Clinical and basic science research is conducted by a cooperative effort between the Hektoen Institute of Medicine, the Division of Nephrology-Hypertension of the Stroger Hospital of Cook County, and the Renal Section of the University of Illinois. Experimental models are designed to better understand the mechanisms of disease and to improve methods of diagnosis, prevention, and treatment.
Diabetes is a systemic disease affecting blood vessels throughout the body, including those of the heart, eyes, kidney (diabetic nephropathy), and of healing wounds.
To understand the mechanism of diabetic nephropathy, Dr. Singh’s team studied collagenase activity, which is responsible for breakdown of extracellular matrix, and found it was depressed in cultured kidney cells exposed to high glucose levels (or to glycated albumin, a product formed in blood from exposure to high glucose). Depression of collagenase activity therefore could damage blood vessels by altering the matrix surrounding the blood vessels. The structure of blood vessels in the kidney is also maintained by paracrine factors secreted by the mesangial cell, which lie adjacent to the endothelial cells. These paracrine factors are vascular endothelial growth factor (VEGF), angiopoietin-1, and angiopoietin-2. They found that high glucose decreased VEGF, and increased angiopoietin-2, thus disturbing the paracrine balance between endothelial and mesangial cells, and causing capillary damage. Dr Singh went on to confirm the important role of vascular factors on blood vessels by implanting perforated polythene tubes under the skin of rats and allowing scar tissue to form around these tubes. He was then able to study the composition of this surrounding tissue layer and to sample the fluid in the tubes. This fluid had very high VEGF levels. These level decreased after the injection of a VEGF receptor antagonist, at which time the blood vessels and the surrounding tissue became disorganized. It thus appears that pre-existing blood vessels are sensitive to ambient VEGF levels, and that a low VEGF resulting from VEGF antagonism can damage blood vessels.
We studied the effect of a high glucose (HG) environment on the vascular factors that are secreted by mesangial cells and regulate endothelial growth and mesangial matrix deposition. To this effect, we measured the vascular factors in the glomeruli of streptozotocin-induced diabetic kidneys and in mesangial cells exposed to a HG concentration. We then transferred the media of mesangial cells previously exposed to high glucose to cultured endothelial cells to study the effects on endothelial growth, matrix formation, and in vitro capillary proliferation. In one week glomerular vascular endothelial growth factor (VEGF) was inhibited by 38% and angiopoietin-1 by 57%, but angiopoietin-2 was increased by 318%. We found similar results in mesangial cells exposed to high glucose. There was a decrease of VEGF (50% by enzyme immunoassay, 27% by mRNA), decrease of angiopoietin-1 (65% by mRNA), and a much greater increase of angiopoietin-2 (280% by immunoassay, 523% by mRNA). Compared to controls, the media of mesangial cells previously exposed to high glucose impaired endothelial cell growth by 61%, increased extracellular matrix by 100%, and decreased capillary formation by 90%. We conclude that high ambient glucose alters the secretion of vascular factors elaborated by mesangial cells, resulting in an expansion of the endothelial cell matrix and disruption of the capillary structure (ref. 1).
Vascular endothelial growth factor (VEGF) is a potent growth factor that is indispensable for the development of blood vessels in the fetus and in wound healing.
It is likely that VEGF continues to play a role in maintaining the blood vessels once they have been formed. However, it is not clear whether a low tissue VEGF caused either by disease or by systemic administration of VEGF antagonists can cause abnormalities in pre-existing blood vessels, especially of wound tissue that requires high local levels of VEGF for healing. Here we studied the effect of VEGF antagonism on blood vessels of a foreign-body granuloma (a model of wound healing tissue). Granulomas were induced by implanting perforated plastic tubes into the subcutaneous tissue of rats and allowed to develop for 14 days, when the implanted tubes were completely encapsulated by the subcutaneous tissue. The encapsulated granulomas consisted of three distinct histological layers, of which the middle layer was well perfused by a rich supply of microvessels. Morphologically, the granuloma remained ‘stable’ after developing for 14 days. At one week the VEGF levels in the granuloma fluid, which is in equilibrium with the interstitial fluid, were 25 times higher than in the plasma and continued to increase for up to three weeks, reflecting the high dependence of the wound tissue to ambient VEGF levels. On injecting the VEGF receptor antagonist in the fully formed granuloma, the pre-existing blood vessels in the middle layer regressed and underwent apoptosis, accompanied by disorganization and expansion of the extracellular matrix (collagen I predominantly) into areas normally devoid of matrix. We conclude that the wound tissue is sensitive to ambient VEGF levels, and that a low VEGF condition resulting from VEGF receptor antagonism can disrupt the healing of wound tissue (ref. 2).
Wound healing in diabetes is impaired largely because of poor new blood vessel formation (angiogenesis).
To better understand the nature of this defect we implanted polyvinyl tubes subcutaneously to form a granulation tissue for two weeks and studied the granulation tissue after inducing diabetes with streptozotocin. By one week of diabetes the granulation tissue was bloody and thinner than controls, its medial layer depleted of microvessels, and the surviving vessels appearing dehisced. These changes were prevented by giving insulin. In further studies we found that after 3 days of diabetes VEGF, which is concentrated in the inner layer of the control granulation tissue, was significantly reduced. Compared to controls, the diabetic tissue showed a greater degree of apoptosis in the microvessels. A strong chemotactic response for the attraction of bone marrow cells was suggested by an intense and equal staining for the chemokines SDF-1α and CXCR-4 in both the control and diabetic granulation tissues. As expected, CD-34 cells were seen in abundance in both control and diabetic granulation tissues. However, while the CD-34 positive cells were mostly integrated in the blood vessels of the control tissue, there were fewer such cells present in the blood vessels of the diabetic tissues, suggesting that they were defective in integrating with the new blood vessels. We thus conclude that diabetes causes a local reduction in VEGF levels, resulting in loss of wound blood vessels by two mechanisms, (a) cell death of newly formed vessels by apoptosis and (b) inhibition of new blood vessel formation because the CD-34 cells are unable to integrate into the vessel structure (ref. 3).
The presence of abnormal amounts of proteins (albumin, globulins, and other proteins) in the urine is a sign of early kidney disease. Clinical diagnosis is enhanced by a better understanding of the mechanisms of protein excretion in health and disease.
We studied the handling of albumin by the kidney, specifically the fate of the albumin filtered by the glomerulus. It is known that a large proportion of the filtered albumin is degraded by the renal tubules, but it has been unclear whether this albumin is degraded partially or completely and what happens to the degradation products. Hitherto it had been believed that all degraded albumin products were reabsorbed into the blood and reutilized. Studying this issue in cultured kidney cells, we found that 35% of the degraded products appeared in the urine and 65% towards the blood. Thus albumin seems to be incompletely degraded by the kidney, and some of the degradation products are excreted and lost in the urine. We extended these studies to experimental animals injected with radioactive albumin, and in a series of experiments found that normally the kidney degrades large amounts of albumin and that the degradation fragments appear in the urine. These findings are in sharp contrast with the established view that degraded albumin is completely reabsorbed into the blood stream. In view of recent reports of large amounts of albumin fragments present in normal urine we investigated the mechanism of albumin handling by the proximal tubule. We injected 125I-albumin intravenously in rats and measured the excretion of intact and degraded 125I-albumin in the urine. The excretion rate of intact 125I-albumin was compared to that obtained by routine radioimmunoassay. Also, human proximal tubular HK-2 cells were used to characterize the albumin receptor and study the degradation of albumin to peptides, establish their size by SDS-PAGE and gel filtration chromatography, and determine the direction in which the degradation products are removed from the cell. Following injection of 125I-albumin intravenously to rats, we recovered large quantities of 125I-albumin fragments in urine and determined that 98% was in a highly degraded form and only 2% was intact. Only the intact albumin could be detected by radioimmunoassay. We observed similar results in the urine of ex-vivo kidneys perfused with 125I-albumin. We found that 125I-albumin was taken up by HK-2 cells via a receptor, degraded in the lysosomes, and the peptides exocytosed to both the apical and basolateral sides of the cells. From these experiments we concluded that normally the kidney degrades large amounts of albumin and that the degradation fragments appear in the urine. These findings are in sharp contrast with the established view that degraded albumin is completely reabsorbed into the blood stream (ref. 4).
Considerable quantities of low molecular weight peptides or protein fragments are present in human urine.
This phenomenon has been long suspected but poorly studied, primarily because of technical difficulty of measuring peptides in the presence of proteins. We estimated these substances in normal subjects and in patients with renal disease by taking the difference between two methods widely used to measure urinary protein, the Lowry method (which detects both proteins and peptides) and the Bradford (Biorad) method (which detects only proteins). We validated this two-assay approach by showing that this difference indeed represented nitrogenous compounds and not inorganic contaminating substances, and found that normal persons excreted less than 200 mg of protein but 3-4 g of peptides / g creatinine, whereas patients with renal disease had a progressive decrease in peptide excretion, ranging from 3 to 0 g / g creatinine. Some nephrotic patients had almost no peptides in the urine, an intriguing finding that suggests renal tubular damage, requires further studies to explain the mechanism, and may offer future diagnostic possibilities. (ref. 5).
Lipoprotein nephropathy is a genetic disease in which deposits of lipoproteins accumulate in the glomerular blood vessels. The deposits gradually block the glomerular capillaries, resulting in kidney failure.
It is not clear if there is an abnormality of the patient’s lipoprotein that makes it adhere to the glomerular blood vessels. To answer this question Dr. Sam investigated the lipoprotein of one such patient to see whether there is a genetic mutation in the molecule. He found that the patient’s lipoprotein had an unusual mutation in which one aminoacid was in the wrong position in the protein part of the lipoprotein. This mutation was responsible for making the lipoprotein bind abnormally to the blood vessels in the kidney.
We studied a case of lipoprotein nephropathy, the second ever reported from the United States, in a Mexican man who had a hitherto undescribed mutation in the lipoprotein E gene (substitution of proline for arginine at position 147, Arg147Pro). In this patient (who had a similarly affected sibling), the glomerular basement membranes displayed double contours and circumferential mesangial extensions, suggesting that the deposition of lipids could be injurious to the endothelial cells. Immunofluorescence staining of the thrombi was positive for apolipoprotein E and B. To study the reason for the lipid deposition in the glomeruli, we incubated normal human kidney sections with the serum from the patient and a normal control. Apolipoprotein E from the patient’s serum exhibited binding to the glomerular capillary wall but the control did not, suggesting enhanced binding of the mutated apolipoprotein E to glomerular capillaries. Apolipoprotein E genotyping by restriction endonuclease digestion of PCR amplified genomic DNA showed it to be of the wild type E3/E3 (ref. 6).
Once a kidney sustains an injury it continues to undergo scarring and further damage even though the initial agent that had caused the injury may no longer be operative.
As scarring involves occurs the transformation of glomerular and tubular kidney cells into fibroblasts, Dr Sam studied the details of this process and induced this transformation after incubating kidney tubular and glomerular cells with transforming growth factor-β.
We carried out studies to determine if epithelial – mesenchymal transformation (EMT), well described in renal tubular epithelial cells, also occurs in glomerular epithelial cells, and whether it is reversible. We incubated cultured glomerular epithelial cells with TGF-β and studied their transformation into myofibroblasts. At four days the cells altered their phenotype, as shown by a change in shape, an increase in intracellular staining for α-smooth muscle actin, and decreased membrane staining for cytokeratin. Changing the medium after four days by excluding TGF-β and adding fetal bovine serum (as a source of EGF and other growth factors) caused the cells to revert to their original epithelial phenotype. By contrast, when the medium was changed in the same manner after eight days of exposure to TGF-β, the cells did not revert but remained myofibroblastic. Staining the cells for expression of epidermal growth factor (EGF) receptor before and after exposure to TGF-β caused this receptor, originally present on the plasma membrane, to become partly intracellular after 4 days of TGF-β exposure and completely intracellular after 8 days of TGF-β exposure. We further stained kidney sections obtained from two models of renal mass reduction and observed in the glomeruli a loss of the epithelial marker (podocalyxin) staining and the acquisition of α-smooth muscle actin staining. We conclude that EMT takes place in glomerular epithelial cells in vivo and in vitro. In cultured glomerular epithelial cells the process can be reversed by early but not by late intervention. It appears that TGF-β exposure progressively down regulates the EGF receptor on the membrane, rendering the cell refractory to EGF signals critical for maintaining the epithelial phenotype (ref. 7).
The omentum is an unusual organ that upon injury to internal organs rapidly expands along with new blood vessels to fuse with the injured site and bring about healing.
To understand the healing power of the omentum Dr. Singh and his team found that the omentum, when activated by injury, becomes a reservoir of stem cells and many growth factors which it supplies to the injured site to bring about healing. To cure diabetes they took advantage of the stem cells and growth factors present in the omentum to regenerate the lost insulin making cells (b-cells) by placing pieces of diabetic pancreas in the artificially activated omentum. Remarkably, new insulin positive cells were seen in the omentum and half of the diabetic animals became normal.
To study the mechanism by which an omental pedicle prevents infections and promotes healing when applied to an injured site, we injected a foreign-body into the abdominal cavity to activate the omentum. One week after the injection we isolated the omentum and measured blood vessel density, growth and angiogenesis factors (VEGF and others), chemotactic factors (SDF-1α), and progenitor cells (CXCR-4, WT-1). We found that the native omentum, which consists mostly of adipose tissue, expanded its non-adipose part (milky spots) mass 15-20 fold. VEGF and other growth factors increased 2-4 fold, blood vessel density 3-fold, and blood content 2-fold. The activated omentum also showed increases in SDF-1α, CXCR-4, and WT-1 cells — factors and cells positively associated with tissue regeneration. Thus we propose that an omentum activated by a foreign-body (or by injury) greatly expands its milky spot tissue and becomes rich in growth factors and progenitor cells that facilitate the healing and regeneration of injured tissue (refs. 8, 9).
In order to determine if pancreatic progenitor cells could be induced to form β-cells, we transplanted fragments of streptozotocin-induced diabetic pancreas into omentum pre-injected with a foreign material (activated omentum- a tissue rich in growth factors and progenitor cells). After transplantation there was immunocytochemical evidence that new insulin secreting β-cells appeared in the omentum. Extracts of these omenta contained measurable insulin in amounts consistent with the number of β-cells seen by immunocytochemical staining. The diabetic animals containing higher amounts of insulin in the omentum became normoglycemic; those with lesser amounts remained hyperglycemic. No insulin was produced when diabetic pancreas fragments were transplanted into the native (unactivated) omentum. Thus our studies show that new β-cells can be regenerated from the diabetic pancreas by introducing pancreatic fragments into an environment of high growth factors such as the activated omentum (ref. 10).
HIV nephropathy is the most common renal manifestation of AIDS and an important cause of kidney failure in infected individuals.
At the CORE Center Dr Mark Kraus is conducting a prospective cohort study of 60 patients aimed at analyzing the clinical importance of asymptomatic albuminuria in HIV+ patients without renal disease. Preliminary data indicates a greater prevalence of albuminuria than previously appreciated in this population. Other investigators have reported an incidence of albuminuria in HIV+ patients as high as 30%. Our ongoing investigation has revealed that with serial assays for albuminuria over one year, the incidence of albuminuria is over 60%. This may indicate ongoing HIV turnover within the kidney despite systemic inactivity of HIV. Dr. Kraus has also been involved in a collaborative study with the Division of Infectious Diseases. The CORE 50 cohort study is evaluating HIV+ patients over the age of 50 for both cardiovascular risk factors and cardiovascular disease. Early data suggests higher levels of proteinuria in an older group of HIV+ patients.
ACE inhibitors and angiotensin receptor blockers are widely used to treat hypertension, reduce proteinuria, and prevent the progression of renal disease
Dr Peter Hart is the principal investigator for the ONTARGET-TRANSCEND trial, an international study of over 26.000 patients in 40 countries including the USA. The study seeks to determine whether combined treatment with ACE inhibitors and angiotensin receptor blockers is more cardioprotective that each drug given alone in high risk cardiovascular patients. This grant funded research is in its fourth year and will begin closing out in 2007.
Research projects over past years have addressed various aspects of the natural history of renal disease, emphasizing prevention of progression to renal insufficiency and improving the treatment of patients maintained by dialysis. Some of these projects have immediate application in the clinical management of patients, others are concerned with understanding the more fundamental aspects of disease. Subjects studied in the past have included the retention products of middle molecular size substances in uremic patients; models of renal failure by injecting xanthopterin crystals; effects of acute and chronic glutathione depletion on renal function; and delineation of the mechanisms of cocaine induced hypertension and renal damage.
A single injection of the toxic substance bromethylamine caused extensive papillary necrosis and sloughing. This initial insult to the kidney is followed over the ensuing weeks by the development of progressive interstitial fibrosis. After one year the rats develop the counterpart of human focal segmental glomerulosclerosis and nephrotic proteinuria. Administering the angiotensin converting enzyme inhibitor enalapril in this model slowed the deterioration of renal function, as did the naturally occurring hormone relaxin. Further studies indicated that collagenase tissue levels were decreased in renal tissue but that transforming growth factor b was increased, suggesting that such changes also take place when renal disease in man progresses to uremia.
Blood lipids have been implicated in causing renal damage. We studied the mechanism by which lipid bearing proteins (lipoproteins) entering the urine from blood are processed and modified in the renal tubule in normal experimental animals and in those made diseased by the induction of nephritis. Although low density lipoprotein (LDL) in kidney tissue was present mostly in the intact form, it was 95% degraded in the urine and present in the form of fragments in the range of 5-15 kD. As this degradation was inhibited by chloroquine, it appears that renal degradation of LDL takes place in the lysosomes. Such degradation is incomplete, and the incompletely degraded fragments released into the urine may well be toxic to the kidney by virtue of their lipid side-chains.
In the course of studies of the mechanism of progression of diabetic nephropathy, we found that collagenase activity was depressed in cultured glomerular epithelial cells exposed to albumin glycated by exposure to high ambient glucose levels. Continuing this work has been a project studying the mechanism by which the fine blood capillaries of the kidney are damaged in diabetes patients, focusing on potential damage to mesangial and endothelial cells. We accordingly designed experiments to study the endothelial - mesangial interaction leading to the production of excessive collagen and other hormone-like messenger molecules that may exacerbate kidney fibrosis. High glucose decreased vascular endothelial growth factor (50% by enzyme immunoassay, 27% by mRNA), increased angiopoietin-1 (65% by mRNA) and increased angiopoietin-2 (280% by immunoassay, 523% by mRNA). The media of mesangial cells previously exposed to high glucose impaired endothelial capillary networks and decreased cadherin expression. Thus high ambient glucose medium seems to disturb the paracrine interaction between endothelial and mesangial cells, causing capillary damage.
It is known that a large proportion of the filtered albumin is degraded by the renal tubules. Is this albumin degraded to small peptides or is it completely broken down to aminoacids? After degradation, are the products excreted to the blood or discarded in the urine? Hitherto, it has been believed that all degraded albumin products are reabsorbed into the blood and reutilized. Studying this issue in cultured kidney cells, we observed a time-dependent increase of degraded albumin in the form of small peptides outside the cells. In the cells intact albumin and larger fragments were present. We then found that 35%% of the degraded products appeared in the urine and 65% towards the blood. Thus albumin seems to be incompletely degraded by the kidney and some of the degradation products are excreted and lost in the urine.
Several researchers have detected in the blood of nephrotic patients a factor that appears to be responsible for causing the protein to leakage out from the blood into urine. The methodology for detecting such a factor to detect is complex and cannot be used in the routine management of nephrotic patients. We described a simpler method using kidney cells cultured on filter membranes, allowing the sampling of the culture medium from both the blood and urine side of the cells. When a permeability factor is present, the radio-albumin added to the blood side in presence of the patient blood will cause the albumin to leak to the urine side This method may allow the nephrologist to assess the severity of the kidney disease and the response to therapy.
These include participation in the prospective NIH funded African American Study of Kidney Disease & Hypertension (AASK) on the effect of different levels of blood pressure reduction on the progression of renal disease. We studied the clinical effects of cocaine, its prevalence in renal diseases, and the mechanisms of causing hypertension and progression to renal failure. We also studied the prevention of parathyroidectomy in dialysis patients with large doses of vitamin D analogues; various diagnostic and therapeutic aspects of lupus nephritis; and the response to corticosteroids of African American patients with focal glomerulosclerosis. Another study addressed the issue whether carbamylated hemoglobin, formed when urea breaks down in-vivo to cyanate, may (by analogy to glycosylated hemoglobin) serve as a marker of the adequacy of dialysis and as an index of compliance with treatment. In order to screen patients for early renal disease by measuring small quantities of protein present in the urine. We reemphasized the utility of tests that the physician can perform himself in the office, exposed some of the inaccuracies of methods currently used to measure total urinary protein, and evaluated a new method of measuring urinary albumin by fluorometry. We found that methods of determining total urinary protein as used in routine clinical laboratories were not accurate enough to detect the small amounts of albumin (so-called microalbuminuria) that mark early renal involvement by diabetes. We also compared albumin excretion measured by radioimmunoassay with urinalysis, and found that a negative sulfosalicylic acid precipitation and "Albustix" test ruled out microalbuminuria with a 98% predictive value. A "trace" finding, however, means microalbuminuria in over 80% of patients, thus reemphasizing the utility of simple tests that can be carried out in the doctor's office.