Linoleic acid and its metabolites, hydroperoxyoctadecadienoic acids, stimulate c-Fos, c-Jun, and c-Myc mRNA expression, mitogen-activated protein kinase activation, and growth in rat aortic smooth muscle cells.

Previous studies from other laboratories suggest that linoleic acid and its metabolites, hydroperoxyoctadecadienoic acids, play an important role in modulating the growth of some cells. A correlation has been demonstrated between hydroperoxyoctadecadienoic acids and conditions characterized by abnormal cell growth such as atherosclerosis and psoriasis. To determine if linoleic acid and its metabolites modulate cell growth in atherosclerosis, we measured DNA synthesis, protooncogene mRNA expression, and mitogen-activated protein kinase (MAPK) activation in vascular smooth muscle cells (VSMC). Linoleic acid induces DNA synthesis, c-fos, c-jun, and c-myc mRNA expression and MAPK activation in VSMC. Furthermore, nordihydroguaiaretic acid, a potent inhibitor of the lipoxygenase system, significantly reduced the growth-response effects of linoleic acid in VSMC, suggesting that conversion of linoleic acid to hydroperoxyoctadecadienoic acids (HPODEs) is required for these effects. HPODEs also caused significant induction of DNA synthesis, protooncogene mRNA expression, and MAPK activation in growth-arrested VSMC, suggesting that linoleic acid and its metabolic products, HPODEs, are potential mitogens in VSMC, and that conditions such as oxidative stress and lipid peroxidation which provoke the production of these substances may alter VSMC growth.

Transcriptional regulation of the human CAD gene during myeloid differentiation.

CAD codes for a trifunctional protein involved in the catalysis of the first three enzymatic activities in the de novo pyrimidine biosynthetic pathway, namely, carbamoyl-phosphate synthetase II (EC 6.3.5.5), aspartate transcarbamylase (EC 2.1.3.2), and dihydroorotase (EC 3.5.2.3). CAD regulation was studied in the human promyelocyte leukemic line HL-60 as it differentiated into monocytic or granulocytic lineages after induction by 12-O-tetradecanoylphorbol-13-acetate or trans-retinoic acid and dibutyryl cyclic AMP, respectively. Within 12 h of induction of HL-60 cells with either inducer, total cellular levels of CAD RNA essentially disappeared. On the other hand, no apparent decreases in beta-actin RNA levels were seen even 48 h after HL-60 cells were induced, as compared with untreated cells. With nuclear runoff assays, it was clearly shown that the inactivation of CAD gene expression during the induction of HL-60 cells with either inducer was at the transcriptional level. The nuclear runoff experiments also demonstrated that the CAD gene expression was shut down in less than 4 h after induction, well before morphological changes were observed in these cells. At the enzymatic level, the activity of aspartate transcarbamylase, one of the three enzymes encoded by the CAD gene, decreased by about half within 24 h of induction, suggesting a CAD protein half-life of 24 h in differentiating HL-60 cells. Nevertheless, this means that significant levels of aspartate transcarbamylase activity remained even after the cells have stopped proliferating. From the RNA data, it is clear that CAD gene expression is rapidly turned off as promyelocytes begin to terminally differentiate into macrophages and granulocytes. We suspect that the inactivation of the CAD gene in induced HL-60 cells is a consequence of the differentiating cells leaving the cell cycle and becoming nonproliferating.

Neoplasms observed in untreated and corn oil gavage control groups of F344/N rats and (C57BL/6N X C3H/HeN)F1 (B6C3F1) mice.

Control data on F344/N rats and (C57BL/6N X C3H/HeN)F1 (B6C3F1) mammary tumor virus-free mice from the National Toxicology Program (NTP) were examined to determine if animals receiving corn oil by gavage showed tumor incidences that differed from those of untreated control animals. Analyses of these data were adjusted for interlaboratory variability, time-related trends, and supplier effects. Two biologically significant effects were found: Male F344/N control rats receiving corn oil by gavage showed a higher (P less than .05) incidence of pancreatic acinar cell adenoma and a lower (P less than .001) incidence of leukemia (primarily mononuclear cell leukemia) than did the corresponding untreated controls. The increased incidences of pancreatic acinar cell adenoma seen in male rats administered corn oil by gavage were associated with elevated body weights observed in these animals relative to untreated controls. Female F344 rats and male and female B6C3F1 mice showed little or no evidence of a difference in tumor incidence between corn oil gavage-treated and untreated controls. A review of nearly 300 carcinogenesis studies done by the National Cancer Institute (NCI) and the NTP revealed that there were no corn oil gavage studies in which increased incidences of pancreatic acinar cell tumors or leukemia in male F344/N rats were the sole evidence of the carcinogenicity of a test chemical. Thus use of corn oil appears to have little impact on the interpretation of NCI-NTP carcinogenicity studies.

Allelotype analysis of 2',3'-dideoxycytidine- and 1,3-butadiene-induced lymphomas in B6C3F1 mice.

To identify potential tumor suppressor genes involved in lymphoma development, we generated allelotypes of 16 2',3'-dideoxycytidine (ddC and 31 1,3-butadiene (BD)-induced lymphomas from C57BL/6 x C3H/He F1 (hereafter called B6C3F1) mice. Two or more anonymous simple sequence length polymorphisms per autosome were examined for loss of heterozygosity (LOH). Allelic losses throughout the genome were generally infrequent, except for markers on chromosome 2, 4, 11 and 12. The highest frequency of allelic losses was observed on chromosome 12, with 38 and 39% in ddC and BD-induced lymphomas, respectively. The most prevalent LOH was localized to the distal region bounded by markers D12Mit263 and D12Nds2. No known tumor suppressor genes have been mapped to this region, and no obvious candidates could be identified, suggesting the presence of novel suppressor gene(s). LOH on chromosome 2 was observed in 31% of ddC-induced lymphomas but in only 3% (1/31) of BD-induced lymphomas, suggesting a ddC-specific genetic effect. Detailed analysis localized a potential tumor suppressor gene residing on the distal region of chromosome 2, between markers D2Mit147 and D2Mit148. Twenty-five % of ddC-induced and 23% of BD-induced lymphomas showed LOH on chromosome 4, and two discrete regions were identified. One of the regions includes the IFN gene cluster and is syntenic to human chromosome 9p2l-22. Candidate tumor suppressor genes, Mts1 (multiple tumor suppressor 1) and Mts2 have been mapped to this region. The second region is located on the distal part of chromosome 4, which is homologous to human chromosome 1p35-36, a region that is frequently deleted in various types of human tumors. Finally, 19% of ddC-induced and 29% of BD-induced lymphomas revealed LOH on chromosome 11 at the Acrb locus, which lies within 1 cM of p53, suggesting that the p53 tumor suppressor gene also plays a role in lymphomagenesis. These results suggest that multiple potential suppressor loci contribute to lymphoma development in B6C3F1 mice.

Carcinogenicity of 2',3'-dideoxycytidine in mice.

2',3'-dideoxycytidine (ddC) is a synthetic pyrimidine nucleoside analogue approved for treatment of HIV-positive patients. Previous studies indicated that ddC has the potential to cause thymic lymphoma in C57BL/6 x C3H F1 (hereafter called B6C3F1) mice. In this study, we evaluated the carcinogenic potential of ddC in two different mouse models. B6C3F1 hybrid mice carry ecotropic endogenous proviral sequences that may be activated to cause lymphoma, whereas NIH Swiss mice lack proviral sequences that can be expressed. The mice were treated with ddC by gavage at 500 and 1000 mg/kg/day for up to 6 months (human dose, 2.25 mg/day) and evaluated for toxicity, plasma levels of ddC, and pathological changes. Lymphocyte cell markers from the thymic lymphomas were assessed by immunophenotyping. Expression of p53 protein was evaluated using immunohistochemical staining. Treatment-related thymic lymphomas were present in both mouse models with a higher incidence in NIH Swiss than in B6C3F1 mice. The lymphomas were more prevalent in females than in males of both mouse models. Most mice with thymic lymphoma died during the course of the study. In addition to the thymus, lymphoma was often present in lymph nodes, spleen, and other organs. Lymphomas arose more frequently in mice that lack endogenous ecotropic retroviral sequences and thus were not due to activation of endogenous provirus. During the third month of the study, a few NIH Swiss mice that died had granulosa cell tumors of the ovary. Treatment-related but reversible thymic atrophy was observed in both mouse models. There was a very high correlation between the internal dose of ddC and the incidence of thymic lymphoma in both mouse models. Most of the lymphocytes from control thymuses and ddC-induced lymphomas were positive for Thy-1.2 (pan-T), heat stable antigen, and CD4 and CD8 markers, with no marked differences in the lymphocyte markers of the tumors between sexes or dose groups. p53 protein was detected in only 20% (23/115) of the ddC-induced lymphomas with mostly minimal expression in scattered cells. Because ddC induced lymphomas in two different mouse models, the potential carcinogenic risk should be considered in long-term treatment of HIV-positive patients, especially children and adolescent patients treated with ddC.

Hydrogen peroxide induces complex formation of SHC-Grb2-SOS with receptor tyrosine kinase and activates Ras and extracellular signal-regulated protein kinases group of mitogen-activated protein kinases.

Protein tyrosine phosphorylation events play determinant roles in cellular processes such as proliferation and differentiation. We have recently reported that hydrogen peroxide, an active oxygen species and a cellular oxidant, stimulates growth response events in vascular smooth muscle cells (VSMC). To understand the mechanisms by which oxidant stress modulates these growth response events, we have studied the effect of hydrogen peroxide on protein tyrosine phosphorylation events in VSMC. Our findings show that hydrogen peroxide stimulates tyrosine phosphorylation of several proteins including epidermal growth factor receptor (EGFR) in VSMC. Hydrogen peroxide-induced tyrosine phosphorylation of EGFR was found to be time dependent; with a threefold increase at 5 min and a 20-fold increase at 30 min of treatment as compared to control levels. Hydrogen peroxide treatment of VSMC also resulted in a time-dependent increase in tyrosine phosphorylation of SHC proteins. In addition, hydrogen peroxide-induced tyrosine-phosphorylated EGFR formed a complex with SHC-Grb2-SOS. These events were followed by activation of Ras and extracellular signal-regulated protein kinases (ERKs) group of mitogen-activated protein kinases (MAPKs). Together these findings demonstrate for the first time that hydrogen peroxide, a cellular oxidant, possess the ability to activate EGFR-mediated signaling events in VSMC. These EGFR-mediated signaling events may be important in oxidant stress-induced cellular responses.

Hydrogen peroxide stimulates transcription of c-jun in vascular smooth muscle cells: role of arachidonic acid.

We reported previously that hydrogen peroxide induces DNA synthesis in rat aortic smooth muscle (RASM) cells. In the present paper we studied the mechanism by which hydrogen peroxide induces c-jun mRNA, an early response gene whose activation is required for mitogen-stimulated cell growth. Hydrogen peroxide induced c-jun mRNA in growth-arrested RASM cells in a time dependent manner. This stimulation was significantly inhibited by mepacrine, a phospholipase A2 (PLA2) inhibitor. Arachidonic acid, a PLA2 product, also increased c-jun mRNA with a time course similar to that of hydrogen peroxide. The increases in c-jun mRNA induced by hydrogen peroxide and arachidonic acid were significantly reduced (55%) by down-regulation of protein kinase C with a phorbol ester. Furthermore, the effect of hydrogen peroxide on c-jun mRNA was also reduced by NDGA, an inhibitor of the lipoxygenase-cytochrome P450 mono-oxygenase system, suggesting that metabolism of arachidonic acid through this pathway is required for the induction of c-jun mRNA by oxidants. Both hydrogen peroxide and arachidonic acid significantly increased c-jun transcription as demonstrated by nuclear run-on assays. Together these observations suggest that: (1) the induction of c-jun mRNA by hydrogen peroxide is mediated by PLA2-dependent arachidonic acid release and metabolism through the lipoxygenase-cytochrome P450 mono-oxygenase system; (2) PKC appears to be involved in this signaling pathway and (3) the induction of c-jun mRNA by hydrogen peroxide in RASM cells is due to increased transcription.

Arachidonic acid activates Jun N-terminal kinase in vascular smooth muscle cells.

We have previously demonstrated that arachidonic acid activates extracellular signal-regulated protein kinases (ERKs) group of mitogen-activated protein kinases (MAPKs) in vascular smooth muscle cells (VSMC). To understand the role of arachidonic acid in cellular signaling events, we have now studied its effect on jun N-terminal kinases (JNKs) group of MAPKs in VSMC. Arachidonic acid activated JNK1 in a time- and concentration-dependent manner with maximum effects at 10 min and 50 microM. Induced activation of JNK1 by arachidonic acid is specific as other fatty acids such as linoleic and stearic acids had no such effect. Indomethacin and nordihydroguaiaretic acid (NDGA), potent inhibitors of the cyclooxygenase (COX) and the lipoxygenase (LOX)/monooxygenase (MOX) pathways, respectively, had no effect on arachidonic acid activation of JNK1 suggesting that the observed phenomenon is independent of its metabolism through either pathway. However, 12-hydroperoxyeicosatetraenoic acid (12-HpETE), the LOX metabolite of arachidonic acid significantly induced JNK1 activity. Protein kinase C (PKC) depletion by prolonged treatment of VSMC with phorbol 12-myristate 13-acetate (PMA) resulted in partial decrease in the responsiveness of JNK1 to arachidonic acid suggesting a role for both PKC-dependent and -independent mechanisms in the activation of JNK1 by this important fatty acid. On the other hand, the responsiveness of JNK1 to 12-HpETE was completely abolished in PKC-depleted cells, suggesting a major role for PKC in 12-HpETE-induced JNK1 activation. IL-1beta and TNF-alpha activated JNK1 in a time-dependent manner with maximum effect at 10 min. Desensitization of JNK1 by arachidonic acid significantly reduced its responsiveness to both the cytokines. In addition, 4-bromophenacyl bromide (4-BPB), a potent and selective inhibitor of phospholipase A2 (PLA2), significantly attenuated the cytokine-induced activation of JNK1. Together, these results show that (1) arachidonic acid and its LOX metabolite, 12-HpETE, activate JNK1 in VSMC, (2) PKC-dependent and -independent mechanisms play a role in the activation of JNK1 by arachidonic acid and 12-HpETE, and (3) arachidonic acid mediates, at least partially, the cytokine-induced activation of JNK1.

Differential regulation of p27kip1 levels and CDK activities by hypertrophic and hyperplastic agents in vascular smooth muscle cells.

To understand the molecular mechanisms that determine the fate of a cell to undergo either hypertrophy or hyperplasia, we studied the effects of angiotensin II (Ang II) and platelet-derived growth factor (PDGF)-BB, hypertrophic and hyperplastic agents, respectively, on the modulation of G1/S transition molecules in smooth muscle cells. Ang II increased protein synthesis while PDGF-BB induced both DNA and protein synthesis. Ang II had no significant effect on the steady-state levels of cyclin-dependent kinase (CDK) inhibitor (CDKI), p27kip1, and on the activities of CDK2 and CDK4, although it caused a modest increase in cyclin E levels. In contrast, PDGF-BB induced depletion of p27kip1 and increased cyclins D1 and E levels and CDK2 and CDK4 activities. Reflecting its lack of effect on CDK activities, Ang II failed to phosphorylate tumor suppressor retinoblastoma protein, Rb. PDGF-BB, on the other hand, induced phosphorylation of Rb, consistent with its ability to activate CDKs. Together, these findings suggest that Ang II-induced hypertrophy may be due to its failure to activate cellular signaling events required for G1/S transition.

Hydrogen peroxide activation of cytosolic phospholipase A2 in vascular smooth muscle cells.

We have reported previously that hydrogen peroxide induces arachidonic acid release from prelabeled vascular smooth muscle cells. Here, we studied the effect of hydrogen peroxide on the phosphorylation of cytosolic phospholipase A2 in these cells. Hydrogen peroxide induced a rapid, time-dependent increase in the phosphorylation of cytosolic phospholipase A2. Hydrogen peroxide also increased arachidonic acid release from prelabeled cells in a time-dependent manner similar to that of phosphorylation of cytosolic phospholipase A2. Protein kinase C depletion significantly inhibited the hydrogen peroxide-stimulated cytosolic phospholipase A2 phosphorylation and arachidonic acid release. Hydrogen peroxide caused a time-dependent increase in mitogen activated protein kinase activity. Taken together, these findings suggest that cytosolic phospholipase A2 may, at least in part, contribute to arachidonic acid release induced by hydrogen peroxide and this effect appears to be mediated by protein kinase C and mitogen activated protein kinase.

cAMP inhibits linoleic acid-induced growth by antagonizing p27(kip1) depletion, but not interfering with the extracellular signal-regulated kinase or AP-1 activities.

To understand the underlying signaling events of polyunsaturated fatty acid-induced growth, we studied the effect of cAMP on early and delayed growth response events induced by linoleic acid in smooth muscle cells (SMC). cAMP significantly inhibited both basal and linoleic acid-induced DNA synthesis. Linoleic acid-induced early growth response events, such as activation of ERKs, induction of expression of c-fos and jun-B and stimulation of AP-1 activity, however, were not affected by cAMP. In contrast, linoleic acid-induced c-jun expression was blocked by cAMP. cAMP alone stimulated ERKs activation, c-fos and jun-B expression and increased AP-1 activity. Linoleic acid induced depletion of p27kip1 and increased CDK2 activity, events required for G1/S transition. In contrast to early growth response events, linoleic acid-induced G1/S transition signals were significantly inhibited by cAMP. These findings suggest that in addition to inducing immediate early growth response events, linoleic acid mimics growth factors in activating cell cycle events that are associated with G1/S transition in SMC and the negative regulation of linoleic acid-induced growth by cAMP is apparently due to its antagonism with linoleic acid-induced p27kip1 depletion and CDK2 activation.

Protein tyrosine kinase activity is required for oxidant-induced extracellular signal-regulated protein kinase activation and c-fos and c-jun expression.

Hydrogen peroxide stimulated tyrosine phosphorylation of several proteins in growth-arrested vascular smooth muscle cells (VSMC). One of these proteins was identified as fibroblast growth factor receptor type I (FGFR1). In addition, induced tyrosine phosphorylation of FGFR1 by hydrogen peroxide resulted in complex formation with Grb2. Hydrogen peroxide also caused a time-dependent activation of extracellular signal-regulated protein kinases (ERKs; p42&p44) group of mitogen-activated protein kinases (MAPKs) in VSMC. The time courses of the hydrogen peroxide-stimulated FGFR1 tyrosine phosphorylation and ERKs activation were followed by induced expression of c-fos and c-jun. Genistein, a potent inhibitor of protein tyrosine kinases, significantly blunted the hydrogen peroxide-induced FGFR1 tyrosine phosphorylation, ERKs activation and c-fos and c-jun expression. PD98059, a specific inhibitor of MEK1, attenuated the hydrogen peroxide-induced ERKs activation and c-fos and c-jun expression. Together, these results suggest that oxidants such as hydrogen peroxide stimulate tyrosine phosphorylation of receptor tyrosine kinases and these, in turn, mediate the down-stream signalling events including the recruitment of Grb2 by the receptor, activation of ERKs and induction of c-fos and c-jun expression.

The Barrie Jones Lecture-Eye care for the neglected population: challenges and solutions.

Globally, pockets of 'neglected populations' do not have access to basic health-care services and carry a much greater risk of blindness and visual impairment. While large-scale public health approaches to control blindness due to vitamin A deficiency, onchocerciasis, and trachoma are successful, other causes of blindness still take a heavy toll in the population. High-quality comprehensive eye care that is equitable is the approach that needs wide-scale application to alleviate this inequity. L V Prasad Eye Institute of India developed a multi-tier pyramidal model of eye care delivery that encompasses all levels from primary to advanced tertiary (quaternary). This has demonstrated the feasibility of 'Universal Eye Health Coverage' covering promotive, preventive, corrective, and rehabilitative aspects of eye care. Using human resources with competency-based training, effective and cost-effective care could be provided to many disadvantaged people.

Bovine lens 23, 21 and 19 kDa intrinsic membrane proteins have an identical amino-terminal amino acid sequence.

We have isolated bovine lens intrinsic membrane proteins (MP) having molecular masses of 19, 21 and 23 kDa. Limited amino acid sequence analysis of the amino-terminal portion of each of these polypeptides revealed a 100% homology in sequence for the number of residues determined (20 amino acids). Northern blot analysis of bovine lens mRNA using a labeled antisense oligonucleotide probe common to the amino acid sequence of these three peptides revealed a single band having an apparent molecular size of 0.8 kb. Taken together, these findings suggest a genetic commonality between these polypeptides.

Posttranscriptional regulation of the expression of CAD gene during differentiation of F9 teratocarcinoma cells by induction with retinoic acid and dibutyryl cyclic AMP.

We have studied the regulation of expression of the carbamoyl-phosphate synthetase II-aspartate transcarbamylase-dihydroorotase gene in F9 teratocarcinoma cells during their differentiation into parietal endoderm cells by induction with a combination of retinoic acid and dibutyryl cyclic AMP. Steady-state levels of CAD mRNA decreased by 7-fold in F9 cells following 120 h of retinoic acid and dibutyryl cyclic AMP induction as compared to levels in uninduced cells. Conversely, no apparent changes were found in the steady-state levels of beta-actin mRNA between induced and uninduced cells. Despite a 7-fold decrease in the steady-state levels of CAD mRNA, its rate of transcription remained the same between induced and uninduced cells, indicating a role for posttranscriptional mechanisms for its down regulation during retinoic acid- and dibutyryl cyclic AMP-induced differentiation of F9 cells. The cellular growth rate of F9 cells as determined by [3H]thymidine uptake and parallel cell counting decreased markedly during their induction with retinoic acid and dibutyryl cyclic AMP. Taken together, it is apparent that the expression of the CAD gene is cell-growth-dependent and its regulation in this system is at the posttranscriptional level.

Retinal transplants into the anterior chamber of the rat eye.

Developing retinas from 13-18-day fetuses and 2-day neonatal Long-Evans rats transplanted into the anterior chamber of adult eyes of the same or different strain (Lewis) survive and differentiate. Light and electron microscopic studies show that the transplants undergo histogenetic differentiation, resulting in the development of neurons and Müller glial cells and formation of nuclear and plexiform layers. Vascular connections develop between the host iris and the retinal transplant. Vessels and nerves, presumably of iridal origin, were seen on the surface of some transplants. Possible manifestations of graft rejection were monitored; signs of tissue rejection in transplants performed in the Long-Evans rats, an outbred strain, were rare and if present they were mild, at least during the survival periods of up to 91 days allowed in these experiments. Transplants into the eyes of Lewis rats were also well tolerated during the survival period. These observations indicate that retinal transplantation to the adult eye of a genetically different host can be successfully achieved and that both embryonic and perinatal retinas are suitable as donor tissue for ocular transplants. The procedure offers ample opportunities for the study of problems related to retinal plasticity.

Free epsilon amino groups and 5-hydroxymethylfurfural contents in clear and cataractous human lenses.

The free epsilon-amino groups and 5-hydroxymethylfurfural (5-HMF) contents were determined in soluble and insoluble proteins of clear human lenses and diabetic and nondiabetic senile cataractous lenses. The free epsilon-amino group content of soluble proteins in diabetic cataracts was decreased by 37% (P less than 0.01), whereas in nondiabetic senile cataracts it did not differ from that of clear lenses. The free epsilon-amino group content of insoluble proteins both in diabetic and nondiabetic cataracts was decreased significantly (P less than 0.001, P less than 0.015, respectively). The 5-HMF content of soluble proteins in diabetic cataracts was increased by 52% (P less than 0.001), whereas in nondiabetic cataracts it did not change from that of clear lenses. The 5-HMF content of insoluble proteins in diabetic as well as in nondiabetic cataracts was increased significantly as compared to that of clear lens (P less than 0.001, P less than 0.001, respectively). The soluble protein of diabetic and nondiabetic cataracts was decreased with an increase in the insoluble protein content. These results suggest that nonenzymatic glycosylation plays a role in the conformational change of lens proteins in both diabetic and nondiabetic cataracts.

Biosynthesis of neopterin, sepiapterin, and biopterin in rat and human ocular tissues.

Neopterin, sepiapterin, and biopterin synthesis by lens, retina, and ciliary body-iris of rat and human indicates pteridine formation from their precursor, GTP. The pteridine biosynthesis was higher in the retina (neopterin 422 +/- 27, 260 +/- 24; sepiapterin 135 +/- 12, 118 +/- 14; biopterin 76 +/- 10, 68 +/- 8 nanomoles/g soluble protein/hr, in rat and human, respectively) than in the ciliary body-iris and lens. The light-sensitive pteridines may protect eye tissues against the effects of sunlight in addition to their role in the hydroxylation of aromatic amino acids.

Cell size-shape relationships in corneal endothelium.

The shape of corneal endothelial cells was studied from specular photomicrographs of 121 normal corneas. The predominant number of cells were hexagonal in shape (48% to 90%), with pentagonal (15% to 35%) and heptagonal (25% to 38%) cells constituting the greater portion of the remaining endothelium. Corneal endothelium with a greater number of hexagonal cells demonstrated minimal variation in cell size.

Endothelial cell density in relation to morphology.

Corneal endothelium of 121 normal corneas was studied with the clinical specular microscope, and the relationship between cell density, cell morphology, and age was examined. Our observations indicate a decrease in cell density with age in homomegethous endothelium but no such correlation in a polymegethous endothelium.