IMP2/p62 induces genomic instability and an aggressive hepatocellular carcinoma phenotype.

Hepatocellular carcinoma (HCC) represents the third leading cause of cancer-related deaths and commonly develops in inflammatory environments. The IGF2 mRNA-binding protein IMP2-2/IGF2BP2-2/p62 was originally identified as an autoantigen in HCC. Aim of this study was to investigate a potential pathophysiological role of p62 in hepatocarcinogenesis. Human HCC tissue showed overexpression of IMP2, which strongly correlated with the fetal markers AFP and DLK1/Pref-1/FA-1 and was particularly elevated in tumors with stem-like features and hypervascularization. Molecular classification of IMP2-overexpressing tumors revealed an aggressive phenotype. Livers of mice overexpressing the IMP2 splice variant p62 highly expressed the stem cell marker DLK1 and secreted DLK1 into the blood. p62 was oncogenic: diethylnitrosamine (DEN)-treated p62 transgenic mice exhibited a higher tumor incidence and multiplicity than wild types. Tumors of transgenics showed a more aggressive and stem-like phenotype and displayed more oncogenic chromosomal aberrations determined with aCGH analysis. DEN-treated p62 transgenic mice exhibited distinct signs of inflammation, such as inflammatory cytokine expression and oxidative stress markers, that is, thiobarbituric acid-reactive substance (TBARS) levels. Reactive oxygen species (ROS) production was elevated in HepG2 cells, which either overexpressed p62 or were treated with DLK1. p62 induced this ROS production by a DLK1-dependent induction and activation of the small Rho-GTPase RAC1, activating NADPH oxidase and being overexpressed in human HCC. Our data indicate that p62/IMP2 promotes hepatocarcinogenesis by an amplification of inflammation.

Tartrate-resistant acid phosphatase isoform 5a as an inflammation marker in end-stage renal disease.

AIM: End-stage renal disease (ESRD) is often complicated by chronic inflammation and malnutrition. We tested whether serum tartrate-resistant acid phosphatase (TRACP) isoform 5a relates to other markers of inflammation in ESRD. MATERIAL: Predialysis serum was collected from 99 ESRD patients (51 male, 48 female) aged 55 +/- 15 years and a control group of 36 healthy subjects (8 male, 28 female) aged 43.2 +/- 10.5 years. METHODS: Serum TRACP 5a activity and protein, TRACP 5b activity and C-reactive protein (CRP) were estimated by in-house immunoassays. Commercial kits were used for serum bone-specific alkaline phosphatase, Ntelopeptides of Type I collagen, interleukin-6 (IL-6) and fetuin-A. Intact parathyroid hormone was determined by chemiluminescent assay. Albumin, cholesterol, triglycerides, ferritin and hemoglobin were compared to the hospital reference ranges. Bone mineral density (BMD) was measured at the heel in 69 patients and all control subjects and expressed as g/cm2 and age-corrected T-score. RESULTS: Mean (median) levels of all serum markers were significantly elevated in ESRD except fetuin-A, which was significantly reduced. Mean BMD (g/cm2) was not different than control, but mean T-score was significantly reduced. TRACP 5a protein correlated with CRP, triglycerides and ferritin, but not with IL-6 or any other nutritional or bone markers or BMD. TRACP 5b activity correlated with all bone markers and BMD, but not with inflammation or nutritional markers. CONCLUSION: Our findings suggest that TRACP 5a may be a useful marker to estimate the degree of inflammation in ESRD patients on chronic hemodialysis.

Regulation of expression of type II sodium-phosphate cotransporters by protein kinases A and C.

The purpose of the present study was to determine the effect of protein kinase A and protein kinase C activation on the membrane expression of NaPi-4, the type II sodium-phosphate cotransporter in OK cells. NaPi-4 expression was measured using polyclonal antisera produced in rabbits against a peptide identical to the carboxy-terminal 12-amino acid sequence of NaPi-4. The antisera identified an apically localized protein by confocal imaging of intact OK cells and a broad band of 110-140 kDa by immunoblot analysis of OK cell membranes. Treatment of OK cells with parathyroid hormone (PTH) decreased the intensity of the 110- to 140-kDa band, which was detectable by 2 h, maximal by 4 h at 62%, and sustained for 24 h. 8-Bromo-cAMP (8-BrcAMP) inhibited NaPi-4 expression for up to 24 h by over 90%. However, phorbol 12-myristate 13-acetate inhibited NaPi-4 expression by less than 10%. PTH-(3-34), a fragment which stimulates only protein kinase C, inhibited phosphate transport but also had no effect on NaPi-4 expression. We conclude that protein kinase A but not protein kinase C inhibits sodium-phosphate uptake in OK cells by downregulation of NaPi-4 expression.

Dopamine regulates phosphate uptake by opossum kidney cells through multiple counter-regulatory receptors.

The purpose of this study was to determine the mechanisms of dopamine regulation of phosphate uptake in opossum kidney (OK) cells, a model of proximal renal tubules. Dopamine stimulated cAMP generation and inhibited radiolabeled phosphate uptake into OK cell monolayers by 14.4 +/- 1.8%. The effect of dopamine was transient, as phosphate uptake returned toward control level by 3 h despite the continued presence of dopamine. Pretreatment with pertussis toxin increased dopamine inhibition of phosphate uptake to 25 +/- 3%, increased the duration of the dopamine effect to at least 3 h, and enhanced cAMP generation. In an OK cell clone that overexpressed cAMP phosphodiesterase, dopamine did not inhibit phosphate uptake, but pharmacologic inhibition of protein kinase A activation did not prevent dopamine inhibition of phosphate uptake. A DA1 receptor agonist inhibited phosphate uptake more potently than dopamine (29.5 +/- 1.1%) or a DA2 receptor agonist (7.9 +/- 2%). However, both DA1 and DA2 receptor antagonists completely blocked dopamine inhibition of phosphate uptake. DA1, but not the DA2, antagonists blocked dopamine-stimulated cAMP generation. Treatment with alpha-adrenergic receptor antagonists potentiated dopamine inhibition of phosphate uptake to the same extent as pertussis toxin and was not additive with pertussis toxin. It is concluded that dopamine inhibits phosphate uptake through DA1 and DA2 receptor stimulation by cAMP-dependent and -independent pathways and activates a pertussis toxin-sensitive counter-regulatory pathway that attenuates this response through alpha-adrenergic receptor stimulation.

Cellular mechanisms involved in the acute adaptation of OK cell Na/Pi-cotransport to high- or low-Pi medium.

Variations in dietary phosphate (Pi) intake in rats lead to alterations of renal Pi reabsorption. These effects are associated with corresponding changes in the abundance of the type II Na/Pi-cotransporter protein in proximal tubular brush-border membranes. In the present study we investigated the regulation of the type II Na/Pi-cotransporter in response to high- and low-Pi medium in opossum kidney (OK) cells, an epithelial cell-line of proximal tubular origin. We show that "acute" (4 h) and "chronic" (24 h) exposures of OK cells to high- or low-Pi medium lead to decreases or increases, respectively, in Na/Pi-cotransport activity which are paralleled by alterations in the total cellular amount of the corresponding type II Na/Pi-cotransporter protein (NaPi-4), but not by changes in the amount of the NaPi-4 mRNA. Also in OK cells transfected with the corresponding rat renal type II Na/Pi-cotransporter (NaPi-2) alterations in the Pi concentration in the medium lead to changes in the amount of NaPi-2 protein but not in the amount of NaPi-2 mRNA. Furthermore we show that lysosomal inhibitors prevent the degradation of the transporter, but do not interfere with its inhibition, in response to "acute" exposure of OK cells to high-Pi medium. Inhibition of lysosomal degradation also leads, in control conditions, to an accumulation of the transporter detectable on Western blot. It is concluded that the lysosomal proteolytic pathway is not only involved in the Pi-induced downregulation of the type II Na/Pi-cotransporter but also in its basic turnover.

Parathyroid hormone leads to the lysosomal degradation of the renal type II Na/Pi cotransporter.

We have studied the involvement of proteolytic pathways in the regulation of the Na/Pi cotransporter type II by parathyroid hormone (PTH) in opossum kidney cells. Inhibition of lysosomal degradation (by leupeptin, ammonium chloride, methylamine, chloroquine, L-methionine methyl ester) prevented the PTH-mediated degradation of the transporter, whereas inhibition of the proteasomal pathway (by lactacystin) did not. Moreover it was found (i) that whereas lysosomal inhibitors prevented the PTH-mediated degradation of the transporter they did not prevent the PTH-mediated inhibition of the Na/Pi cotransport and (ii) that treating opossum kidney cells with lysosomal inhibitors led to an increased expression of the transporter without any concomitant increase in the Na/Pi cotransport. Further analysis by subcellular fractionation and morphological techniques showed (i) that the Na/Pi cotransporter is constitutively transported to and degraded within late endosomes/lysosomes and (ii) that PTH leads to the increased degradation of the transporter in late endosomes/lysosomes.

Biotransformation of perchloroethene: dose-dependent excretion of trichloroacetic acid, dichloroacetic acid, and N-acetyl-S-(trichlorovinyl)-L-cysteine in rats and humans after inhalation.

Chronic exposure of rodents to perchloroethene (PER) increased the incidence of liver tumors in male mice and resulted in a small but significant increase in the incidence of renal tumors in male rats. The tumorigenicity of PER is mediated by metabolic activation reactions. PER is metabolized by cytochrome P450 and by conjugation with glutathione. Cytochrome P450 oxidation of PER results in trichloroacetyl chloride which reacts with water to trichloroacetic acid (TCA) which is excreted. The formation of S-(trichlorovinyl)glutathione (TCVG) from PER results in nephrotoxic metabolites. TCVG is cleaved to S-(trichlorovinyl)-L-cysteine (TCVC) and acetylated to N-acetyl-S-(trichlorovinyl)-L-cysteine (N-ac-TCVC), which is excreted with urine. TCVC is also cleaved in the kidney by cysteine conjugate beta-lyase to dichlorothioketene which may react with water to dichloroacetic acid (DCA) or with cellular macromolecules. The object of this study was to comparatively quantify the dose-dependent excretion of PER metabolites in urine of humans and rats after inhalation exposure. Three female and three male human volunteers and three female and three male rats were exposed to 10, 20, and 40 ppm PER for 6 h, and three female and three male rats to 400 ppm. A dose-dependent increase in the excretion of TCA and N-ac-TCVC after exposure to PER was found both in humans and in rats. A total of 20.4 +/- 7.77 mumol of TCA and 0.21 +/- 0.05 mumol of N-ac-TCVC were excreted in urine of human over 78 h after the start of exposure to 40 ppm PER; only traces of DCA were present. After identical exposure conditions, rats excreted 1.64 +/- 0.42 mumol of TCA, 0.006 +/- 0.002 mumol of N-ac-TCVC and 0.18 +/- 0.04 mumol of DCA. Excretion of N-ac-TCVC in male rats exposed to 400 ppm PER (103.7 nmol) was significantly higher, compared to female rats (31.5 nmol) exposed under identical conditions. N-ac-TCVC was rapidly eliminated with urine both in humans (t1/2 = 14.1 h) and in rats (t1/2 = 7.5 h). When comparing the urinary excretion of N-ac-TCVC, a potential marker for the formation of reactive intermediates in the kidney, humans received a significantly lower dose (3 nmol/kg at 40 ppm) compared to rats (23.0 nmol/kg) after identical exposure conditions. In addition, rats excreted large amounts of DCA which likely is a product of the beta-lyase-dependent metabolism of TCVC in the kidney. The obtained data suggest that glutathione conjugate formation and beta-lyase-dependent bioactivation of TCVC in PER metabolism is significantly higher in rats than in humans. Thus, using rat tumorigenicity data for human risk assessment of PER exposure may overestimate human tumor risks.

A calculator based program to optimize the simulation of breast irradiation.

The simulation of breast fields using an isocentric set-up technique can be a lengthy process involving the placement of the isocentre, the determination of the gantry angles, and the selection of the lung shields, which in our center is one of six standard blocks. We show that with a body contour taken through central axis, five measurements and a calculator program, it is possible to significantly decrease the amount of time required to simulate a breast patient. We have developed a program for an HP48GX handheld calculator to determine the gantry angles, the isocentre, the field width, the standard angled block, and the couch and collimator rotation. The calculations are based on measurements of the field length, the horizontal distance between midline and mid axillary line, and the vertical distances from the mid axillary line to the inferior and superior beam border and central axis at midline. We use spherical geometry to perform the calculations to reflect the true environment and do not make any assumptions about the average patient's shape. For the simulation process a jig was developed that is inserted into the tray holder of the simulator to show the optical and radiological shadow of the calculated shielding along the patient's midline for clinical assessment during simulation and on the simulation film. The jig also has a holder for an aluminum wedge to improve the image quality of the simulation film. We admit that the lung shield increases the dose to the contralateral breast because of increased scatter and transmission through the shield; however, the block decreases the volume of irradiated lung while keeping the beam edge along the midline of the patient. The technique has been in use for two years and has resulted in time savings of up to 30% per patient. It has proven to be an easy and accurate way of setting up isocentric treatments to the breast.

Azotemia, TNF alpha, and LPS prime the human neutrophil oxidative burst by distinct mechanisms.

The oxidative burst of neutrophils from azotemic patients (AzoPMNs) is primed for an enhanced response compared to neutrophils from normal subjects (NorPMNs). The mechanism for this priming is unknown, although TNF alpha does not further prime AzoPMNs. The present study examines the hypothesis that azotemia and TNF alpha prime neutrophils by the same mechanism. Formyl peptide receptor expression and degranulation were not primed in AzoPMNs, but were primed by both LPS and TNF alpha. LPS was also able to prime the AzoPMN oxidative burst. Guanine nucleotide exchange by multiple guanine nucleotide binding proteins, including heterotrimeric G-proteins and low molecular weight GTP-binding proteins (LMWGs), was increased in AzoPMNs, as demonstrated by GTP gamma S binding and azidoanilide GTP photoaffinity labeling. The plasma membrane density of G-protein alpha i2, alpha i3, and alpha s subunits and the density in the cytosol of the LMWG, Rap1A, was present in significantly greater amounts on plasma membranes from AzoPMNs. FMet-Leu-Phe-stimulated phospholipase D activity, but not basal activity, was significantly greater in AzoPMNs. Finally, incubation of NorPMNs in plasma from azotemic patients resulted in a significant increase in basal GTP gamma S binding. These results demonstrate that priming of AzoPMNs is restricted to oxidative burst activity and that it occurs by a mechanism distinct from that utilized by TNF alpha and LPS. While the exact mechanism remains unknown, it appears to involve a plasma factor and changes in LMWG expression or activity.

Anthropomorphic phantom measurements for the validation of a treatment planning system.

A series of eight irradiations of an anthropomorphic phantom has been performed as part of a programme of treatment planning system (TPS) validation. The "treatment' configurations used represent a realistic cross section of those carried out routinely in a radiotherapy centre. The dose distributions within the phantom were determined using 75 LiF TLD chips per irradiation and the measured dose distributions were compared with calculations performed on a TPS for three ranges of dose gradient. The accuracy of the technique per measurement point is estimated to be 3% in dose (low dose gradient) and 3 mm in position (high dose gradient). The measurement and data analysis techniques used permit evaluation of a TPS against recently recommended criteria of acceptability although it is noted that is is not possible to isolate TPS performance from factors depending on, for example, open beam data fitting and treatment set up accuracy.

P2 purinoceptor stimulation attenuates PTH inhibition of phosphate uptake by a G protein-dependent mechanism.

Purinergic P2 receptors are present on proximal renal tubules, but their function is unknown. Because P2 agonists antagonize vasopressin-stimulated water transport in the distal tubule by inhibiting activation of adenylyl cyclase, we postulated that P2 receptor activation blocks parathyroid hormone (PTH) inhibition of phosphate uptake in proximal tubule by preventing PTH-stimulated adenosine 3',5'-cyclic monophosphate (cAMP) generation. PTH inhibition of sodium-dependent phosphate uptake was attenuated by alpha,beta-methylene-ATP (AMP-CPP), a P2x receptor agonist, but not by 2-methyl-thio-ATP, a P2y receptor agonist, in a dose-dependent manner. AMP-CPP did not attenuate inhibition of phosphate uptake produced by direct activation of adenylyl cyclase with forskolin, by addition of the cAMP analogue 8-bromo-cAMP, or by inhibition of cAMP phosphodiesterase with RO-20-1724. Additionally, AMP-CPP had no effect on basal or PTH-stimulated cAMP production. As PTH also stimulates protein kinase C activation, the effect of AMP-CPP on inhibition of phosphate uptake stimulated by phorbol 12-myristate 13-acetate (PMA) was tested. AMP-CPP had no effect on PMA-induced inhibition of phosphate uptake. Pretreatment with pertussis toxin abolished the attenuating effect of AMP-CPP on PTH inhibition of sodium-dependent phosphate uptake. We conclude that activation of purinergic P2 receptors attenuates the inhibitory effect of PTH on sodium-dependent phosphate uptake by a G protein-dependent mechanism that is independent of cAMP generation protein kinase A activation, or protein kinase C activation.

Heat-inactivated Sendai virus can enter multiple MHC class I processing pathways and generate cytotoxic T lymphocyte responses in vivo.

We have earlier described an alternative MHC class I processing pathway for Sendai virus (SV) in H-2Kb-transfected T2 cells (T2Kb). These cells have deleted genes for transporters associated with Ag processing (TAP1/2) and proteasome subunits LMP2/7 but can still process SV for the presentation of an immunodominant nucleoprotein CTL epitope (nucleoprotein peptide 324-332, FAPGNYPAL, SV9), even in the presence of the fungal metabolite brefeldin A (BFA). Presently we have compared live and heat-inactivated SV to investigate whether infectious virus, including early events such as binding and fusion at the host cell membrane, is important for nonclassical MHC class I processing and immunogenicity. We have found that heated virus (56 degrees C, boiled or autoclaved) with no fusion and hemagglutinin-neuraminidase activities, behaves similar to live SV in T2kb cells by entering a TAP-independent and BFA-resistant pathway. In EL-4 cells, which do not express this nonclassical TAP-independent and BFA-resistant pathway, heat-treated SV is processed in a BFA-sensitive way. In T1Kb- and TAP1/2-transfected T2Kb cells, as in T2Kb cells, processing of heat-inactivated SV was completely BFA resistant. Heat-inactivated SV was also found to prime CTLs in vivo. We conclude that heat-inactivated SV can enter both BFA-sensitive and -resistant MHC class I processing pathways and that SV in this respect may be particularly efficient. What property in the SV that is important for this characteristic is presently not clear but might be useful for the deliberate generation of CTL responses in vivo.

Effect of prenylcysteine analogues on chemoattractant receptor-mediated G protein activation.

The hypothesis that carboxylmethylation of gamma subunits plays a role in G protein activation was tested by examining the ability of N-acetyl-S-farnesyl-L-cysteine (AFC) and its methyl ester (AFC-ME) to inhibit G protein-mediated signalling in intact HL-60 granulocytes and isolated HL-60 plasma membranes. Incubation of HL-60 granulocytes with AFC or AFC-ME inhibited superoxide release stimulated by fMet-Leu-Phe, but not by opsonized bacteria. AFC-ME, but not AFC, inhibited NaF- and PMA-stimulated superoxide release. Addition of AFC to HL-60 membranes inhibited fMet-Leu-Phe-, leukotriene B4- (LTB4) and C5a-stimulated GTP gamma S binding and GTP hydrolysis more potently than it inhibited basal guanine nucleotide exchange. AFC-ME inhibited basal- and ligand-stimulated G protein activation with equal potency, but less potently than AFC. AFC also inhibited mastoparan-stimulated GTP gamma S binding. Binding of fMet-Leu-Phe and LTB4 to HL-60 membranes was completely inhibited by AFC, while AFC-ME inhibited ligand binding by less than 50%. Neither AFC nor AFC-ME inhibited pertussis toxin or cholera toxin-catalysed ADP-ribosylation of alpha i. It was concluded that AFC interrupts signal propagation in G protein-dependent pathways by multiple mechanisms, including inhibition of ligand-receptor interactions, of receptor-G protein coupling and of guanine nucleotide binding to G proteins. Carboxylmethylation alters the specificity of AFC interruption of signal propagation in intact cells and isolated membranes.

Role of carboxylmethylation in chemoattractant receptor-stimulated G protein activation and functional responses.

The role of G protein gamma subunit carboxylmethylation was examined in HL-60 granulocytes using an inhibitor of S-adenosylmethionine-dependent methylation, periodate-oxidized adenosine (Adox). A 40-60% reduction in gamma subunit carboxyl-methylation was associated with attenuation of fMet-Leu-Phe-stimulated GTP gamma S binding and GTP hydrolysis, while plasma membrane density of formyl peptide receptors, alpha i2, alpha i3, beta, gamma 5, and gamma 7 were not reduced. Reduced pertussis toxin-catalyzed ADP-ribosylation was re-established by in vitro methylation or addition of transducin beta gamma subunits. Superoxide release and inositol phosphate generation stimulated by fMet-Leu-Phe were significantly inhibited by Adox treatment. Carboxylmethylation contributes to transmembrane signalling and functional responses by enhancing association of alpha and beta gamma subunits.

Role of isoprenoid metabolism in chemotactic peptide receptor-mediated G protein activation.

The role of isoprenylation in formyl peptide receptor-mediated G protein activation was studied using plasma membranes isolated from normal HL-60 granulocytes and from cells in which isoprenylation was inhibited with mevastatin. Plasma membrane expression of formyl peptide receptors and G protein beta subunits, but not alpha i2 and alpha i3, was significantly reduced by inhibition of isoprenylation. This reduced expression resulted in impaired basal and fMet-Leu-Phe-stimulated G protein activation.

Unusual glomerular lesion in a patient receiving long-term interferon alpha.

A 60-year-old man with long-standing chronic myelogenous leukemia presented with renal insufficiency and proteinuria after more than 6 years of therapy with daily interferon alpha injections. He also manifested unusual skin lesions and a low-titer antinuclear antibody (ANA). Percutaneous renal biopsy disclosed an unusual glomerular lesion characterized by global, diffuse, and marked widening of the lamina rara interna, and focal segmental mesangial proliferation. Discontinuation of the drug resulted in resolution of the proteinuria, but not the renal insufficiency. These glomerular changes have not been reported previously as a complication of this form of malignancy and are similar to lesions reported in newborn rats and mice receiving interferon alpha. The potential role of interferon alpha in the development of this glomerular disease is discussed.

TMB-8 prevents the hydroosmotic response to ADH in rabbit cortical collecting tubules.

Both AVP and dDAVP effect a transient increase in cytosolic free calcium (iCa2+) in cortical collecting tubule (CCT) cells. To investigate the physiological role of this increase in iCa2+, we examined the effect of TMB-8, a putative inhibitor of iCa2+ release, on the initial and sustained phase of AVP- and dDAVP-stimulated water permeability (Pf) in isolated, perfused CCTs. Pretreatment of tubules with TMB-8, 50 microM, suppressed the increase in osmotic water permeability (Pf) induced by 10 microU/ml AVP and dDAVP, but had no effect on the sustained phase of the response. When increased to 100 microM. TMB-8 inhibited the sustained phase of AVP action. A similar pattern was observed on AVP-stimulated adenyly cyclase activity in rabbit renal membranes. Pretreatment of tubules with 50 microM TMB-8 attenuated the initial increase in Pf in response to cholera toxin but not to 8-Br-cAMP or forskolin. There was no effect of this concentration of TMB-8 on the sustained phase of these agonists. These studies suggest that, in lower concentrations, TMB-8 inhibits the mobilization of iCa2+, which is important for the interaction of Gs with the catalytic unit of adenylyl cyclase and the initial increase in AVP-stimulated Pf. In higher concentrations, TMB-8 inhibits adenylyl cyclase activity directly.

Electron microscopic study of an unusual posttransplant glomerular lesion.

Glomerular lesions are frequently seen in renal allografts and are usually classified into transplant glomerulopathy and posttransplant glomerulonephritis. The latter is subdivided into donor-related, de novo, and recurrent glomerulonephritis. We report a distinctive posttransplant glomerular lesion that does not fit into any diagnostic category mentioned above. This lesion was characterized by the presence of global, diffuse, subepithelial, electron-lucent deposits, in addition to the usual features of transplant glomerulopathy. This unusual usual lesion, to the best of our knowledge, has been reported only once and, indeed, was recognized in only one of 297 renal allograft biopsy specimens in our file. Although the classification, pathogenesis, and origin of this rare lesion remain to be elucidated, it can be associated with nephrotic syndrome, deterioration of renal function, and eventual graft loss.

Paired tumour infiltrating lymphocyte (TIL) and tumour cell line from bladder cancer: a new approach to study tumour immunology in vitro.

This paper reports the first example of tumour infiltrating lymphocytes (TILs) and a tumour cell line from the same individual and analyses their characteristics. The tumour cell line (CAT), derived from a patient with well-differentiated (G3pTa) TCC, has been in culture for 24 months and subcultured more than 100 times. Epithelial origin was established by electronmicroscopy and use of a range of monoclonal antibodies (Mabs) against cytokeratins. The TILs isolated from the same tumour expressed all the phenotypic characteristics of normal activated T cells and demonstrated low levels of cytotoxicity against the autologous tumour line (CAT). Comparison of cell surface molecules of these cells revealed the loss of HLA-B7, B44 and Bw6 from the CAT cells whilst maintaining HLA-A2, A3 and Bw4. Karyotypic analysis demonstrated three rearranged chromosomes (between chromosomes 4 and 11, 10 and 13, 11 and 17) on CAT cells. The potential that study of paired autologous tumour cells and TILs in culture offers for studying the role of MHC antigens in tumour rejection and the impact of different approaches to correcting the defect are reviewed.