Concentrations of IL-10 in preterm human milk and in milk from mothers of infants with necrotizing enterocolitis.

BACKGROUND: Despite the protective effects of human milk against necrotizing enterocolitis, the incidence is highest in the extremely premature infant, and only minimally decreased with feeding human milk. This suggests that certain protective agents may be lower in milk from mothers delivering extremely premature infants. The anti-inflammatory cytokine IL-10 was one possibility. AIM: We hypothesized that low concentrations of IL-10 in preterm milk contribute to the development of necrotizing enterocolitis in extremely premature infants. METHODS: IL-10 in human milk collected at weeks 1, 2, and 4 postpartum was measured by ELISA in mothers of infants born extremely premature at 23-27 wk gestation (group EP), premature at 32-36 wk gestation (group P), and term at 38-42 wk gestation (group T). Single milk samples were collected from a separate group of mothers whose infants developed necrotizing enterocolitis. RESULTS: There were no significant differences in concentrations of milk IL-10 among groups EP, P, or T. Concentrations of IL-10 declined as lactation progressed (p < 0.001). IL-10 in milk was frequently undetected in all groups, but even more so in the milk of the group of women whose infants had necrotizing enterocolitis (86%) than in groups EP (40%) and P (27%) (p < 0.01). CONCLUSION: IL-10 was present in preterm milk from most women, and the concentrations in preterm and term milk were not significantly different. A paucity of IL-10 in human milk was found in certain mothers in each group, especially in those whose infants developed necrotizing enterocolitis.

Interleukin-8, aquaporin-1, and inducible nitric oxide synthase in smoke and burn injured sheep treated with percutaneous carbon dioxide removal.

We previously showed that a percutaneous arteriovenous gas exchanger was effective in removing CO2 and reversing respiratory failure in an ovine model of adult respiratory distress syndrome (ARDS) produced by smoke inhalation and burn injury (Alpard et al., Ann Surg 230:215-224, 1999). In this study, we tested the hypothesis that arteriovenous CO2 removal (AVCO2R) lessened endogenous inflammation in the lung. Myeloperoxidase activity, aquaporin-1 (AQP-1), interleukin-8 (IL-8), and inducible nitric oxide synthase mRNAs as well as aquaporin-1, and IL-8 protein were measured in ovine lung tissue. Lung tissue was taken at 96 h (time of sacrifice) from animals with combined smoke inhalation and 40% third degree dermal burn and subsequently treated with AVCO2R or sham (ventilator alone) after onset of ARDS (PaO2:FiO2 ratio of < 200). Myeloperoxidase activity was 1.862 +/- 0.302 U/mg protein in the ventilator group and 0.830 +/- 0.141 in the AVCO2R plus ventilator group. AQP-1 mRNA was 140,482 +/- 31,702 copies/microg total RNA in the ventilator group and 61,854 +/- 22,433 copies/microg total RNA in the AVCO2R plus ventilator group (p = 0.076). mRNA for IL-8 mRNA in the ventilator alone treated animals was 74,000 +/- 3,300 copies/microg total RNA compared to < 1,000 copies/microg total RNA in the ventilator plus AVCO2R group. This result was highly significant (p < 0.001) Inducible nitric oxide synthase mRNA was 7,853 +/- 2,229 copies/microg total RNA for the AVCO2R group and 5,854 +/- 2,070 copies/microg total RNA for the ventilator managed animals. These differences were not statistically significant (p = 0.54). Percutaneous AVCO2R produced a specific decrease in IL-8 in the smoke and burn injured animals. Furthermore, this effect was consistent with cell signaling mechanisms that increase the expression of IL-8 by cyclic stretching and the observed reduction in the number of neutrophils in the lung parenchyma. Therefore, we speculate that the mechanism by which CO2 removal exerts a beneficial effect may be due to both decreases in ventilatory requirements, with an accompanying reduction in alveolar stretching, and reduction of neutrophil numbers in lung tissue.

Genesis of progressive T-cell deficiency owing to a single missense mutation in the common gamma chain gene.

Patients with a moderate X-linked combined immunodeficiency (XCID) owing to a single missense mutation in the common gamma chain (gammac) gene (L-->Q271) were found to have a progressive T-cell deficiency. Blood T cells from four older subjects with XCIDL-->Q271 were studied to ascertain the basis of that progression. Few CD4+ T cells displayed the phenotype (CD45RA+ CD62L+) or deletion circles from T-cell receptor (TCR) Vbeta-gene rearrangements found in recent thymic emigrants. These deficiencies were more severe in older males with XCIDL-->Q271. Relative frequencies of fresh CD4+ and CD8+ T cells that bound annexin V, an early indicator of programmed cell death, or propidium iodide, an indicator of cell necrosis, were greater in XCIDL-->Q271 T cells than in normal fresh T cells. The binding of annexin V and propidium iodide to XCIDL-Q271 T cells increased marginally after stimulation with anti-CD3, but binding by fresh or stimulated XCIDL-Q271 T cells exceeded that found in normal stimulated T cells. Also, telomeres from XCIDL-->Q271 CD4+ T cells were shortened in these patients compared to normal young adults. It therefore appears that the thymus is dysfunctional and that mature T cells are not effectively rescued from apoptosis or replication senescence via gamma-mediated pathways in XCIDL-->Q271.

Blood gammadelta T cells and gammadelta TCR V gene specificities in a single missense mutation (L-->Q271) in the common gamma chain gene.

The numbers of blood CD4+, CD8+, and CD4-CD8- T cells bearing alphabeta T-cell receptor (TCR) or gammadelta TCR molecules in males with a single missense mutation (L-->Q271) in the common gamma chain gene (gamma(c)) were investigated by flow cytometry. Virtually all XCIDL-->Q271 blood T cells that were CD4+ or CD8+ displayed alphabeta TCR but no gammadelta TCR. In contrast, CD4-CD8- T cells from affected males usually displayed gammadelta TCR, but no alphabeta TCR. The gammadelta TCR specificities were also studied. Except for the oldest subject, there was a direct relationship between blood CD3+ T cells that displayed gammadelta TCR and Vgamma9 and Vdelta2a specificities. Relative frequencies of CD3+ blood T cells that were Vgamma9+ or Vdelta2a+ were inversely related to age. In the oldest patient, the only detected gammadelta TCR specificity was Vdelta1. Thus, in contrast to mice with no gamma(c), XCIDL-Q271 blood T cells that bear gammadelta TCR with Vgamma9/Vdelta2a specificities develop but then decline in late childhood and thereafter. TCR with the Vdelta1 specificity then appear in older survivors with XCIDL-->Q271.

Immunodeficiency due to a unique protracted developmental delay in the B-cell lineage.

A unique immune deficiency in a 24-month-old male characterized by a transient but protracted developmental delay in the B-cell lineage is reported. Significant deficiencies in the number of B cells in the blood, the concentrations of immunoglobulins in the serum, and the titers of antibodies to T-dependent and T-independent antigens resolved spontaneously by the age of 39 months in a sequence that duplicated the normal development of the B-cell lineage: blood B cells followed by immunoglobulin M (IgM), IgG, IgA, and specific IgG antibodies to T-independent antigens (pneumococcal polysaccharides). Because of the sequence of recovery, the disorder could have been confused with other defects in humoral immunity, depending on when in the course of disease immunologic studies were conducted. Investigations of X-chromosome polymorphisms suggested that the disorder was not X linked in that the mother appeared to have identical X chromosomes. An autosomal recessive disorder involving a gene that controls B-cell development and maturation seems more likely. In summary, this case appears to be a novel protracted delay in the development of the B-cell lineage, possibly due to an autosomal recessive genetic defect.

Decreased interleukin-10 production by neonatal monocytes and T cells: relationship to decreased production and expression of tumor necrosis factor-alpha and its receptors.

The production of IL-10 by human neonatal blood mononuclear leukocytes (BML) stimulated with lipopolysaccharide (LPS), tumor necrosis factor-alpha (TNF-alpha), antibodies to CD3, or phorbol 12-myristate 13-acetate (PMA) was measured. The production of IL-10 by neonatal BML cultured with LPS or TNF-alpha was approximately 20 and approximately 15%, respectively, of adult BML. The combination of human recombinant TNF-alpha and LPS failed to augment IL-10 production in neonatal BML. The decreased production of IL-10 by neonatal leukocytes was not due to an autocrine feedback mechanism because only low concentrations of IL-10 were found in newborn sera. A connection with TNF-alpha could not be ruled out, because TNF-alpha production by LPS-stimulated newborn BML and the expression of TNF-alpha receptors on newborn monocytes were reduced. Mean +/- SD of concentrations of IL-10 in supernatants from adult and neonatal BML after stimulation with antibodies to human CD3 for 48 or 72 h were 914 +/- 386 and 178 +/- 176 pg/mL, respectively (p < 0.0001). In experiments with enriched populations of neonatal T cells, the addition of PMA failed to augment IL-10 production. This suggested that newborn T cells may be in a different state of activation than adult T cells Thus, IL-10 production in neonatal monocytes and T cells is reduced and this study suggests that the reduction may be secondary in part to regulatory processes involving TNF-alpha and its receptors.

Deficient quantitative expression of CD45 isoforms on CD4+ and CD8+ T cell subpopulations and subsets of CD45RA(low)CD45RO(low) T cells in newborn blood.

Deficiencies in the quantitative expression of CD45RA and CD45RO on CD4+ and CD8+ T cells and in a population of CD45RA(low)CD45RO(low) T cells in blood from term newborn infants were found by flow cytometry. The relative frequencies of CD45RO on CD4+ T cells from adults and newborn infants were 72 and 58%, respectively. However, in newborn infants greater than 70% of T cells expressing CD45RO also expressed CD45RA. In addition, the quantitative expression of CD45RA and CD45RO on newborn T cells was significantly less than that found on adult blood T cells.

Interleukin-10 in human milk.

The concentrations of immunoreactive IL-10 in the aqueous fraction of 20 specimens of human milk obtained during the first 80 h of lactation and stored at -60 degrees C ranged from 66 to 9301 pg/mL (mean +/- SD, 3304 +/- 3127 pg/mL). IL-10 was present also in the lipid layer of milk. Gel filtration revealed that IL-10 was located in a high molecular weight fraction, where certain other cytokines in human milk have been found. In addition, immunoreactive IL-10 in milk increased after treatment with sodium taurocholate. Bioactive IL-10 was demonstrated by the finding that human milk inhibited [3H]thymidine uptake by human blood lymphocytes and that inhibition was partly overcome by concomitant incubation with antibodies to human IL-10. IL-10 mRNA but no protein product was found in cultured human mammary epithelial cells. Some IL-10 was associated with preparations of human milk leukocytes, but the data did not suggest that the cells were producing the cytokine. Bioactive IL-10 in a possible protected compartment suggests that IL-10 in human milk may have immunomodulating, antiinflammatory effects on the alimentary tract of the recipient infant.

Production of interleukin-6 and interleukin-8 by human mammary gland epithelial cells.

The production of transforming growth factor-beta 2 (TGF-beta 2), interleukin-1 beta (IL-1 beta), IL-6, IL-8, and tumor necrosis factor-alpha (TNF-alpha) by spontaneously immortalized human mammary gland epithelial cells of non-malignant origin and the effect of prolactin upon the production of those cytokines were investigated. Cells were cultured on plastic with epithelial growth factor, insulin, and hydrocortisone. Cytokines were quantified by enzyme-immunoassays. The cells produced IL-6 and IL-8, but no detectable TGF-beta 2, IL-1 beta, or TNF-alpha. Although prolactin enhanced the uptake of [3H]thymidine, it did not alter the production of cytokines/interleukins. Because of the marked production of IL-8 by mammary epithelium and a past report of TGF activity in human milk, those agents were quantified in human milk. The mean +/- S.D. concentrations of IL-8 and TGF-beta 2 in human milk obtained in the first 3 days of lactation were 3684 +/- 2910 and 130 +/- 108 pg/ml, respectively. Thus, IL-8 and TGF-beta 2 are normal constituents in human milk, and human mammary gland epithelium may be responsible for producing some of the IL-6 and IL-8 in human milk.

Activated neutrophils and neutrophil activators in human milk: increased expression of CD11b and decreased expression of L-selectin.

Human milk neutrophils and macrophages were examined by flow cytometry to determine whether they displayed phenotypic markers of activation. The markers were CD11b and L-selectin, which are increased or shed, respectively, from the surface of activated neutrophils. Phenotypic features of milk neutrophils and macrophages were similar to blood neutrophils stimulated with fMLP: plasma membrane expression of CD11b was increased and L-selectin was decreased. After blood neutrophils were incubated in acellular milk, their expression of CD11b increased and L-selectin decreased. The activation was not affected by trypsin but was significantly decreased by treating acellular milk with chloroform or ether. Sedimentation studies suggested that particulate fractions of milk were active. Further, the activation was partly blocked by treating target blood neutrophils with cytochalasin B. Thus, human milk neutrophils are activated, and the activation may be due partly to phagocytosis of membranous structures in milk.

Interleukin-6 in human milk.

Interleukin-6 (IL-6) in human milk collected during the first 2 days of lactation was investigated by a competitive radioimmunoassay (RIA) and column chromatography. The concentrations of IL-6 in the aqueous phase of fresh human milk were 151 +/- 89 pg/ml. The concentrations of IL-6 in milk increased after storage at 4 degrees C and decreased after storage at -20 degrees C (P < 0.01). Column chromatography revealed two molecular weight peaks of IL-6 in human milk, the first > or = 100 kDa and the second between 25 and 30 kDa. The 25-30-kDa peak corresponded to known isoforms of human IL-6 and to the elution pattern for 125I-labeled recombinant human IL-6, whereas the higher molecular weight peak may be in keeping with a bound or compartmentalized form of that cytokine. The precise molecular forms of this protein, the compartmentalization of or binding proteins for this cytokine and the in vivo effects of IL-6 in human milk upon the mammary gland or the recipient infant remain to be explored.

Repertoire of V alpha and V beta regions of T cell antigen receptors on CD4+ and CD8+ peripheral blood T cells in a novel X-linked combined immunodeficiency disease.

The repertoire of V regions of alpha/beta+ T cell receptor (TcR) on CD4+ and CD8+ T cells from the peripheral blood of six patients with a novel X-linked combined immunodeficiency was investigated by flow cytometry. The relative frequencies of V region subfamilies V alpha 2a on CD4+ T lymphocytes and V beta 5b and V beta 12a on CD8+ T lymphocytes from the peripheral blood from the affected males were decreased significantly. Also, the relative frequencies of certain other V region subfamilies on CD4+ or CD8+ T cells from the peripheral blood of some patients were either considerably below or above the ranges found in normal subjects. Although there may be other explanations including an extrathymic event, we suggest that the abnormalities in the TcR repertoire of peripheral blood T cells are consistent with a dysregulation of the intrathymic maturation/selection of T cells that leads to deficiencies in T cell populations in the peripheral blood of males with this disease.

Activated and memory T lymphocytes in human milk.

Since activated macrophages and cytokines are found in human milk (HM), a flow cytometry study was conducted to determine whether T cells in HM display phenotypic markers of recent or previous activation. HM was collected during the first 3 d of lactation. The Paint-a-Gate program was used to optimize gating on the lymphocyte population. A mean +/- 1 SD of 4 +/- 3% of total HM leukocytes were lymphocytes and 96 +/- 3% were macrophages and granulocytes (N = 33 subjects). HM lymphocyte populations were further analyzed in five subjects. T cells (CD3+) represented 83 +/- 11% and B cells (CD19+) were 6 +/- 4% of HM lymphocytes. The mean CD4/CD8 ratio of T cells in HM was 0.88 (range 0.40-1.25). This ratio was significantly decreased compared to the peripheral blood (PB) of control adults (P less than 0.02) and postpartum women (P less than 0.02), due mostly to a significant increase in CD8+ CD3+ cells in HM compared to the PB of control adults (P less than 0.002) and postpartum women (P less than 0.05). T cells bearing markers of recent activation were significantly increased in HM compared to the PB of control adults: 85 +/- 7% of CD3+ cells in HM were HLA-DR+ (controls, 10 +/- 4%; P less than 0.001), and 15 +/- 6% of CD3+ cells in HM were IL-2R+ (controls, 6 +/- 2%; P less than 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)

Tumor necrosis factor-alpha in human milk.

We previously demonstrated that certain biologic activities in human milk were partially blocked by antibodies directed against human tumor necrosis factor-alpha (TNF-alpha). In this study, immunochemical methods were used to verify the presence of TNF-alpha in human milk obtained during the first few days of lactation. Gel filtration revealed the presence of TNF-alpha by RIA in molecular weight fractions between 80 and 195 kD. TNF-alpha could not be detected consistently by conventional Western blotting or cytotoxic assays. Although immunoreactive bands were detected by a Western blot-125I protein A technique in TNF-alpha-positive fractions from gel filtration, those bands proved to be nonspecific. TNF-alpha in milk was reliably quantified by the competitive RIA. Those studies revealed that the concentrations of TNF-alpha in milk were 620 +/- 183 pg/mL. Although RNA to TNF-alpha was detected in milk leukocytes by Northern blotting, little TNF-alpha was found in those cells before or after stimulation with N-formyl-l-methionyl-l-leucyl-l-phenylalanine or 4 beta-phorbol-12 beta-myristate-13 alpha-acetate. The origin of this cytokine in human milk remains unclear. Nevertheless, this study suggests that TNF-alpha is present in early human milk in sufficient quantities to exert possible biologic effects upon the mammary gland of the mother or the immune system of the infant.