Ischemia/reperfusion of the liver induces heart injury in rats.

OBJECTIVE: We evaluated the cardiovascular injury induced by ischemia and reperfusion (I/R) of the liver by measuring changes in blood levels of cardiac troponin I (cTNI), an index of cardiovascular injury, as well as levels of selected indicators of an inflammatory response. MATERIALS AND METHODS: Ischemia was induced in the rat liver by clamping the common hepatic artery and portal vein for 40 minutes, after which flow was restored, and the liver reperfused for 90 minutes. Blood samples were collected prior to ischemia and after reperfusion. cTNI as well as levels of tumor necrosis factor alpha (TNFalpha), hydroxyl radical (.OH), nitric oxide (NO), and alanine transferase (ALT) were measured. RESULTS: I/R of the liver induced a significant increase in ALT (P<.001). Increased cTNI levels (P<.05) were associated with inflammatory responses, such as elevated levels of TNFalpha (P<.001), . OH (P<.001), and NO (P<.001). After administration of 3-aminobenzamide, a poly(ADP-ribose) polymerase (PARP) inhibitor, liver and heart injuries were significantly attenuated (P<.05). CONCLUSIONS: I/R-induced liver injury was associated with cardiovascular injury, perhaps resulting from inflammatory responses triggered by elevated levels of reactive radical species of nitric oxide, superoxide, and peroxynitrite, by which PARP was activated. 3-Aminobenzamide, significantly attenuated I/R-induced liver and heart injuries.

Ischemia/reperfusion-induced low reactivity of the rat superior mesenteric vascular bed is associated with expression of nitric oxide synthases.

UNLABELLED: Our objective was to investigate the mRNA and protein expressions of eNOS and iNOS in the mesenteric vascular bed after ischemia and reperfusion of the rat superior mesenteric artery (SMA) and the role of nitric oxide (NO) in the response of the vascular bed to vasoconstrictors following reperfusion of the SMA. METHODS: Real-time polymerase chain reaction and immunohistochemistry were used to monitor the mRNA and protein expression of eNOS and iNOS after I/R challenge to the rat SMA. Ischemia was induced by clamping the SMA for 40 minutes, after which the flow was restored and the vessels were reperfused for 300 minutes. Blood samples were collected for assays of lactic dehydrogenase, tumor necrosis factor (TNF), hydroxyl radical, and NO. After ischemia/reperfusion, the vascular beds were separated for analysis of the expression of eNOS and iNOS. The SMA with its associated intestinal tissue was isolated and perfused in vitro with Tyrode's solution (N = 8) then challenged with phenylephrine. RESULTS: Reperfusion of the SMA induced an increase in blood concentrations of lactic dehydrogenase (P < .001; N = 8), hydroxyl radical (P < .05), TNF (P < .001), and NO (P < .05). ENOS and iNOS mRNA expression increased 1.3 +/- 0.1-fold and 19.6 +/- 3.5-fold, respectively when compared to the sham-operated group. Protein expression increased 1.9 +/- 0.4-fold and 12.6 +/- 3.1-fold, respectively, after reperfusion (N = 3) when compared with sham-treated rats. In vitro challenge showed that administration of phenylephrine (10(-8) approximately 10(-4) nmol) produced vasoconstriction in a dose-related manner. Maximum contractile responses to phenylephrine were attenuated in reperfused SMA. Addition of the NOS inhibitor N(G)-nitro-L-arginine (L-NNA, 10(-4) M) resulted in full recovery of the response to phenylephrine. CONCLUSIONS: Ischemia/reperfusion of the SMA results in a decrease in vascular reactivity of the mesenteric vessels that is dependent on NOS expression by the intestinal vascular bed.

Lack of a protective effect of insulin on three reperfusion-liver injury models in rats and mice.

UNLABELLED: Our objective was to investigate the potential protective effects of insulin on the liver injury induced in three ischemia and reperfusion (I/R) models. METHODS: Three I/R models were used: (1) I/R of the liver was produced in isolated, perfused rat livers; (2) in in situ I/R of the liver in rats, ischemia was induced by clamping off the hepatic artery and portal vein for 40 minutes, the flow then restored, and the liver reperfused for 90 minutes; (3) in in situ I/R of the liver in mice, ischemia was induced by clamping off the hepatic artery for 15 minutes, the flow then restored, and the liver reperfused for 45 minutes. In all three cases, blood samples collected before ischemia and after reperfusion were analyzed for sGOT. Plasma nitrate/nitrite, hydroxyl radicals, and tumor necrosis factor were also measured. In each model, a dose of insulin sufficient to induce euglycemia was administered to assess its protective effect on liver injury and inflammation. RESULTS: These I/R protocols resulted in a significant increase in sGOT and in three inflammatory parameters; nitric oxide, hydroxyl radicals, and tumor necrosis factor. Pretreatment with insulin did not attenuate the liver injury in any of the three I/R models. CONCLUSIONS: Although insulin has been reported to provide anti-inflammatory benefits by reducing oxidative and nitrosative stress and cytokine release, none of these protective effects was seen in the three I/R-induced liver injury models we tested.

Oxygen radicals and matrix metalloproteinases mediate reperfusion liver injury.

Many pathological processes involve the breakdown and remodeling of the extracellular matrix, which is mediated by the family of important enzymes known as matrix metalloproteinases (MMPs). One such process is warm ischemia/reperfusion (I/R) injury, the most important cause of dysfunction of liver allografts. We monitored protein expression of MMP-9 by Western blotting in rat liver after I/R. We also monitored changes in total MMP activity in the serum before and after I/R. Ischemia was induced by clamping the common hepatic artery and portal vein for 40 minutes and reperfusing for 90 minutes. Blood samples collected before ischemia and after reperfusion were analyzed for AST, hydroxyl radical, and tumor necrosis factor (TNFalpha). This protocol resulted in a high level of MMP-9 expression in liver tissue. Total MMP activity in serum was also significantly increased. Levels of AST, hydroxyl radicals, and TNF alpha were concomitantly increased. Ilomastat, an MMP inhibitor, attenuated the I/R-induced liver injury. After administration of the oxygen radical scavenger N-acetylcysteine (NAC), total MMP activity was suppressed, and liver injury was again attenuated. These results indicated that reperfusion liver injury induced an increase in MMP-9 protein expression and in serum MMP activity. The protective effects of an MMP inhibitor and NAC indicate that oxygen radical production is involved in MMP expression and liver injury associated with I/R.

Abnormality of the DNA double-strand-break checkpoint/repair genes, ATM, BRCA1 and TP53, in breast cancer is related to tumour grade.

The role of the DNA double-strand-break (DSB) checkpoint/repair genes, ATM, BRCA1 and TP53, in sporadic breast cancer requires clarification, since ATM and BRCA1 mutations are rare in sporadic tumours. In an attempt to explain this phenomenon, we postulated that (i) in addition to genetic deletion, abnormal expression of DSB checkpoint/repair proteins might abolish the function of these genes and (ii) there might be a combined effect of individual defective genes during breast cancer pathogenesis. Using a largely homogenous group of 74 specimens of early-onset (< or =35 years of age) infiltrating ductal carcinomas, we examined associations between pathological grade and genetic deletion and/or abnormal protein expression of ATM, BRCA1 and TP53. The results showed that high-grade tumours displayed a high frequency of loss of heterozygosity (LOH) at, and/or abnormal expression of, ATM, BRCA1 and TP53. Multigenetic analysis showed abnormalities in BRCA1 to be independently associated with high-grade tumours. ATM and TP53 appeared to play an assistant role, abnormalities in these genes significantly increasing the possibility of poor differentiation in tumours with abnormalities in BRCA1. Furthermore, a higher number of abnormalities (LOH or abnormal expression) in these three genes correlated with poor tumour differentiation. Thus, this study suggests that combined changes in several DSB checkpoint/repair genes belonging to a common functional pathway are associated with breast cancer pathogenesis.

Enhanced malignant progression of nasopharyngeal carcinoma cells mediated by the expression of Epstein-Barr nuclear antigen 1 in vivo.

Nasopharyngeal carcinoma (NPC) is consistently associated with Epstein-Barr virus (EBV) and mostly classified as poorly differentiated squamous cell carcinoma or undifferentiated carcinoma with early metastasis and a rapidly progressive clinical course. The EBV-encoded latent proteins, Epstein-Barr nuclear antigen 1 (EBNA 1) and latent membrane proteins (LMPs), may be expressed in NPC, but their biological effects are poorly understood. EBNA 1 may predispose B lymphocytes to lymphomagenesis in transgenic mice, but its biological effects in NPC are still unknown. This study investigated the biological effects of EBNA 1 by expressing it in an EBV-negative NPC cell line (HONE-1), which was then inoculated into both nude and severe combined immunodeficiency mice. The EBNA 1 caused HONE-1 cells to grow in a less differentiated pattern and to progress more rapidly, as well as increasing their tumourigenicity and metastatic capability. These data suggest that EBNA 1 may play a critical role in the progressive evolution of NPC.

Epizootic of low-virulence hepatotropic murine hepatitis virus in a nude mice breeding colony in Taiwan.

A natural outbreak of mouse hepatitis virus infection developed in a breeding colony of nude mice in Taiwan. The outbreak was unique in that morbidity was high in both adult and suckling mice, but only sucklings died. In contrast, all suckling heterozygous (nu/+) mice survived, and no lesions were found in adult female heterozygous (nu/+) mice. Adult male nude mice had chronic, active, necrotizing hepatitis with syncytial giant cells, but no lesions were detected in other tissues. Immunohistochemistry with anti-A59 and anti-JHM serum revealed mouse hepatitis virus antigen in the liver of infected adult and suckling nude mice, although less intensively in the kidney of adult nude mice. Suckling BALB/c mice inoculated with filtrates of the liver of adult nude mice developed hepatitis similar to that in the naturally infected nude mice. Virus was isolated by inoculating cell-free liver filtrate from infected adult nude mice onto 3T3 cells. Electron microscopy of purified virus revealed 100-nm-diameter enveloped particles with characteristic petal-shaped surface projections. We conclude that the outbreak was caused by a weakly virulent, highly hepatotropic murine hepatitis virus.

Assessment of p53 expression in nasopharyngeal carcinoma.

We analyzed the expression of the p53 protein by immunohistochemical methods from 101 patients with nasopharyngeal carcinoma (NPC): 24 with NPC and dysplastic lesions adjacent to carcinoma and 14 with primary and metastatic specimens. Ninety-six of 101 lesions (95%) had detectable p53 protein in the nuclei of tumor cells, indicating that overexpression of the p53 protein might be closely associated with NPC. Among 24 patients who had NPC and dysplastic lesions adjacent to carcinoma, 19 of the dysplastic lesions (79.2%) and 22 of the carcinomas (91.7%) showed positive staining for the p53 protein. In dysplastic epithelia p53 antigenicity was generally in a basal location. The significant association of p53 expression in NPC and dysplastic lesions adjacent to carcinoma (P < .0001, Fisher's exact probability test) suggests that p53 overexpression seems to occur at an early stage in the development of NPC. p53 expression in NPC does not correlate with histological grading, degree of lymphocytic infiltration between tumor cells, clinical stage, sex, or age (P > .05, chi-squared test). A comparison of p53 expression between primary and metastatic NPC was performed in 14 lesions. Although the p53 protein was consistently expressed in primary and metastatic tumor cells, there was no significant difference in p53 expression in both distinct but related lesions (P > .05, paired t-test). Our results suggest that the association of overexpression of the p53 protein in NPC may not be indicative of a mutant type p53 protein.

Xenotransplantation of microencapsulated canine islets into diabetic rats.

Islets of Langerhans were isolated in high yields from canine pancreata. In the procedure, the pancreata were perfused and digested with collagenase, and the islets were then purified on histopaque density gradients. As many as 60,000 islets were isolated from a single pancreas. Islets were encapsulated in alginate-polylysine-alginate membranes with the aid of an air-jet droplet generator. In vitro studies demonstrated that the isolated and encapsulated islets secreted insulin in response to glucose and IBMX challenge for at least 9 weeks. In in vivo studies 6 diabetic Wistar rats were transplanted with 5,000 to 8,000 encapsulated islets each. The diabetic condition was reversed in all recipients for up to 112 days. In control animals, which received free, unencapsulated islets, the xenografts remained functional for fewer than 21 days. Microcapsules retrieved from normoglycemic transplant recipients 1 and 2 months posttransplantation were shown to contain viable islet tissue, and no cellular overgrowth was observed on capsular surfaces. The results of the study indicate a considerable clinical potential of microencapsulated canine islet xenografts.

Mass isolation of highly purified canine islets using an automated method and Histopaque Ficoll gradients.

Eighteen pancreata from adult mongrel dogs were used for the study of islet isolation. The pancreas was distended with collagenase in Hanks' solution. The automated screen method and Histopaque Ficoll gradients were used to isolate and purify the canine islets. In vitro, the viability of isolated islets was assessed by both histology and perifusion studies. In vivo, the islet function was evaluated by using a nude mice xenograft model. Fair to good isolation and purification was found in 12 experiments. Before and after purification, the isolated islet count was 4767.1 +/- 560.1 and 3637.7 +/- 333.4 islet equivalence (I.E.)/gm pancreatic tissue. The purity was above 90%. Aldehyde Fuchsin stain disclosed islets with copious beta granules. The stimulation index of islets responding to 16.7 mM glucose plus 1 mM 3-isobutyl-1-methylxanthine (IBMX) versus 1.67 mM glucose was 12.93 +/- 4.75. Normoglycemia was restored and maintained for up to 2 weeks in 7 of 10 and up to 3 weeks in 5 of 10 diabetic nude mice transplanted with canine islets. In conclusion, the automated screen method and Histopaque Ficoll gradients afford a good yield of highly purified canine islets, and functional viability was verified both in vitro and in vivo. This will be an ideal model for isolation of human islets.

A new method of tissue processing that causes no shrinkage or distortion.

BACKGROUND: The routine method of tissue processing or dehydration usually causes shrinkage and/or distortion. More noticeable changes are seen in large specimens rich on hydrated tissues, such as embryos. Most likely this is due to the step-wise concentration change of the solvents. EXPERIMENTAL DESIGN: A new method of gradual solvent or medium exchange for tissue processing was developed. By using a peristaltic pump, pure solvent or embedding medium was added slowly to the processing chamber and the overflowing processed solution was slowly drained off. A computer spreadsheet program was used to calculate the concentration changes. After adding four to five times the processing volume of the solvent, the concentration of the adding solvent in the processing chamber could reach 98.15 to 99.32%. After reaching a desired endpoint, pure solvent could be used to replace the processing solution. RESULTS: We compared microencapsulated pancreatic islets and liver tissue processed traditionally and with the new method. By using this design, a smooth linear and gradual change of the concentration of the processing solutions could be obtained. No shrinkage or distortion of tissues due to the dehydration artifact was noticed, and the tissue seemed to be ideally suited for accurate quantitative histologic measurements. CONCLUSIONS: A very simple but efficient method for gradual solvent and medium exchange had been designed. This could efficiently prevent the shrinkage and distortion commonly produced by traditional tissue processing methods and prepare the tissue for quantitative histologic and pathologic studies.

Microencapsulated pancreatic islets: a pathologic study.

Dog pancreatic islets isolated by an enzymatic digestion method were encapsulated in an alginate-poly L-lysine-alginate membrane. These microencapsulated pancreatic islets were cultured in vitro to study their ability of insulin secretion. Portions of these in vitro-cultured microencapsulated pancreatic islets were taken out for a viability dye exclusion study as well as for pathologic studies to correlate them with insulin secretion ability. We found that there was a strong correlation between them. Good insulin-secreting microcapsules showed well-preserved cell membranes and beta-cell granules. An in vitro culture for one to two days in RPMI-1640 made the islets more stable, the cellular surface became smoother and the beta-granules were in better shape. The microencapsulated pancreatic islets were also injected into the peritoneum of streptozotocin-induced diabetic CDF1 mice. Blood glucose levels dropped and stayed low for up to 60 days. But, when non-encapsulated dog pancreatic islets were used, the blood glucose levels remained low for only about 14 days. A small portion of the injected microcapsules were washed out at specific times for pathologic study. Up to 28 days after injection, only a few of the injected microcapsules showed pericapsular cellular infiltrate. However, after 56 days, most of the microcapsules showed dense pericapsular cellular infiltrate. Immunohistochemical analysis of these infiltrates showed that the majority of cells were fibroblasts and macrophages. Most of the cells located in the inner portion of the infiltrate were fibroblasts, while the macrophages were located mainly on the outer portion. Both scanning and transmission electron microscopy showed that the surface of the microcapsule outer wall was much smoother than the inner wall. The size of the microcapsules was approximately 0.6-0.8 mm and the thickness of the wall measured around 10 nm. The smaller the microcapsule is, the less chance there is of rupture with release of the xenographic islets. Once the wall of the transplanted microcapsules was ruptured, the inner surface showed more increased inflammatory cell and fibroblast infiltration than the outer surface.

Extracellular matrix alteration in chronic hepatitis.

We used immunofluorescence and immunohistochemical PAP methods on 22 paraffin-embedded liver tissue specimens for observation and analysis of the distribution of extracellular matrix (ECM) elements in chronic hepatitis and cirrhosis. Our study revealed that in CLH there was only mild increases of types III, V collagen and fibronectin in spotty necrosis areas. In CPH, types III, V collagen and fibronectin revealed mild to moderate increase in portal area and lobular sinusoid. In CAH, moderate to marked increases of types III, V collagen and fibronectin and mild increase of type IV collagen in portal area, sinusoid lining, piecemeal necrosis and fibroseptum were found. Types I, IV collagen in fibroseptum were also noted. Some periportal hepatocytes showed abundant intracellular fibronectin. In cirrhosis, cases expressed similar finding to CAH except much more type IV collagen deposition. In addition, the basement membrane components including type IV collagen and laminin were found in the "capillarization" of periportal sinusoids in fibrotic liver tissue. In areas of piecemeal necrosis, the hepatocytes, single or assembled in "rosettes", were underlined by linear deposits of laminin and type IV collagen. Our study revealed that, histologically, the ECMs distribution of CAH is similar to that of cirrhosis but could be clearly distinguished from CPH and CLH. The prominent changes of ECMs, especially the basement membrane components, in case of CAH and cirrhosis are consistent with the fact that ECM may play a central role in liver function impairment and fibrogenesis.

Detection of acute myocardial infarction by evaluation of ultrasonic gray levels in dogs.

The study tested whether an experimental myocardial infarction can be detected from two-dimensional echocardiograms (2DE) by analysis of regional gray levels. The mid-left anterior descending coronary artery was ligated for 3 hours in 14 dogs (group 1) and for 5 hours in 6 dogs (group 2). 2DE were performed before, and after 3 and 5 hours of coronary artery ligation. The ultrasonic amplitude of the control and infarcted regions were obtained from digitized 2DE in the short axis view, at the mid-papillary muscle level. The mean gray levels (+/- SD) of the control and infarcted regions were compared during end-diastolic stop frames. After sacrifice, the hearts were cut into 1 cm thick slices and stained with 1% triphenyl tetrazolium chloride (TTC) solution. The myocardium was then studied by light microscopy. Computerized tomographic scans were also obtained in vitro from 3 hearts of both groups. There was no difference in mean gray levels of the control region during the experiments. However, in the region of wall motion abnormality (area of infarction), the mean gray levels increased from 49.9 +/- 3.5 (before ligation) to 62.0 +/- 7.4 (after 3 hours of ligation, p less than 0.005) in 10 group 1 dogs, but no differences were seen in mean gray levels (49.6 +/- 3.8 vs 50.4 +/- 4.0) in those without a myocardial infarction in 4 group 1 dogs; gray levels also increased from 50.4 +/- 2.9 (before ligation) to 58.6 +/- 6.1 (after 3 hours of ligation, p less than 0.05) and to 65.0 +/- 4.2 (after 5 hours of ligation, p less than 0.005) in group 2 dogs. The area of left ventricular asynergy corresponded precisely to the area of myocardial infarction, determined by both TTC staining and the computerized tomographic scans. The light microscopy of the infarcted area also demonstrated interstitial edema and polymorphonuclear cell infiltration.

Detection of human papillomavirus DNA in human cervical lesions by tissue in situ nucleic acid hybridization.

Cervical cancer is one of the most common female cancers in Taiwan. Certain types of human papillomavirus (HPV) are frequently detected in the epithelial precancerous and cancerous lesions of the cervix. By the use of tissue in situ hybridization, we investigated the relationship of various types of HPV (group I, HPV-6 & 11, group II, HPV-16 & 18, group III, HPV-31, 33 & 35) with cervical condyloma, carcinoma as well as precancerous lesions. Group I HPV DNAs were mainly found in cervical condylomatous lesions (2/2) of the cervix and cervical intraepithelial neoplasia I (CIN I) (2/4), but were only occasionally found in CIN II (1/4), CIN III (1/9) or non-keratinized squamous cell carcinoma (1/15). HPV DNAs of groups II and III were mainly detected in lesions of CIN III (5/9) and invasive squamous cell carcinoma (large cell, keratinized type: 4/7; large cell, non-keratinized type: 11/15). HPV DNA sequences were invariably detectable only in the cell nuclei of condyloma or dysplastic epithelium or invasive carcinoma. However, they could not only be detected in the upper layer dysplastic cells and koilocytes but also in the well and poorly differentiated cervical cancer cells. The distribution of HPV DNA positive cells in the carcinomas fell into four different patterns: (1) upper zone and non-invasive regions of the carcinoma (11/22, 50%), (2) basal zone and invasive regions (2/22, 9%), (3) randomly scattered (7/22, 32%), and (4) extensively distributed over the whole tumor lesions (2/22, 9%). Thus, our results are consistent with a strong correlation between the presence of HPV-16, 18, 31, 33 and 35 and malignant conversion of cervical epithelial cells.

Microencapsulation [corrected] of tumor cells and assay for selecting anticancer drugs.

A microencapsulation of living tumor cells by an improved membrane and droplet forming technique was established in our laboratory. This semipermeable microencapsulating membrane was impermeable to serum albumins (M.W. 66,000 or 45,000) and human hemoglobin (M.W. 64,000), but permitted passage of low molecular weight substances (alpha-Lactalbumin, or Trypsinogen; M.W. 14,200 or 24,000). The in vivo results showed that microencapsulated tumor cell lines (KB, human oral epidermoid cell; P-388 lymphocytic leukemia; GBM 8401/TSGH, glioma) and human colorectal carcinoma cells grew and proliferated exponentially within twenty days. The in vivo growth exhibited better than that in vitro. Histological and morphological findings of these four different kinds of tumor cells are similar to those of original tumor cells. Treatment of the microencapsulated tumor cells (MTC) with cytotoxic drugs (adriamycin, 5-fluorouracil and cyclophosphamide) in vitro showed no significant difference in percent inhibition (p greater than 0.05) between the encapsulated and non-encapsulated cells. The in vivo data indicated that different anti-cancer drugs had different inhibition effects. The results showed that the MTC model was useful for screening an appropriate cytotoxic drug and could be applied to clinical medicine in the near future.