Issue published August 1, 1998

W e report here the role of the CXC chemokine, epithelial neutrophil activating peptide (ENA-78), as an angiogenic factor in human non-small cell lung cancer (NSCLC). In freshly isolated human specimens of NSCLC, elevated levels of ENA-78 were found that strongly correlated with the vascularity of the tumors. In a SCID mouse model of human NSCLC tumorigenesis, expression of ENA-78 in developing tumors correlated with tumor growth in two different NSCLC cell lines. Furthermore, passive immunization of NSCLC tumor-bearing mice with neutralizing anti-ENA-78 antibodies reduced tumor growth, tumor vascularity, and spontaneous metastases, while having no effect on the proliferation of NSCLC cells either in vitro or in vivo. These findings suggest that ENA-78 is an important angiogenic factor in human NSCLC.

A major goal of the transplant field is to tolerize donor T cells to prevent graft-versus-host disease (GVHD) (1). We describe an ex vivo approach in which the blockade of CD40 ligand (CD40L:CD154):CD40 interactions, a pathway required for optimal T cell expansion, induces donor CD4(+) T cells to become tolerant to host alloantigens (2). High doses of tolerized cells did not cause GVHD lethality in vivo. T cells had intact responses to antigens not present during tolerization. Tolerance was long lived and not readily reversible in vivo. These data have significant implications for the use of tolerization approaches to prevent human GVHD.

T he role of mutations of the granulocyte colony-stimulating factor receptor (G-CSFR) in the pathogenesis of severe congenital neutropenia (SCN) and the subsequent development of acute myeloid leukemia (AML) is controversial. Mice carrying a targeted mutation of their G-CSFR that reproduces the mutation found in a patient with SCN and AML have been generated. The mutant G-CSFR allele is expressed in a myeloid-specific fashion at levels comparable to the wild-type allele. Mice heterozygous or homozygous for this mutation have normal levels of circulating neutrophils and no evidence for a block in myeloid maturation, indicating that resting granulopoiesis is normal. However, in response to G-CSF treatment, these mice demonstrate a significantly greater fold increase in the level of circulating neutrophils. This effect appears to be due to increased neutrophil production as the absolute number of G-CSF-responsive progenitors in the bone marrow and their proliferation in response to G-CSF is increased. Furthermore, the in vitro survival and G-CSF-dependent suppression of apoptosis of mutant neutrophils are normal. Despite this evidence for a hyperproliferative response to G-CSF, no cases of AML have been detected to date. These data demonstrate that the G-CSFR mutation found in patients with SCN is not sufficient to induce an SCN phenotype or AML in mice.

S ince dopamine receptors are important in the regulation of renal and cardiovascular function, we studied the cardiovascular consequences of the disruption of the D3 receptor, a member of the family of D2-like receptors, expressed in renal proximal tubules and juxtaglomerular cells. Systolic and diastolic blood pressures were higher (approximately 20 mmHg) in heterozygous and homozygous than in wild-type mice. An acute saline load increased urine flow rate and sodium excretion to a similar extent in wild-type and heterozygous mice but the increase was attenuated in homozygous mice. Renal renin activity was much greater in homozygous than in wild-type mice; values for heterozygous mice were intermediate. Blockade of angiotensin II subtype-1 receptors decreased systolic blood pressure for a longer duration in mutant than in wild-type mice. Thus, disruption of the D3 receptor increases renal renin production and produces renal sodium retention and renin-dependent hypertension.

P aracrine effect of transforming growth factor-beta1 (TGF-beta1) on autoimmune insulitis and diabetes was studied by transgenic production of the active form of porcine TGF-beta1 (pTGF-beta1) in pancreatic islet (islet) alpha cells in nonobese diabetic (NOD) mice under the control of rat glucagon promoter (RGP) (NOD-RGP-TGF-beta1). None of 27 NOD-RGP-TGF- beta1 mice developed diabetes by 45 wk of age, in contrast to 40 and 71% in male and female nontransgenic mice, respectively. None of the NOD-RGP-TGF-beta1 mice developed diabetes after cyclophosphamide (CY) administration. Adoptive transfer of splenocytes of NOD-RGP-TGF-beta1 mice to neonatal NOD mice did not transfer diabetes after CY administration. Adoptive transfer of three types of diabetogenic lymphocytes to NOD-RGP-TGF-beta1 and nontransgenic mice after CY administration led to the lower incidence of diabetes in NOD-RGP-TGF-beta1 mice versus that in nontransgenic mice: 29 vs. 77% for diabetogenic splenocytes, 25 vs. 75% for islet beta cell-specific Th1 clone cells, and 0 vs. 50% for islet beta cell-specific CD8(+) clone cells, respectively. Based on these, it is concluded that autoimmune diabetes in NOD mice is not a systemic disease and it can be completely prevented by the paracrine TGF-beta1 in the islet compartment through protection against CD4(+) and CD8(+) effector lymphocytes.

G lycogen storage disease type 0 (GSD-0) is a rare form of fasting hypoglycemia presenting in infancy or early childhood and accompanied by high blood ketones and low alanine and lactate concentrations. Although feeding relieves symptoms, it often results in postprandial hyperglycemia and hyperlactatemia. The glycogen synthase (GS) activity has been low or immeasurable in liver biopsies, whereas the liver glycogen content has been only moderately decreased. To investigate whether mutations in the liver GS gene (GYS2) on chromosome 12p12.2 were involved in GSD-0, we determined the exon-intron structure of the GYS2 gene and examined nine affected children from five families for linkage of GSD-0 to the GYS2 gene. Mutation screening of the 16 GYS2 exons was done by single-strand conformational polymorphism (SSCP) and direct sequencing. Liver GS deficiency was diagnosed from liver biopsies (GS activity and glycogen content). GS activity in the liver of the affected children was extremely low or nil, resulting in subnormal glycogen content. After suggestive linkage to the GYS2 gene had been established (LOD score = 2.9; P < 0.01), mutation screening revealed several different mutations in these families, including a premature stop codon in exon 5 (Arg246X), a 5'-donor splice site mutation in intron 6 (G+1T--> CT), and missense mutations Asn39Ser, Ala339Pro, His446Asp, Pro479Gln, Ser483Pro, and Met491Arg. Seven of the affected children carried mutations on both alleles. The mutations could not be found in 200 healthy persons. Expression of the mutated enzymes in COS7 cells indicated severely impaired GS activity. In conclusion, the results demonstrate that GSD-0 is caused by different mutations in the GYS2 gene.

R esident macrophages have been suggested to participate in the initiation of beta cell damage during the development of autoimmune diabetes. The purpose of this study was to determine if the endogenous production and release of interleukin 1 (IL-1) in human islets of Langerhans by resident macrophages results in the inhibition of beta cell function. Treatment of human islets with a combination of tumor necrosis factor (TNF) + lipopolysaccharide (LPS) + interferon-gamma (IFN-gamma) stimulates inducible nitric oxide synthase (iNOS) expression, nitric oxide production, and inhibits glucose-stimulated insulin secretion. The IL-1 receptor antagonist protein (IRAP) prevents TNF + LPS + IFN-gamma-induced iNOS expression and nitrite production, and attenuates the inhibitory effects on glucose-stimulated insulin secretion by human islets. Inhibition of iNOS activity by aminoguanidine also attenuates TNF + LPS + IFN-gamma-induced inhibition of insulin secretion by human islets. These results indicate that the inhibitory effects of TNF + LPS + IFN-gamma are mediated by nitric oxide, produced by the actions of IL-1 released endogenously within human islets. Reverse transcriptase polymerase chain reaction was used to confirm that TNF + LPS + IFN-gamma stimulates the expression of both IL-1alpha and IL-1beta in human islets. Two forms of evidence indicate that resident macrophages are the human islet cellular source of IL-1: culture conditions that deplete islet lymphoid cells prevent TNF + LPS + IFN-gamma-induced iNOS expression, nitric oxide production, and IL-1 mRNA expression by human islets; and IL-1 and the macrophage surface marker CD69 colocalize in human islets treated with TNF + LPS + IFN-gamma as determined by immunohistochemical analysis. Lastly, nitric oxide production is not required for TNF + LPS + IFN-gamma-induced IL-1 release in human islets. However, cellular damage stimulates IL-1 release by islet macrophages. These findings support the hypothesis that activated islet macrophages may mediate beta cell damage during the development of insulin-dependent diabetes by releasing IL-1 in human islets followed by cytokine-induced iNOS expression by beta cells.

M itochondrial fatty acid beta-oxidation is important for energy production, which is stressed by the different defects found in this pathway. Most of the enzyme deficiencies causing these defects are well characterized at both the protein and genomic levels. One exception is carnitine palmitoyltransferase I (CPT I) deficiency, of which until now no mutations have been reported although the defect is enzymatically well characterized. CPT I is the key enzyme in the carnitine-dependent transport across the mitochondrial inner membrane and its deficiency results in a decreased rate of fatty acid beta-oxidation. Here we report the first delineation of the molecular basis of hepatic CPT I deficiency in a new case. cDNA analysis revealed that this patient was homozygous for a missense mutation (D454G). The effect of the identified mutation was investigated by heterologous expression in yeast. The expressed mutant CPT IA displayed only 2% of the activity of the expressed wild-type CPT IA, indicating that the D454G mutation is the disease-causing mutation. Furthermore, in patient's fibroblasts the CPT IA protein was markedly reduced on immunoblot, suggesting that the mutation renders the protein unstable.

D uring each reproductive cycle, a preovulatory surge of gonadotropins induces meiotic maturation of the oocyte in the preovulatory follicle followed by ovulation. Although gonadotropins stimulate cAMP production in somatic cells of the follicle, a decrease in intra-oocyte cAMP levels is required for resumption of meiosis in oocytes. Based on the observed compartmentalization of the cAMP-degrading enzyme, phosphodiesterase, in follicular somatic and germ cells, inhibitors of phosphodiesterase 3 were used to block meiosis in ovulating oocytes in rodents. By this strategy, we demonstrated that fertilization and pregnancy could be prevented without disturbing follicle rupture and normal estrous cyclicity. In contrast to conventional contraceptive pills that disrupt ovarian steroidogenesis and reproductive cycles, the present strategy achieves effective contraception by selective blockage of oocyte maturation and development without alterations in ovulation and reproductive cyclicity.

L iver fibrosis results from the excessive secretion of matrix proteins by hepatic stellate cells (HSC), which proliferate during fibrotic liver injury. We have studied a model of spontaneous recovery from liver fibrosis to determine the biological mechanisms mediating resolution. Livers were harvested from rats at 0, 3, 7, and 28 d of spontaneous recovery from liver fibrosis induced by 4 wk of twice weekly intraperitoneal injections with CCl4. Hydroxyproline analysis and histology of liver sections indicated that the advanced septal fibrosis observed at time 0 (peak fibrosis) was remodeled over 28 d of recovery to levels close to control (untreated liver). alpha-Smooth muscle actin staining of liver sections demonstrated a 12-fold reduction in the number of activated HSC over the same time period with evidence of HSC apoptosis. Ribonuclease protection analysis of liver RNA extracted at each recovery time point demonstrated a rapid decrease in expression of the collagenase inhibitors TIMP-1 and TIMP-2, whereas collagenase mRNA expression remained at levels comparable to peak fibrosis. Collagenase activity in liver homogenates increased through recovery. We suggest that apoptosis of activated HSC may vitally contribute to resolution of fibrosis by acting as a mechanism for removing the cell population responsible for both producing fibrotic neomatrix and protecting this matrix from degradation via their production of TIMPs.

H uman invasive soft-tissue infections caused by group A Streptococcus are associated with significant morbidity and mortality. To investigate the pathogenesis of these serious infections, we characterized the host response to bacterial challenge with an M-type 3 isolate recovered from a patient with necrotizing fasciitis, or with isogenic gene replacement mutants deficient in cysteine protease, hyaluronic acid capsule, or M protein in a murine model of human invasive soft-tissue infection. Animals challenged with the wild-type or cysteine protease-deficient strain developed spreading tissue necrosis at the site of inoculation, became bacteremic, and subsequently died. Histopathologic examination of the necrotic lesion revealed bacteria throughout inflamed subcutaneous tissue. Arterioles and venules in the subcutaneous layer were thrombosed and the overlying tissue was infarcted. In contrast, animals challenged with either an acapsular or M protein-deficient mutant developed a focal area of tissue swelling at the site of inoculation without necrosis or subsequent systemic disease. Histopathologic examination of the soft-tissue lesion demonstrated bacteria confined within a well-formed subcutaneous abscess. We conclude that the group A streptococcal hyaluronic acid capsule and M protein, but not the cysteine protease, are critical for the development of tissue necrosis, secondary bacteremia, and lethal infection in a murine model of human necrotizing fasciitis.

I L-12 has significant antitumor activity in mice that may be mediated by CD8(+) T cells. We show in this report that repeated subcutaneous injections of IL-12 in patients with cancer resulted in the selective expansion of a subset of peripheral blood CD8(+) T cells. This T cell subset expressed high levels of CD18 and upregulated IL-12 receptor expression after IL-12 treatment in vivo. In normal subjects, these CD3(+)CD8(+)CD18(bright) T cells expressed IL-12 and IL-2 receptors and adhesion/costimulatory molecules to a greater degree than other CD8(+) and CD4(+) T cells. They appeared morphologically as large granular lymphocytes, although they did not express NK cell markers such as CD56. In addition, CD8(+)CD18(bright) T cells were almost exclusively T cell receptor (TCR) alphabeta+, and exhibited a TCR Vbeta repertoire that was strikingly oligoclonal, whereas the Vbeta repertoire of CD18(dim) T cells was polyclonal. Although CD8+CD18(bright) T cells demonstrated little functional responsiveness to IL-12 or IL-2 alone in vitro, they responded to the combination of IL-12+IL-2 with strong IFN-gamma production and proliferation and enhanced non-MHC-restricted cytolytic activity. In contrast, CD18(dim) T cells were not activated by IL-12 or IL-2, alone or in combination. These findings demonstrate that CD8+CD18(bright) T cells are a unique population of peripheral blood lymphocytes with features of both memory and effector cells that are capable of TCR-independent activation through combined stimulation with IL-12+IL-2. As this activation results in IFN-gamma production and enhanced cytolytic activity, these T cells may play a role in innate as well as acquired immunity to tumors and infectious pathogens. Additional studies will be necessary to determine whether CD8+CD18(bright) T cells mediate the antitumor effect of IL-12 or IL-2 administered to cancer patients, and if so, whether maximal activation of these T cells with the combination of IL-12+IL-2 in vivo can augment the clinical effectiveness of these cytokines.

F ormation of the atherosclerotic intima must involve altered metabolism of the elastin-rich arterial extracellular matrix. Proteases potentially involved in these processes remain unclear. This study examined the expression of the potent elastases cathepsins S and K in human atheroma. Normal arteries contained little or no cathepsin K or S. In contrast, macrophages in atheroma contained abundant immunoreactive cathepsins K and S. Intimal smooth muscle cells (SMC), especially cells appearing to traverse the internal elastic laminae, also contained these enzymes. Extracts of atheromatous tissues had approximately twofold greater elastase-specific activity than extracts of uninvolved arteries, mostly due to cysteine proteases. Cultured human SMC displayed no immunoreactive cathepsins K and S and exhibited little or no elastolytic activity when incubated with insoluble elastin. SMC stimulated with the atheroma-associated cytokines IL-1beta or IFN-gamma secreted active cathepsin S and degraded substantial insoluble elastin (15-20 microg/10(6) cells/24 h). A selective inhibitor of cathepsin S blocked > 80% of this elastolytic activity. The presence of cathepsins K and S at sites of vascular matrix remodeling and the ability of SMC and macrophages to use these enzymes to degrade elastin supports a role for elastolytic cathepsins in vessel wall remodeling and identifies novel therapeutic targets in regulating plaque stability.

T he etiology of bronchopulmonary dysplasia (BPD), a chronic lung disease of infants surviving respiratory distress syndrome, remains fundamentally enigmatic. BPD is decreasing in severity but continues to be a major problem in pediatric medicine, being especially prevalent among very premature infants. Increased numbers of pulmonary neuroendocrine cells containing bombesin-like peptide (BLP) have been reported to occur in human infants with BPD. We tested the hypothesis that BLP mediates BPD using the hyperoxic baboon model. Urine BLP levels increased soon after birth only in 100% O2-treated 140-d animals which developed BPD, correlating closely with severity of subsequent chronic lung disease. Similar elevations in urine BLP were observed in the 125-d baboon "interrupted gestation" model of BPD. Postnatal administration of anti-BLP antibody attenuated clinical and pathological evidence of chronic lung disease in the hyperoxic baboon model. Urine BLP could be a biological predictor of infants at risk for BPD, and blocking BLP postnatally could be useful for BPD prevention.

N itric oxide (NO) is a free radical involved in the regulation of many cell functions and in the expression of several diseases. We have found that the antimalarial and antiinflammatory drug, chloroquine, is able to stimulate NO synthase (NOS) activity in murine, porcine, and human endothelial cells in vitro: the increase of enzyme activity is dependent on a de novo synthesis of some regulatory protein, as it is inhibited by cycloheximide but is not accompanied by an increased expression of inducible or constitutive NOS isoforms. Increased NO synthesis is, at least partly, responsible for chloroquine-induced inhibition of cell proliferation: indeed, NOS inhibitors revert the drug-evoked blockage of mitogenesis and ornithine decarboxylase activity in murine and porcine endothelial cells. The NOS-activating effect of chloroquine is dependent on its weak base properties, as it is exerted also by ammonium chloride, another lysosomotropic agent. Both compounds activate NOS by limiting the availability of iron: their stimulating effects on NO synthesis and inhibiting action on cell proliferation are reverted by iron supplementation with ferric nitrilotriacetate, and are mimicked by incubation with desferrioxamine. Our results suggest that NO synthesis can be stimulated in endothelial cells by chloroquine via an impairment of iron metabolism.

T hymic nurse cells are known to interact with T cells and play a role in their functional maturation. However, the role of nurse cells in B cell maturation and differentiation is less well established, especially at extralymphoid sites. To address this issue, nurse-like cell clones from bone marrow and synovial tissue of patients with RA (RA-NLC) were established and characterized. RA-NLC constitutively expressed CD29, CD49c, CD54 (ICAM-1), CD106 (VCAM-1), CD157 (BST-1), and class I MHC molecules, and secreted IL-6, IL-7, IL-8, granulocyte-macrophage colony-stimulating factor (GM-CSF) and granulocyte colony-stimulating factor (G-CSF). Bone marrow-derived and synovial RA-NLC differed in that the former secreted IL-7 and expressed a greater density of CD157 constitutively and after stimulation with IFNgamma, whereas the latter secreted G-CSF and more IL-6. Stimulation of both bone marrow and synovial RA-NLC induced expression of CD40 and class II MHC, but not CD154 (CD40L) or CD35. RA-NLC rescued peripheral B cells from spontaneous apoptosis and promoted survival of B cells for > 4 wk. B cell survival was blocked by antibodies to CD106 or CD157. RA-NLC also increased Ig production from B cells. After long-term culture (4-6 wk) with RA-NLC, but not alone or with fibroblasts, outgrowth of B cells was observed. All B cell lines derived from these cultures had been transformed by EBV, although the RA-NLC themselves were not infected with EBV. Precursor frequency analysis indicated that approximately 1 in 12,500 peripheral B cells could give rise to these EBV-transformed B cell lines upon coculture with RA-NLC. These results indicate that RA-NLC from bone marrow and synovium have the capacity to rescue B cells from spontaneous apoptosis, facilitate Ig production, and promote the outgrowth of EBV-transformed B lymphoblastoid cells. These findings suggest that RA-NLC may play a role in the local and systemic hyperreactivity of B cells characteristic of rheumatoid arthritis.

R elease of glucose by liver and kidney are both increased in diabetic animals. Although the overall release of glucose into the circulation is increased in humans with diabetes, excessive release of glucose by either their liver or kidney has not as yet been demonstrated. The present experiments were therefore undertaken to assess the relative contributions of hepatic and renal glucose release to the excessive glucose release found in type 2 diabetes. Using a combination of isotopic and balance techniques to determine total systemic glucose release and renal glucose release in postabsorptive type 2 diabetic subjects and age-weight-matched nondiabetic volunteers, their hepatic glucose release was then calculated as the difference between total systemic glucose release and renal glucose release. Renal glucose release was increased nearly 300% in diabetic subjects (321+/-36 vs. 125+/-15 micromol/min, P < 0.001). Hepatic glucose release was increased approximately 30% (P = 0.03), but increments in hepatic and renal glucose release were comparable (2.60+/-0.70 vs. 2.21+/-0.32, micromol.kg-1.min-1, respectively, P = 0.26). Renal glucose uptake was markedly increased in diabetic subjects (353+/-48 vs. 103+/-10 micromol/min, P < 0.001), resulting in net renal glucose uptake in the diabetic subjects (92+/-50 micromol/ min) versus a net output in the nondiabetic subjects (21+/-14 micromol/min, P = 0.043). Renal glucose uptake was inversely correlated with renal FFA uptake (r = -0.51, P < 0.01), which was reduced by approximately 60% in diabetic subjects (10. 9+/-2.7 vs. 27.0+/-3.3 micromol/min, P < 0.002). We conclude that in type 2 diabetes, both liver and kidney contribute to glucose overproduction and that renal glucose uptake is markedly increased. The latter may suppress renal FFA uptake via a glucose-fatty acid cycle and explain the accumulation of glycogen commonly found in the diabetic kidney.

H ypertriglyceridemia is a metabolic complication of retinoid therapy. In this study, we analyzed whether retinoids increase the expression of apo C-III, an antagonist of plasma triglyceride catabolism. In men, isotretinoin treatment (80 mg/d; 5 d) resulted in elevated plasma apo C-III, but not apo E concentrations. In human hepatoma HepG2 cells, retinoids increased apo C-III mRNA and protein production. Transient transfection experiments indicated that retinoids increase apo C-III expression at the transcriptional level. This increased apo C-III transcription is mediated by the retinoid X receptor (RXR), since LG1069 (4-[1-(5,6,7,8-tetrahydro-3,5,5,8, 8-pentamethyl-2-naphtalenyl)ethenyl]benzoic acid), a RXR-specific agonist, but not TTNPB ((E)- 4-[2-(5,6,7,8-tetrahydro-5,5,8, 8-tetramethyl-2-naphtalenyl)propenyl]benzoic acid), a retinoic acid receptor (RAR)-specific agonist, induced apo C-III mRNA in HepG2 cells and primary human hepatocytes. Mutagenesis experiments localized the retinoid responsiveness to a cis-element consisting of two imperfect AGGTCA sequences spaced by one oligonucleotide (DR-1), within the previously identified C3P footprint site. Cotransfection assays showed that RXR, but not RAR, activates apo C-III transcription through this element either as a homo- or as a heterodimer with the peroxisome proliferator-activated receptor. Thus, apo C-III is a target gene for retinoids acting via RXR. Increased apo C-III expression may contribute to the hypertriglyceridemia and atherogenic lipoprotein profile observed after retinoid therapy.

E xtracellular mobilization of Group IIA 14-kD phospholipase A2 (PLA2) in glycogen-induced rabbit inflammatory peritoneal exudates is responsible for the potent bactericidal activity of the inflammatory fluid toward Staphylococcus aureus (1996. J. Clin. Invest. 97:250-257). Because similar levels of PLA2 are induced in plasma during systemic inflammation, we have tested whether this gives rise to plasma bactericidal activity not present in resting animals. Baboons were injected intravenously (i.v.) with a lethal dose of Escherichia coli and plasma or serum was collected before and at hourly intervals after injection. After infusion of bacteria, PLA2 levels in plasma and serum rose > 100-fold over 24 h to approximately 1 microg PLA2/ml. Serum collected at 24 h possessed potent bactericidal activity toward S. aureus, Streptococcus pyogenes, and encapsulated E. coli not exhibited by serum collected from unchallenged animals. Bactericidal activity toward S. aureus and S. pyogenes was nearly completely blocked by a monoclonal antibody to human Group IIA PLA2 and addition of purified human Group IIA PLA2 to prechallenge serum conferred potent antistaphylococcal and antistreptococcal activity equal to that of the 24 h post-challenge serum. PLA2-dependent bactericidal activity was enhanced approximately 10x by factor(s) present constitutively in serum or plasma. Bactericidal activity toward encapsulated E. coli was accompanied by extensive bacterial phospholipid degradation mediated, at least in part, by the mobilized Group IIA PLA2 but depended on the action of other bactericidal factors in the 24-h serum. These findings further demonstrate the contribution of Group IIA PLA2 to the antibacterial potency of biological fluids and suggest that mobilization of this enzyme during inflammation may play an important role in host defense against invading bacteria.