.J., Yao, R., Cooper, G.M., Segal, R.A., Kaplan, D.R., Greenberg, M.E., 1997.

Regulation of neuronal survival by the serine–threonine protein kinase Akt. Science 275, 661–665.

Dupont, J., Derouet, M., Simon, J., Taouis, M., 1998. Effect of nutritional state on the formation of a complex involving insulin receptor IRS-1, the 52 kDa Src homology/collagen protein (Shc) isoform and phosphatidylinositol 30 -kinase activity. Biochem. J. 335, 293–300.

Earnshaw, W.C., Martins, L.M., Kaufmann, S.H., 1999. Mammalian caspases: structure, activation, substrates, and functions during apoptosis. Annu. Rev. Biochem. 68, 383–424.

Ebina, Y., Ellis, L., Jarnagin, K., Edery, M., Graf, L., Clauser, E., Ou, J.H., Masiarz, F., Kan, Y.W., Goldfine, I.D., 1985. The human insulin receptor cDNA: the structural basis for hormone-activated transmembrane signalling. Cell 40, 747–758.

560

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

Escurat, M., Djabali, K., Huc, C., Landon, F., Becourt, C., Boitard, C., Gros, F., Portier, M.M., 1991. 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. Dev. Neurosci. 13, 424–432.

Federici, M., Lauro, D., D’Adamo, M., Giovannone, B., Porzio, O., Mellozzi, M., Tamburrano, G., Sbraccia, P., Sesti, G., 1998.

Expression of insulin/IGF-I hybrid receptors is increased in skeletal muscle of patients with chronic primary hyperinsulinemia.

Diabetes 47, 87–92.

Feman, S.S., Turinsky, J., Lam, K.-W., 1978. The insulin concentration in human ocular fluids. Am. J. Ophthalmol. 85, 387–391.

Ferguson, T.A., Griffith, T.S., 1997. A vision of cell death: insights into immune privilege. Immunol. Rev. 156, 167–184.

Fernandez, R., Tabarini, D., Azpiazu, N., Frasch, M., Schlessinger, J., 1995. The Drosophila insulin receptor homolog: a gene essential for embryonic development encodes two receptor isoforms with different signaling potential. EMBO J. 14, 3373–3384.

Fischer, A.J., McGuire, C.R., Dierks, B.D., Reh, T.A., 2002. Insulin and fibroblast growth factor 2 activate a neurogenic program in Muller glia of the chicken retina. J. Neurosci. 22, 9387–9398.

Folli, F., Saad, M.J., Backer, J.M., Kahn, C.R., 1992. Insulin stimulation of phosphatidylinositol 3-kinase activity and association with insulin receptor substrate 1 in liver and muscle of the intact rat. J. Biol. Chem. 267, 22171–22177.

Folli, F., Bonfanti, L., Renard, E., Kahn, C.R., Merighi, A., 1994.

Insulin receptor substrate-1 (IRS-1) distribution in the rat central nervous system. J. Neurosci. 14, 6412–6422.

Fruman, D.A., Meyers, R.E., Cantley, L.C., 1998. Phosphoinositide kinases. Annu. Rev. Biochem. 67, 481–507.

Gambineri, A., Pelusi, C., Vicennati, V., Pagotto, U., Pasquali, R., 2002. Obesity and the polycystic ovary syndrome. Int. J. Obesity 26, 883–896.

Ganong, W.F., 1997. Review of Medical Physiology. Appleton & Lange, Stamford, CT. Garcia-de Lacoba, M., Alarcon, C., de la Rosa, E.J., de Pablo, F., 1999. Insulin/insulin-like growth factor-I hybrid receptors with high affinity for insulin are developmentally regulated during neurogenesis. Endocrinology 140, 233–243.

Gardner, T.W., Antonetti, D.A., Barber, A.J., Quillen, D.A., Neely, K.A., 2002. Recent advances in diabetic retinopathy research: bench and bedside. Curr. Conc. Ophthalmol. 10, 43–46.

Garofalo, R.S., 2002. Genetic analysis of insulin signaling in Drosophila. Trends Endocrinol. Metabol. 13, 156–162.

Ghirlanda, G., Di Leo, M.A., Caputo, S., Cercone, S., Greco, A.V., 1997. From functional to microvascular abnormalities in early diabetic retinopathy. Diabet.-Metabol. Rev. 13, 15–35.

Gillies, M.C., Su, T., Naidoo, D., 1995. Electrical resistance and macromolecular permeability of retinal capillary endothelial cells in vitro. Curr. Eye Res. 14, 435–442.

Goldberg, J.L., Espinosa, J.S., Xu, Y., Davidson, N., Kovacs, G.T., Barres, B.A., 2002. Retinal ganglion cells do not extend axons by default: promotion by neurotrophic signaling and electrical activity. Neuron 33, 689–702.

Gosbell, A., Favilla, I., Jablonski, P., 1996. The effects of insulin on the electroretinogram of bovine retina in vitro. Curr. Eye Res. 15, 1132–1137.

Gosbell, A.D., Favilla, I., Jablonski, P., 2002. The location of insulin receptors in bovine retina and isolated retinal cells. Clin. Exp. Ophthalmol. 30, 124–130.

Hachiya, H.L., Halban, P.A., King, G.L., 1988. Intracellular pathways of insulin transport across vascular endothelial cells. Am. J. Physiol. 255, C459–C464.

Harik, S.I., Kalaria, R.N., Whitney, P.M., Andersson, L., Lundahl, P., Ledbetter, S.R., Perry, G., 1990. Glucose transporters are

abundant in cells with ‘‘occluding’’ junctions at the blood-eye barriers. Proc. Natl. Acad. Sci. USA 87, 4261–4264.

Harrower, A.D.B., Clark, B.F., 1976. Diabetic retinopathy with normal glucose tolerance. Br. J. Ophthalmol. 60, 459–463.

Haskell, J.F., Meezan, E., Pillion, D.J., 1984. Identification and characterization of the insulin receptor of bovine retinal microvessels. Endocrinology 115, 698–704.

Havrankova, J., Roth, J., Brownstein, M., 1978a. Insulin receptors are widely distributed in the central nervous system of the rat. Nature 272, 827–829.

Havrankova, J., Schmechel, D., Roth, J., Brownstein, M., 1978b.

Identification of insulin in rat brain. Proc. Natl. Acad. Sci. USA 75, 5737–5741.

Hedo, J.A., Kasuga, M., Van Obberghen, E., Roth, J., Kahn, C.R., 1981. Direct demonstration of glycosylation of insulin receptor subunits by biosynthetic and external labeling: evidence for heterogeneity. Proc. Natl. Acad. Sci. USA 78, 4791–4795.

Heidenreich, K.A., Brandenburg, D., 1986. Oligosaccharide heterogeneity of insulin receptors. Comparison of N-linked glycosylation of insulin receptors in adipocytes and brain. Endocrinology 118, 1835–1842.

Heidenreich, K.A., Zahniser, N.R., Berhanu, P., Brandenburg, D., Olefsky, J.M., 1983. Structural differences between insulin receptors in the brain and peripheral target tissues. J. Biol. Chem. 258, 8527–8530.

Heidenreich, K.A., de Vellis, G., Gilmore, P.R., 1988. Functional properties of the subtype of insulin receptor found on neurons.

J. Neurochem. 51, 878–887.

Hermann, C., Assmus, B., Urbich, C., Zeiher, A.M., Dimmeler, S., 2000. Insulin-mediated stimulation of protein kinase Akt: a potent survival signaling cascade for endothelial cells. Arterioscler.

Thromb. Vasc. Biol. 20, 402–409.

Hernandez-Sanchez, C., Lopez-Carranza, A., Alarcon, C., de La Rosa, E.J., de Pablo, F., 1995. Autocrine/paracrine role of insulin-related growth factors in neurogenesis: local expression and effects on cell proliferation and differentiation in retina. Proc. Natl. Acad. Sci. USA 92, 9834–9838.

Holdengreber, V., Ren, Y., Ben-Shaul, Y., Hausman, R.E., 1998.

Co-localization of the insulin receptor, jun protein and choline acetyltransferase in embryonic chick retina. Exp. Eye Res. 66, 307–313.

Im, J.H., Pillion, D.J., Meezan, E., 1986. Comparison of insulin receptors from bovine retinal blood vessels and nonvascular retinal tissue. Invest. Ophthalmol. Vis. Sci. 27, 1681–1690.

Ito, Y., Ariga, M., Takahashi, S., Takenaka, A., Hidaka, T., Noguchi, T., 1997. Changes in tyrosine phosphorylation of insulin receptor and insulin receptor substrate-1 (IRS-1) and association of p85 of phosphatidylinositol 3-kinase with IRS-1 after feeding in rat liver in vivo. J. Endocrinol. 154, 267–273.

James, C.R., Cotlier, E., 1983. Fate of insulin in the retina: an autoradiographic study. Br. J. Ophthal. 67, 80–88.

Jialal, I., Crettaz, M., Hachiya, H.L., Kahn, C.R., Moses, A.C., Buzney, S.M., King, G.L., 1985. Characterization of the receptors for insulin and the insulin-like growth factors on micro- and macrovascular tissues. Endocrinology 117, 1222–1229.

Johansen, K., 1969. Diabetic angiopathy in a patient with normal glucose tolerance and plasma insulin response. Am. J. Med. 47, 487–491.

Kandel, E.S., Hay, N., 1999. The regulation and activities of the multifunctional serine/threonine kinase Akt/PKB. Exp. Cell Res. 253, 210–229.

Kimura, K.D., Tissenbaum, H.A., Liu, Y., Ruvkun, G., 1997. daf-2, an insulin receptor-like gene that regulates longevity and diapause in Caenorhabditis elegans. Science 277, 942–946.

King, G.L., Johnson, S.M., 1985. Receptor-mediated transport of insulin across endothelial cells. Science 227, 1583–1586.

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562 King, G.L., Goodman, A.D., Buzney, S., Moses, A., Kahn, C.R., 1985. Receptors and growth-promoting effects of insulin and insulinlike growth factors on cells from bovine retinal capillaries and aorta. J. Clin. Invest. 75, 1028–1036.

King, G.L., Kunisaki, M., Nishio, Y., Inoguchi, T., Shiba, T., Xia, P., 1996. Biochemical and molecular mechanisms in the development of diabetic vascular complications. Diabetes 45 (Suppl. 3), S105–S108.

Kohn, A.D., Summers, S.A., Birnbaum, M.J., Roth, R.A., 1996.

Expression of a constitutively active Akt Ser/Thr kinase in 3T3-L1 adipocytes stimulates glucose uptake and glucose transporter 4 translocation. J. Biol. Chem. 271, 31372–31378.

Kohner, E.M., Aldington, S.J., Stratton, I.M., Manley, S.E., Holman, R.R., Matthews, D.R., Turner, R.C., 1998. United Kingdom Prospective Diabetes Study, 30: diabetic retinopathy at diagnosis of non-insulin-dependent diabetes mellitus and associated risk factors.

Arch. Ophthalmol. 116, 297–303.

Kowluru, R.A., Tang, J., Kern, T.S., 2001. Abnormalities of retinal metabolism in diabetes and experimental galactosemia. VII. Effect of long-term administration of antioxidants on the development of retinopathy. Diabetes 50, 1938–1942.

Kumagai, A.K., 1999. Glucose transport in brain and retina: implications in the management and complications of diabetes.

Diabet.-Metabol. Rev. 15, 261–273.

Kyriakis, J.M., Hausman, R.E., Peterson, S.W., 1987. Insulin stimulates choline acetyltransferase activity in cultured embryonic chicken retina neurons. Proc. Natl. Acad. Sci. USA 84, 7463–7467.

Lansel, N., Niemeyer, G., 1997. Effects of insulin under normal and low glucose on retinal electrophysiology in the perfused cat eye.

Invest. Ophthalmol. Vis. Sci. 38, 792–799.

Leconte, I., Auzan, C., Debant, A., Rossi, B., Clauser, E., 1992.

N-linked oligosaccharide chains of the insulin receptor beta subunit are essential for transmembrane signaling. J. Biol. Chem. 267, 17415–17423.

LeRoith, D., Adamo, M., Shemer, J., Waldbillig, R., Lesniak, M.A., dePablo, F., Hart, C., Roth, J., 1988. Insulin-related materials in the nervous system of vertebrates and non-vertebrates: possible extrapancreatic production. Hormone Metabol. Res. 20, 411–420.

Lieth, E., Gardner, T.W., Barber, A.J., Antonetti, D.A., The Penn State Retina Research Group, 2000a. Retinal neurodegeneration: early pathology in diabetes. Clin. Exp. Ophthalmol. 28, 3–8.

Lieth, E., LaNoue, K.F., Antonetti, D.A., Ratz, M., The Penn State Retina Research Group, 2000b. Diabetes reduces glutamate oxidation and glutamine synthesis in the retina. Exp. Eye Res. 70, 723–730.

Lin, K., Hsin, H., Libina, N., Kenyon, C., 2001. Regulation of the Caenorhabditis elegans longevity protein DAF-16 by insulin/IGF-1 and germline signaling. Nat. Gen. 28, 139–145.

Maneschi, F., Mashiter, K., Kohner, E.M., 1983. Insulin resistance and insulin deficiency in diabetic retinopathy of non-insulindependent diabetes. Diabetes 32, 82–87.

Martin, F.I.R., Hopper, J.L., 1987. The relationship of acute insulin sensitivity to the progression of vascular disease in long-term Type I (insulin-dependent) diabetes mellitus. Diabetologia 30, 149–153.

Meyer-Franke, A., Kaplan, M.R., Pfrieger, F.W., Barres, B.A., 1995.

Characterization of the signaling interactions that promote the survival and growth of developing retinal ganglion cells in culture.

Neuron 15, 805–819.

Myers Jr., M.G., Backer, J.M., Sun, X.J., Shoelson, S., Hu, P., Schlessinger, J., Yoakim, M., Schaffhausen, B., White, M.F., 1992.

IRS-1 activates phosphatidylinositol 30 -kinase by associating with src homology 2 domains of p85. Proc. Natl. Acad. Sci. USA 89, 10350–10354.

Naeser, P., 1997. Insulin receptors in human ocular tissues. Immunohistochemical demonstration in normal and diabetic eyes. Upsala J. Med. Sci. 102, 35–40.

561

Niswender, K.D., Morrison, C.D., Clegg, D.J., Olson, R., Baskin, D.G., Myers Jr., M.G., Seeley, R.J., Schwartz, M.W., 2003. Insulin activation of phosphatidylinositol 3-kinase in the hypothalamic arcuate nucleus: a key mediator of insulin-induced anorexia.

Diabetes 52, 227–231.

Nystrom, F.H., Quon, M.J., 1999. Insulin signalling: metabolic pathways and mechanisms for specificity. Cell Signal 11, 5 63–574.

Ogg, S., Paradis, S., Gottlieb, S., Patterson, G.I., Lee, L., Tissenbaum, H.A., Ruvkun, G., 1997. The Fork head transcription factor DAF16 transduces insulin-like metabolic and longevity signals in C. elegans. Nature 389, 994–999.

Ota, A., Shemer, J., Pruss, R.M., Lowe Jr., W.L., LeRoith, D., 1988.

Characterization of the altered oligosaccharide composition of the insulin receptor on neural-derived cells. Brain Res. 443, 1–11.

Pansky, B., Das, A., Budd, G.C., Reid, T.W., 1986. Insulin-like immunoreactivity in human retinoblastoma Y79 cell line. Neurosci.

Lett. 68, 187–191.

Paradis, S., Ruvkun, G., 1998. Caenorhabditis elegans Akt/PKB transduces insulin receptor-like signals from AGE-1 PI3 kinase to the DAF-16 transcription factor. Genes Dev. 12, 2488–2498.

Pierce, S.B., Costa, M., Wisotzkey, R., Devadhar, S., Homburger, S.A., Buchman, A.R., Ferguson, K.C., Heller, J., Platt, D.M., Pasquinelli, A.A., Liu, L.X., Doberstein, S.K., Ruvkun, G., 2001.

Regulation of DAF-2 receptor signaling by human insulin and ins1, a member of the unusually large and diverse C. elegans insulin gene family. Genes Dev. 15, 672–686.

Poulaki, V., Qin, W., Joussen, A.M., Hurlbut, P., Wiegand, S.J., Rudge, J., Yancopoulos, G.D., Adamis, A.P., 2002. Acute intensive insulin therapy exacerbates diabetic blood–retinal barrier breakdown via hypoxia-inducible factor-1alpha and VEGF. J. Clin. Invest. 109, 805–815.

Puro, D.G., Agardh, E., 1984. Insulin-mediated regulation of neuronal maturation. Science 225, 1170–1172.

Rajala, R.V., Anderson, R.E., 2001. Interaction of the insulin receptor beta-subunit with phosphatidylinositol 3-kinase in bovine ROS. Invest. Ophthalmol. Vis. Sci. 42, 3110–3117.

Rajala, R.V., McClellan, M.E., Ash, J.D., Anderson, R.E., 2002. In vivo regulation of phosphoinositide 3-kinase in retina through light-induced tyrosine phosphorylation of the insulin receptor beta subunit. J. Biol. Chem. 277, 43319–43326.

Reiter, C.E.N., Sandirasegarane, L., Wolpert, E.B., Klinger, M., Barber, A.J., Simpson, I.A., Antonetti, D.A., Kester, M., Gardner, T.W., 2003. Characterization of early insulin signaling events in rat retina, submitted.

Ren, Y., Holdengreber, V., Ben-Shaul, Y., Shah, B.H., Varanasi, J., Hausman, R.E., 1997. Causal role for jun protein in the stimulation of choline acetyltransferase by insulin in embryonic chick retina.

Biochem. Biophys. Res. Commun. 232, 788–793.

Rodrigues, M., Waldbillig, R.J., Rajagopalan, S., Hackett, J., LeRoith, D., Chader, G.J., 1988. Retinal insulin receptors: localization using a polyclonal anti-insulin receptor antibody.

Brain Res. 443, 389–394.

Rungger-Brandle, E., Dosso, A.A., Leuenberger, P.M., 2000. Glial reactivity, an early feature of diabetic retinopathy. Invest.

Ophthalmol. Vis. Sci. 41, 1971–1980.

Saltiel, A.R., Pessin, J.E., 2002. Insulin signaling pathways in time and space. Trends Cell Biol. 12, 65–71.

Schechter, R., Beju, D., Gaffney, T., Schaefer, F., Whetsell, L., 1996.

Preproinsulin I and II mRNAs and insulin electron microscopic immunoreaction are present within the rat fetal nervous system.

Brain Res. 736, 16–27.

Schmelzle, T., Hall, M.N., 2000. TOR, a central controller of cell growth. Cell 103, 253–262.

Schwartz, M.W., Sipols, A., Kahn, S.E., Lattemann, D.F., Taborsky Jr., G.J., Bergman, R.N., Woods, S.C., Porte Jr., D., 1990. Kinetics

562

C.E.N. Reiter, T.W. Gardner / Progress in Retinal and Eye Research 22 (2003) 545–562

and specificity of insulin uptake from plasma into cerebrospinal fluid. Am. J. Physiol. 259, E378–E383.

Schwartz, M.W., Figlewicz, D.P., Baskin, D.G., Woods, S.C., Porte Jr., D., 1992. Insulin in the brain: a hormonal regulator of energy balance. Endocrine Rev. 13, 387–414.

Seino, S., Bell, G.I., 1989. Alternative splicing of human insulin receptor messenger RNA. Biochem. Biophys. Res. Commun. 159, 312–316.

Shires, T.K., Braddock, K.J., Pulido, J.S., 1992. Insulin in the vitreous of the normal and streptozotocin-induced diabetic rat. Peptides 13, 671–675.

Simon, J., Rosebrough, R.W., McMurtry, J.P., Steele, N.C., Roth, J., Adamo, M., LeRoith, D., 1986. Fasting and refeeding alter the insulin receptor tyrosine kinase in chicken liver but fail to affect brain insulin receptors. J. Biol. Chem. 261, 17081–17088.

Stella Jr., S.L., Bryson, E.J., Thoreson, W.B., 2001. Insulin inhibits voltage-dependent calcium influx into rod photoreceptors. Neuro Report 12, 947–951.

Stitt, A.W., Anderson, H.R., Gardiner, T.A., Bailie, J.R., Archer, D.B., 1994. Receptor-mediated endocytosis and intracellular trafficking of insulin and low-density lipoprotein by retinal vascular endothelial cells. Invest. Ophthalmol. Vis. Sci. 35, 3384–3392.

Sugasawa, K., Deguchi, J., Okami, T., Yamamoto, A., Omori, K., Uyama, M., Tashiro, Y., 1994. Immunocytochemical analyses of distributions of Na, K-ATPase and GLUT1, insulin and transferrin receptors in the developing retinal pigment epithelial cells. Cell Struct. Funct. 19, 21–28.

Tamemoto, H., Kadowaki, T., Tobe, K., Yagi, T., Sakura, H., Hayakawa, T., Terauchi, Y., Ueki, K., Kaburagi, Y., Satoh, S., et al., 1994. Insulin resistance and growth retardation in mice lacking insulin receptor substrate-1. Nature 372, 182–186.

Testa, J.R., Bellacosa, A., 2001. AKT plays a central role in tumorigenesis. Proc. Natl. Acad. Sci. USA 98, 10983–10985.

Ullrich, A., Bell, J.R., Chen, E.Y., Herrera, R., Petruzzelli, L.M., Dull, T.J., Gray, A., Coussens, L., Liao, Y.C., Tsubokawa, M., 1985.

Human insulin receptor and its relationship to the tyrosine kinase family of oncogenes. Nature 313, 756–761.

Varandani, P.T., Organisciak, D.T., Nafz, M.A., 1982. The occurrence of glutathione-insulin transhydrogenase in the retina. Invest.

Ophthalmol. Vis. Sci. 22, 715–719.

Virkamaki, A., Ueki, K., Kahn, C.R., 1999. Protein–protein interaction in insulin signaling and the molecular mechanisms of insulin resistance. J. Clin. Invest. 103, 931–943.

Waldbillig, R.J., Fletcher, R.T., Chader, G.J., Rajagopalan, S., Rodrigues, M., LeRoith, D., 1987a. Retinal insulin receptors. 1.

Structural heterogeneity and functional characterization. Exp. Eye Res. 45, 823–835.

Waldbillig, R.J., Fletcher, R.T., Chader, G.J., Rajagopalan, S., Rodrigues, M., LeRoith, D., 1987b. Retinal insulin receptors. 2.

Characterization and insulin-induced tyrosine kinase activity in bovine retinal rod outer segments. Exp. Eye Res. 45, 837–844.

Waldbillig, R.J., Fletcher, R.T., Somers, R.L., Chader, G.J., 1988.

IGF-I receptors in the bovine neural retina: structure, kinase activity and comparison with retinal insulin receptors. Exp. Eye Res. 47, 587–607.

Waldbillig, R.J., Arnold, D.R., Fletcher, R.T., Chader, G.J., 1991.

Insulin and IGF-I binding in developing chick neural retina and pigment epithelium: a characterization of binding and structural differences. Exp. Eye Res. 53, 13–22.

White, M.F., 2002a. IRS proteins and the common path to diabetes.

Am. J. Physiol. Endocrinol. Metab. 283, E413–E422.

White, M.F., 2002b. IRS proteins and the common path to diabetes [abstract]. 2002 Annual Meeting Abstract and Program Planner accessed at www.arvo.org. Association for Research in Vision and Ophthalmology, Abstract 6.

Withers, D.J., Gutierrez, J.S., Towery, H., Burks, D.J., Ren, J.M., Previs, S., Zhang, Y., Bernal, D., Pons, S., Shulman, G.I., BonnerWeir, S., White, M.F., 1998. Disruption of IRS-2 causes type 2 diabetes in mice. Nature 391, 900–904.

Wolkow, C.A., Kimura, K.D., Lee, M.S., Ruvkun, G., 2000.

Regulation of C. elegans life-span by insulinlike signaling in the nervous system. Science 290, 147–150.

Wolter, J.R., 1961. Diabetic retinopathy. Am. J. Ophthalmol. 51, 1123–1139.

Yamagishi, S., Kawakami, T., Fujimori, H., Yonekura, H., Tanaka, N., Yamamoto, Y., Urayama, H., Watanabe, Y., Yamamoto, H., 1999. Insulin stimulates the growth and tube formation of human microvascular endothelial cells through autocrine vascular endothelial growth factor. Microvasc. Res. 57, 329–339.

Yamaguchi, A., Tamatani, M., Matsuzaki, H., Namikawa, K., Kiyama, H., Vitek, M.P., Mitsuda, N., Tohyama, M., 2001. Akt activation protects hippocampal neurons from apoptosis by inhibiting transcriptional activity of p53. J. Biol. Chem. 276, 5256–5264.

Yang, J.W., Fellows, R.E., 1980. Characterization of insulin stimulation of the incorporation of radioactive precursors into macromolecules in cultured rat brain cells. Endocrinology 107, 1717–1724.

Yanoff, M., Fine, S.L., 1989. Ocular Pathology. Mosby-Wolfe, London, p. 564.

Yeh, T.C., Ogawa, W., Danielsen, A.G., Roth, R.A., 1996. Characterization and cloning of a 58/53-kDa substrate of the insulin receptor tyrosine kinase. J. Biol. Chem. 271, 2921–2928.

Yenush, L., White, M.F., 1997. The IRS-signalling system during insulin and cytokine action. Bioessays 19, 491–500.

Yorek, M.A., Dunlap, J.A., Ginsberg, B.H., 1987. Amino acid and putative neurotransmitter transport in human Y79 retinoblastoma cells. Effect of insulin and insulin-like growth factor. J. Biol. Chem.

262, 10986–10993.

Zeng, X.X., Ng, Y.K., Ling, E.A., 2000. Neuronal and microglial response in the retina of streptozotocin-induced diabetic rats. Vis. Neurosci. 17, 463–471.

Zetterstrom, C., Benjamin, A., Rosenzweig, S.A., 1992. Differential expression of retinal insulin receptors in STZ-induced diabetic rats.

Diabetes 41, 818–825.

Zhang, L., Krzentowski, G., Albert, A., Lefebvre, P.J., 2001. Risk of developing retinopathy in Diabetes Control and Complications Trial type 1 diabetic patients with good or poor metabolic control.

Diabetes Care 24, 1275–1279.