Glycaemia and correlates of patient-reported outcomes in ACCORD trial participants.

AIMS: Post-hoc evaluation of relationships between first-year change in glycaemic control (HbA(1c) ) and change in patient-reported outcomes among ACCORD health-related quality of life (HRQoL) substudy participants. METHODS: Data from 2053 glycaemia-trial subjects were analysed. We assessed physical and mental health status (36-Item Short Form Health Survey, Version-2), symptom count and severity (Diabetes Symptoms Distress Checklist) and treatment satisfaction (Diabetes Treatment Satisfaction Questionnaire). Linear mixed models were used to test relationships between 1-year changes in HbA(1c) and patient reported outcomes sequentially adjusting for correlates (baseline characteristics, baseline patient reported outcomes, treatment assignment, frequency of clinical contact and post-randomization weight change plus new complications). RESULTS: Poorer baseline control of HbA(1c) and cardiovascular disease risk factors predicted greater one-year improvements in treatment satisfaction. Similarly, poorer baseline patient reported outcome scores all individually predicted greater 1-year improvement in that same outcome. Accounting for baseline and post-randomization characteristics and treatment arm, 1-year change in HbA(1c) was unrelated to changes in overall physical or mental health; however, every one percentage-point (10.9 mmol/mol) reduction in HbA(1c) was associated with lower symptom count (ß = 0.599; P = 0.012), lower symptom distress (ß = 0.051; P = 0.001), and higher treatment satisfaction (ß = -2.514; P < 0.001). CONCLUSIONS: Independent of all relevant covariates, better glycaemic control over 1 year was associated with reduced patient-reported diabetes symptoms and symptom distress, and increased treatment satisfaction, but not overall physical and mental health. Further investigation is required to understand the specific psychosocial mechanisms that affect how patients value health and treatments.

Molecular cloning of CD31, a putative intercellular adhesion molecule closely related to carcinoembryonic antigen.

cDNA clones encoding CD31 have been isolated by transient expression. The sequence of CD31 expressed on human umbilical vein endothelial cells (HUVEC) is identical to that expressed on the monocyte-like cell line HL60. In HUVEC. CD31 is concentrated in regions of cell-cell contacts. CD31 is a member of the Ig superfamily and is most closely related to the carcinoembryonic antigen CEA, consisting of four contiguous C2 domains. The localization of CD31 to regions of cell-cell contacts, and the sequence similarity to CEA, a known intercellular adhesion molecule (ICAM), strongly suggest that CD31 may function as an ICAM, possibly mediating endothelial cell-cell contacts and also promoting interactions between leukocytes and endothelial cells.

CD59, an LY-6-like protein expressed in human lymphoid cells, regulates the action of the complement membrane attack complex on homologous cells.

A novel cell surface antigen has been identified on a wide range of lymphoid cells and erythrocytes. A mAb YTH 53.1 (CD59) against this antigen enhanced the lysis of human red cells and lymphocytes by homologous complement. Studies of reactive lysis using different species of C56, and of whole serum used as a source of C7-9, indicated that the inhibitory activity of the CD59 antigen is directed towards the homologous membrane attack complex. CD59 antigen was purified from human urine and erythrocyte stroma by affinity chromatography using the mAb YTH 53.1 immobilized on Sepharose, and, following transient expression of a human T cell cDNA library in COS cells, the corresponding cDNA also identified using the antibody. It was found that the CD59 antigen is a small protein (approximately 20 kD as judged by SDS-PAGE, 11.5 kD predicted from the isolated cDNA) sometimes associated with larger components (45 and 80 kD) in urine. The sequence of CD59 antigen is unlike that of other complement components or regulatory proteins, but shows 26% identity with that of the murine LY-6 antigen. CD59 antigen was released from the surface of transfected COS cells by phosphatidylinositol-specific phospholipase C, demonstrating that it is attached to the cell membrane by means of a glycolipid anchor; it is therefore likely to be absent from the surface of affected erythrocytes in the disease paroxysmal nocturnal hemoglobinuria.

Mapping the homotypic binding sites in CD31 and the role of CD31 adhesion in the formation of interendothelial cell contacts.

CD31 is a member of the immunoglobulin superfamily consisting of six Ig-related domains. It is constitutively expressed by platelets, monocytes, and some lymphocytes, but at tenfold higher levels on vascular endothelial cells. CD31 has both homotypic and heterotypic adhesive properties. We have mapped the homotypic binding sites using a deletion series of CD31-Fc chimeras and a panel of anti-CD31 monoclonal antibodies. An extensive surface of CD31 is involved in homotypic binding with domains 2 and 3 and domains 5 and 6 playing key roles. A model consistent with the experimental data is that CD31 on one cell binds to CD31 on an apposing cell in an antiparallel interdigitating mode requiring full alignment of the six domains of each molecule. In addition to establishing intercellular homotypic contacts. CD31 binding leads to augmented adhesion via beta 1 integrins. The positive cooperation between CD31 and beta 1 integrins can occur in heterologous primate cells (COS cells). The interaction is specific to both CD31 and beta 1 integrins. Neither intercellular adhesion molecule-1 (ICAM-1)/leukocyte function-associated antigen-1 (LCAM-1) nor neural cell adhesion molecule (NCAM)/NCAM adhesion leads to recruitment of beta 1 integrin adhesion pathways. Establishment of CD31 contacts have effects on the growth and morphology of endothelial cells. CD31(D1-D6)Fc inhibits the growth of endothelial cells in culture. In addition, papain fragments of anti-CD31 antibodies (Fab fragments) disrupt interendothelial contact formation and monolayer integrity when intercellular contacts are being formed. The same reagents are without effect once these contacts have been established, suggesting that CD31-CD31 interactions are critically important only in the initial phases of intercellular adhesion.

Involvement of the "I" domain of LFA-1 in selective binding to ligands ICAM-1 and ICAM-3.

To analyze the binding requirements of LFA-1 for its two most homologous ligands, ICAM-1 and ICAM-3, we compared the effects of various LFA-1 activation regimes and a panel of anti-LFA-1 mAbs in T cell binding assays to ICAM-1 or ICAM-3 coated on plastic. These studies demonstrated that T cell binding to ICAM-3 was inducible both from the exterior of the cell by Mn2+ and from the interior by an agonist of the "inside-out" signaling pathway. T cells bound both ICAM ligands with comparable avidity. A screen of 29 anti-LFA-1 mAbs led to the identification of two mAbs specific for the alpha subunit of LFA-1 which selectively blocked adhesion of T cells to ICAM-3 but not ICAM-1. These two mAbs, YTH81.5 and 122.2A5, exhibited identical blocking properties in a more defined adhesion assay using LFA-1 transfected COS cells binding to immobilized ligand. Blocking was not due to a steric interference between anti-LFA-1 mAbs and N-linked carbohydrate residues present on ICAM-3 but not ICAM-1. The epitopes of mAbs YTH81.5 and 122.2A5 were shown to map to the I domain of the LFA-1 alpha subunit. A third I domain mAb, MEM-83, has been previously reported to uniquely activate LFA-1 to bind ICAM-1 (Landis, R. C., R. I. Bennett, and N. Hogg. 1993. J. Cell Biol. 120:1519-1527). We now show that mAb MEM-83 is not able to stimulate binding of T cells to ICAM-3 over a wide concentration range. Failure to induce ICAM-3 binding by mAb MEM-83 was not due to a blockade of the ICAM-3 binding site on LFA-1. This study has demonstrated that two sets of functionally distinct mAbs recognizing epitopes in the I domain of LFA-1 are able to exert differential effects on the binding of LFA-1 to its ligands ICAM-1, and ICAM-3. These results suggest for the first time that LFA-1 is capable of binding these two highly homologous ligands in a selective manner and that the I domain plays a role in this process.

Rel-associated pp40: an inhibitor of the rel family of transcription factors.

The Rel-associated protein pp40 is functionally related to I kappa B, an inhibitor of the transcription factor NF-kappa B. Purified pp40 inhibits the DNA binding activity of the NF-kappa B protein complex (p50:p65 heterodimers), p50:c-Rel heteromers, and c-Rel homodimers. The sequence of the complementary DNA encoding pp40 revealed similarity to the gene encoding MAD-3, a protein with mammalian I kappa B-like activity. Protein sequencing of I kappa B purified from rabbit lung confirmed that MAD-3 encodes a protein similar to I kappa B. The sequence similarity between MAD-3 and pp40 includes a casein kinase II and consensus tyrosine phosphorylation site, as well as five repeats of a sequence found in the human erythrocyte protein ankyrin. These results suggest that rel-related transcription factors, which are capable of cytosolic to nuclear translocation, may be held in the cytosol by interaction with related cytoplasmic anchor molecules.

Cross-talk between cell adhesion molecules regulates the migration velocity of neutrophils.

BACKGROUND: Although the adhesive mechanisms underlying the capture and immobilization of circulating neutrophils in inflamed blood vessels have been well described, factors controlling the subsequent migration of neutrophils over and through the blood vessel endothelium are poorly understood. Directional rearrangement of the actin cytoskeleton within the neutrophil, along with modulation of integrin-mediated adhesion, are necessary for neutrophil migration. Signals from chemotactic agents and from the adhesive substrate may regulate these processes, but little is known about their relative importance or their mode of integration. RESULTS: We examined the kinetics of neutrophil migration after formyl tripeptide or platelet-activating factor was perfused over neutrophils that were already rolling on the adhesion molecule P-selectin, which was presented either on the surface of immobilized platelets or in purified form coated on glass capillaries. Upon activation, neutrophils stopped rolling, spread and began to migrate; each of these processes was dependent on beta2 integrin (CD11b/CD18). The rate of migration increased over a period of about 8 minutes and was modulated directly by both the P-selectin and the CD31 surface receptors. Antibody blockade of either CD31 or P-selectin on platelets resulted in a reduction in the velocity of migration, and simultaneous blockade of both receptors reduced velocity further. Purified CD31 and P-selectin (but not a control adhesion molecule, ICAM-1) increased migration velocity in a concentration-dependent and additive manner that reconstituted the migratory behaviour observed on platelets. CONCLUSIONS: These studies show that binding of ligands to CD31 and/or P-selectin modifies the rate of integrin-supported neutrophil migration. This novel example of 'cross-talk' between surface receptors suggests that cell adhesion molecules might generally transduce accessory signals between adjacent cells to modify their migratory responses to chemotactic signals.

Bacterial lipopolysaccharide-mediated fetal death. Production of a newly recognized form of inducible cyclooxygenase (COX-2) in murine decidua in response to lipopolysaccharide.

Maternal infection is a cause of spontaneous abortion and preterm labor in humans, but the pathophysiology is unclear. We hypothesized that eicosanoids play an important role in infection-driven pregnancy loss. To investigate this hypothesis, we administered lipopolysaccharide (LPS) to pregnant C3H/HeN mice and found that LPS administration caused fetal death in a dose-dependent fashion. Pretreatment with indomethacin significantly decreased the proportion of fetal death from 83% to < 25% in mice injected with 10 micrograms of LPS. Also, decidual explants from LPS-treated mice produced significantly more inflammatory eicosanoids, including prostaglandins E2 and F2 alpha and thromboxane B2, than controls. We investigated the regulatory mechanisms responsible for increased decidual prostanoid production in response to LPS. Western and Northern blots demonstrated that decidual protein and mRNA levels of a recently recognized highly inducible form of cyclooxygenase, COX-2, were substantially increased in mice treated with LPS. Induction of COX-2 was rapid: mRNA was detected 30 min after LPS injection. In contrast, another form of cyclooxygenase, COX-1, was only minimally induced in response to LPS. Our data indicate that LPS induces decidual prostanoid production via increased COX-2 expression. Since LPS-mediated fetal death is markedly diminished by pretreatment with indomethacin, COX-2-mediated eicosanoid production is likely a key pathophysiologic event in LPS-mediated fetal death.

Identification of an early-growth-response gene encoding a novel putative protein kinase.

Early-growth-response genes, also known as immediate-early genes, play important roles in regulating cell proliferation. We have identified a new type of early-growth-response gene product, a 77,811-Da putative serine/threonine kinase, which is highly inducible by serum and phorbol ester. mRNA encoding this putative kinase is markedly elevated within 1 h after treatment with mitogen, and this induction is synergistically increased by cycloheximide. Dexamethasone blocks serum induction of the kinase mRNA, as does transformation by v-Ki-ras. The kinase mRNA was detected in mouse brain, lung, and heart. This new putative kinase, which we term Snk, for serum-inducible kinase, showed similarity in its proposed catalytic domain to many other protein kinases; however, no other kinase showed enough sequence similarity with Snk to suggest the existence of a common function. Hence, Snk represents a new type of protein kinase involved in the early mitogenic response whose activity is transcriptionally and posttranscriptionally regulated.

Rapid repression of quiescence-specific gene expression by epidermal growth factor, insulin, and pp60v-src.

We isolated a cDNA for p20K, a secreted protein preferentially synthesized in nonproliferating cells. p20K mRNA and protein levels declined rapidly following treatment with various mitogens. DNA sequence analysis of the p20K cDNA predicted a novel protein distantly related to alpha 2 mu-globulin and plasma retinol-binding protein.

In search of an enzyme: the beta-secretase of Alzheimer's disease is an aspartic proteinase.

The deposition of beta-amyloid (Abeta) in the brain is a neuropathological feature of Alzheimer's disease. Abeta is cleaved from its precursor protein (APP) by processing at its N and C termini by enzymes known as beta- and gamma-secretases,respectively. The identity of these enzymes has been elusive but the search for the N-terminal secretase might have ended recently with the almost simultaneous publication by five major laboratories claiming a transmembrane aspartic proteinase to be the long sought after beta-secretase. Even at this early stage of its characterization, this aspartic proteinase fulfils many of the key criteria necessary for beta-secretase. The race is now on to develop inhibitors that could prove effective in halting the progression of Alzheimer's disease.

Normal human tissues, in addition to some tumors, express multiple different CD44 isoforms.

At least 20 different isoforms of the human CD44 lymphocyte-homing receptor/hyaluronan receptor have been described to date that arise from the differential splicing of up to 10 alternative exons (termed v1-v10) encoding the membrane-proximal extracellular domain. Although numerous analyses at the mRNA level have indicated tissue-specific expression of CD44 variants, few analyses have been performed at the protein level because of limited availability of suitable monoclonal antibodies. Recently, however, exon-specific monoclonal antibodies have been generated using bacterial fusion proteins, and these have been reported to detect high levels of vCD44 containing the v6 exon on human tumors. Together with earlier evidence linking this particular exon with tumor metastasis in the rat, these latter experiments have led to the interpretation that v6 splice variants play a causative role in tumor dissemination. In this paper we describe the use of a new and comprehensive panel of CD44 exon-specific monoclonal antibodies generated against a recombinant CD44(v3-10)-immunoglobulin chimera to study vCD44 expression in a large number of normal and neoplastic tissues. We show that the expression of vCD44 varies greatly among different human tumors and that some express either very low levels of vCD44 or no CD44 at all. Furthermore, we demonstrate that expression is not limited to isoforms containing the v6 exon but includes variants carrying v3, v4/5, and v8/9. Additionally, normal epithelial tissues are shown to express considerable levels of these same vCD44 isoforms. Such results argue against a ubiquitous role for vCD44 isoforms in promoting tumor growth and metastasis.

Cloning and expression of a chicken c-rel cDNA: unlike p59v-rel, p68c-rel is a cytoplasmic protein in chicken embryo fibroblasts.

We isolated and sequenced a 3727 bp clone of the c-rel proto-oncogene from a chicken embryo fibroblast (CEF) cDNA library. Sequence comparison to the retroviral oncogene v-rel showed conclusively that the v-rel protein is truncated at both the amino- and carboxy-termini as compared to the c-rel protein. In vitro transcription and translation of this clone yielded a 68,000 dalton polypeptide that co-migrated on SDS polyacrylamide gels with p68c-rel from avian spleen cells. We inserted this c-rel cDNA clone into an avian retroviral vector (pJD214c-rel), and over-expressed p68c-rel in CEF. Over-expression of p68c-rel did not induce morphological transformation of these cells. Unlike p59v-rel, which is a nuclear protein in CEF, indirect immunofluorescence showed that p68c-rel in JD214c-rel infected CEF is located exclusively in the cytoplasm of these cells, even though the sequence of p68c-rel showed that it contains a nuclear localizing sequence identical to the one previously identified in p59v-rel. Furthermore, the c-rel protein does contain a nuclear localizing sequence which can function in CEF since replacement of the v-rel nuclear localizing sequence with the homologous domain from c-rel resulted in a hybrid rel protein that was located in the nucleus in CEF. Mutant c-rel proteins, deleted of the carboxy-terminal sequences not present in p59v-rel, localized to the nucleus in CEF. Our results show that the carboxy-terminus of p68c-rel inhibits nuclear localization of the protein, and suggest that subcellular location may be a form of regulation of the activity of p68c-rel.

The deleted in colon cancer (DCC) gene is consistently expressed in colorectal cancers and metastases.

The DCC (deleted in colorectal cancer) gene was originally identified as a candidate tumour suppressor gene in colon carcinogenesis on the basis of allelic losses in chromosome 18q.21 in 70% of colon cancers. Reverse transcriptase polymerase chain reaction (RT-PCR) of DCC mRNA suggests that DCC expression may also be reduced in colon cancers. We have used monoclonal antibodies generated against the DCC immunoglobulin-like domain to investigate DCC isoforms and DCC protein expression during colon cancer progression. Normal mucosa and colonic tumour specimens representative of the range of colonic tumour progression from benign adenomatous polyps to metastases were compared by Western blot analyses. We show that while M(r) 194 000 DCC is present in normal colonic mucosa and adenomatous polyps, it is also similarly expressed in colorectal carcinomas and colonic metastases in the liver. The presence of DCC protein is consistent with the presence of DCC mRNA transcripts in the same tissue specimens. Notably DCC was not completely lost in any colonic tumour specimens examined, even those that had progressed to metastatic cancers. Quantitation of DCC protein expression in tissue specimens by densitometry demonstrated that both normal and malignant specimens exhibit a wide range of DCC protein levels and there was no significant correlation between diminished DCC protein expression and colon cancer progression. These results demonstrate the pattern of expression of the DCC gene product in colonic tumour progression and show that absence of DCC expression is not associated with colonic tumour progression.

Sequence specificity of cytosine methylation in the DNA of the Chinese hamster ovary (CHO-K1) cell line.

We have determined the DNA renaturation kinetics for those DNA sequences of the Chinese hamster ovary (CHO-K1) cells in which enzymatic cytosine methylation occurred immediately after strand synthesis and for those in which methylation was delayed after strand synthesis. DNA sequences showing immediate or delayed methylation were found to be distributed throughout all repetition classes of the DNA of these cells, with a slight concentration of immediate methylation in moderately repetitive sequences and with delayed methylation being slightly over-represented in the highly repetitive fraction. However, DNA sequences showing both classes of methylation were represented equally in unique DNA sequences. We interpret these data to mean that the methylase acting near the replication forks (the 'immediate' methylase) is a relatively inefficient enzyme, missing some 20% of hemimethylated sites produced by DNA replication in these cells. We suggest that the methylase performing maintenance methylation at sites remote from the replication forks (the 'delayed' methylase) is simply a back-up enzyme for the first and that it has no true sequence specificity. The implications of this for the function(s) of DNA methylation in mammalian cells are discussed.

Identification and characterisation of human Junctional Adhesion Molecule (JAM).

It is widely believed that migrating immune cells utilise the intercellular junctions as routes of passage, and in doing so cause the transient disruption of junctional structures. Thus there is much interest in the molecules that have been identified at cell-cell contact points and their potential involvement in the control of leukocyte diapedesis. In this report we describe the human orthologue to Junctional Adhesion Molecule (JAM), a recently identified member of the immunoglobulin superfamily expressed at intercellular junctions (Martin-Padura et al., 1998). The human protein shares a highly conserved structure and sequence with the murine protein. However it is distinct in that it is constitutively expressed on circulating neutrophils, monocytes, platelets and lymphocyte subsets. This broad expression pattern is similar to another IgSF molecule, CD31, expressed at intercellular junctions, and may indicate further complexities in the control of leukocyte/ endothelial interactions.

Progression of micronutrient alteration and hepatotoxicity following acute PCB126 exposure.

Polychlorinated Biphenyls (PCBs) are industrial chemicals that have become a persistent threat to human health due to ongoing exposure. A subset of PCBs, known as dioxin-like PCBs, pose a special threat given their potent hepatic effects. Micronutrients, especially Cu, Zn and Se, homeostatic dysfunction is commonly seen after exposure to dioxin-like PCBs. This study investigates whether micronutrient alteration is the byproduct of the ongoing hepatotoxicity, marked by lipid accumulation, or a concurrent, yet independent event of hepatic damage. A time course study was carried out using male Sprague-Dawley rats with treatments of PCB126, the prototypical dioxin-like PCB, resulting in 6 different time points. Animals were fed a purified diet, based on AIN-93G, for three weeks to ensure micronutrient equilibration. A single IP injection of either tocopherol-stripped soy oil vehicle (5 mL/kg) or 5 µmol/kg PCB126 dose in vehicle was given at various time points resulting in exposures of 9h, 18 h, 36 h, 3 days, 6 days, and 12 days. Mild hepatic vacuolar change was seen as early as 36 h with drastic changes at the later time points, 6 and 12 days. Micronutrient alterations, specifically Cu, Zn, and Se, were not seen until after day 3 and only observed in the liver. No alterations were seen in the duodenum, suggesting that absorption and excretion may not be involved. Micronutrient alterations occur with ROS formation, lipid accumulation, and hepatomegaly. To probe the mechanistic underpinnings, alteration of gene expression of several copper chaperones was investigated; only metallothionein appeared elevated. These data suggest that the disruption in micronutrient status is a result of the hepatic injury elicited by PCB126 and is mediated in part by metallothionein.

Regulation of hen granulosa cell prostaglandin production by transforming growth factors during follicular development: involvement of cyclooxygenase II.

PGs are important physiological regulators of granulosa cell proliferation during ovarian follicular development. The rate-limiting step in the synthesis of PGs from arachidonic acid is catalyzed by cyclooxygenase (COX). Although two isoforms of COX (COX I and COX II) have been identified in the ovary, the nature and physiological significance of their regulation by transforming growth factor (TGF) alpha and TGF beta are unclear. The object of the present studies was to investigate the regulation of hen granulosa cell PG production by TGF alpha and TGF beta and the role of COX II in this process. Granulosa cells from the first (F1) and fifth and sixth (F5-F6) largest hen preovulatory follicles were cultured for up to 21 h in the presence of TGF alpha (0-10 ng/ml) and/or TGF beta (0-20 ng/ml). COX protein and messenger RNA (mRNA) levels were determined by Western blot and Northern analysis, respectively. Granulosa cell PGF and PGE secretion was increased by TGF alpha but suppressed by TGF beta in vitro. The increase in PG secretion was accompanied by an elevation of COX II content, which was dose and time dependent, with maximum response observed within 6-12 h. Maximal increase in COX II mRNA abundance was evident at 4 and 8 h in cells from F1 and F5-F6, respectively. Although TGF alpha-stimulated PG secretion was higher in cells from mature follicle (F1), the magnitude of change in COX II mRNA abundance and protein content was, however, not dependent on follicular maturation. TGF beta significantly suppressed basal and TGF alpha-induced COX II transcript levels. COX I transcript, however, was undetectable irrespective of the presence of TGF alpha, duration of culture or the stage of follicular development. In conclusion, the mitogenic response of granulosa cells to TGF alpha is mediated by changes in PG secretion, which are regulated at the level of COX II. Unlike the control of PG secretion, the regulation of COX II by TGF alpha is independent of follicular maturation. Whereas COX II may be important in the signaling cascade for TGF alpha in the regulation of granulosa cell proliferation, the mechanism(s) of regulation of follicular stage-dependent PG secretion by the growth factor is complex and requires further investigation.

The role of the cysteine-rich region of the beta2 integrin subunit in the leukocyte function-associated antigen-1 (LFA-1, alphaLbeta2, CD11a/CD18) heterodimer formation and ligand binding.

The cysteine-rich region (CRR) of the beta2 integrin subunit was replaced by that of beta1 to give the chimera beta2NV1. Beta2NV1 can combine with alphaL to form a variant leukocyte-function-associated antigen (LFA)-1 on COS cell surface, suggesting that the specificity of the beta2 interaction with alphaL does not lie in the CRR. Unlike those expressing wild-type LFA-1, COS cells expressing alphaL beta2NV1 are constitutively active in intercellular adhesion molecule (ICAM)-1 adhesion. These results suggest that activation of LFA-1 involves the release of an intramolecular constraint, which is maintained, in part, by the authentic beta2 CRR.

COX-2 inhibition, apoptosis, and chemoprevention by nonsteroidal anti-inflammatory drugs.

Non-steroidal anti-inflammatory drugs (NSAIDs) have as their common mechanism the inhibition of cyclooxygenase (COX) enzymes, of which two isoforms (COX-1 and COX-2) exist. The effect of NSAIDs on chemoprevention and tumor regression has been shown in animal models, epidemiologic studies, and in treatment of patients. The exact biochemical and cellular mechanisms underlying each of these phenomena is only partially understood. Processes that have been recently implicated as being important include the inhibition of tumor cell growth, prevention of angiogenesis, and induction of apoptosis in neoplastic cells.