Tumour class prediction and discovery by microarray-based DNA methylation analysis

Aberrant DNA methylation of CpG sites is among the earliest and most frequent alterations in cancer. Several studies suggest that aberrant methylation occurs in a tumour type-specific manner. However, large-scale analysis of candidate genes has so far been hampered by the lack of high throughput assays for methylation detection. We have developed the first microarray-based technique which allows genome-wide assessment of selected CpG dinucleotides as well as quantification of methylation at each site. Several hundred CpG sites were screened in 76 samples from four different human tumour types and corresponding healthy controls. Discriminative CpG dinucleotides were identified for different tissue type distinctions and used to predict the tumour class of as yet unknown samples with high accuracy using machine learning techniques. Some CpG dinucleotides correlate with progression to malignancy, whereas others are methylated in a tissue-specific manner independent of malignancy. Our results demonstrate that genome-wide analysis of methylation patterns combined with supervised and unsupervised machine learning techniques constitute a powerful novel tool to classify human cancers.     

Production of matrix metalloproteinase-9 in lipopolysaccharide-stimulated human amnion occurs through an autocrine and paracrine proinflammatory cytokine-dependent system

The objective of this study was to determine the presence of autocrine/paracrine regulation of matrix metalloproteinase-9 (MMP-9) expression mediated by proinflammatory cytokines in human fetal membranes. Fetal membranes obtained from women who underwent cesarean delivery before labor were manually separated into amnion and chorion layers and maintained in culture. These explants were stimulated with tumor necrosis factor alpha (TNFalpha), interleukin-1beta (IL-1beta), and either lipopolysaccharide (LPS) alone or LPS with anti-TNFalpha or anti-IL-1beta-neutralizing antibodies. Levels of proMMP-9 in culture media were evaluated by zymography. Enzyme-linked immunosorbant assay was performed to measure the quantity of IL-1beta, TNFalpha, and tissue inhibitor of matrix metalloproteinases-1 (TIMP-1) after LPS stimulation. ProMMP-9 activity was upregulated after stimulation of the amnion by LPS, TNFalpha, and IL-1beta. The increased activity of proMMP-9 resulting from LPS stimulation in the amnion was blocked by the addition of TNFalpha neutralizing antibody but not with anti-IL-1beta. No significant effect of LPS, TNFalpha, or IL-1beta on proMMP-9 expression was observed in the chorion; however, the chorion produced both cytokines when stimulated with LPS. In contrast, TIMP-1 levels remained unchanged in all cultures incubated in the presence of LPS. Therefore, these data indicate that proMMP-9 is produced by the amnion but not the chorion in response to LPS. Because anti-TNFalpha-neutralizing antibody inhibits proMMP-9 activity in the amnion, TNFalpha appears to upregulate proMMP-9 production by the amnion in an autocrine fashion. Meanwhile, TNFalpha and IL-1beta produced by the chorion may upregulate amnionic proMMP-9 production in a paracrine manner.     

Control of receptor-induced signaling complex formation by the kinetics of ligand/receptor interaction

Tumor necrosis factor (TNF) exists both as a membrane-integrated type II precursor protein and a soluble cytokine that have different bioactivities on TNFR2 (CD120b) but not on TNFR1 (CD120a). To identify the molecular basis of this disparity, we have investigated receptor chimeras comprising the cytoplasmic part of Fas (CD95) and the extracellular domains of the two TNF receptors. The membrane form of TNF, but not its soluble form, was capable of inducing apoptosis as well as activation of c-Jun N-terminal kinase and NF-kappaB via the TNFR2-derived chimera. In contrast, the TNFR1-Fas chimera displayed strong responsiveness to both TNF forms. This pattern of responsiveness is identical to that of wild type TNF receptors, demonstrating that the underlying mechanisms are independent of the particular type of the intracellular signaling machinery and rather are controlled upstream of the intracellular domain. We further demonstrate that the signaling strength induced by a given ligand/receptor interaction is regulated at the level of adaptor protein recruitment, as shown for FADD, caspase-8, and TRAF2. Since both incidents, strong signaling and robust adapter protein recruitment, are paralleled by a high stability of individual ligand-receptor complexes, we propose that half-lives of individual ligand-receptor complexes control signaling at the level of adaptor protein recruitment.     

Scale-up of stirring as foam disruption (SAFD) to industrial scale

Foam disruption by agitation—the stirring as foam disruption (SAFD) technique—was scaled up to pilot and production scale using Rushton turbines and an up-pumping hydrofoil impeller, the Scaba 3SHP1. The dominating mechanism behind SAFD—foam entrainment—was also demonstrated at production scale. The mechanistic model for SAFD defines a fictitious liquid velocity generated by the (upper) impeller near the dispersion surface, which is correlated with complete foam disruption. This model proved to be scalable, thus enabling the model to be used for the design of SAFD applications. Axial upward pumping impellers appeared to be more effective with respect to SAFD than Rushton turbines, as demonstrated by retrofitting a 12,000 l bioreactor, i.e. the triple Rushton configuration was compared with a mixed impeller configuration from Scaba with a 20% lower ungassed power draw. The retrofitted impeller configuration allowed 10% more broth without risking excessive foaming. In this way a substantial increase in the volumetric productivity of the bioreactor was achieved. Design recommendations for the application of SAFD are given in this paper. Using these recommendations for the design of a 30,000 l scale bioreactor, almost foamless Escherichia coli fermentations were realised. Electronic Publication     

Enantioseparation of racemic 4-aryl-3,4-dihydro-2(1H)-pyrimidones on chiral stationary phases based on 3,5-dimethylanilides of N-(4-alkylamino-3,5-dinitro)benzoyl L-alpha-amino acids

Three novel chiral packing materials for high-performance liquid chromatography were prepared by covalently binding of (2S)-N-(3,5-dimethylphenyl)-2-[(4-chloro-3,5-dinitrophenyl)carbonylamino]propan-amide (7), (2S)-N-(3,5-dimethylphenyl)-2-[(4-chloro-3,5-dinitrophenyl)carbonylamino]-4-methylpentanamide (8), and (2S)-N-(3,5-dimethylphenyl)-2-[(4-chloro-3,5-dinitrophenyl)carbonyl-amino]-2-phenylacetamide (9) to aminopropyl silica. The resulting chiral stationary phases (CSPs 1-3) proved effective for the resolution of racemic 4-aryl-3,4-dihydro-2(1H)-pyrimidone derivatives (TR 1-14). The mechanism of their enantioselection, supported by the elution order of (S)-TR 13 and (R)-TR 13 and molecular modeling of the complex of the slower running (S)-TR 13 with CSP 1 is discussed.     

Measles Virus DNA Vaccination: Antibody Isotype Is Determined by the Method of Immunization and by the Nature of both the Antigen and the Coimmunized Antigen

Plasmids encoding the measles virus hemagglutinin (HA) and nucleoprotein (NP) proteins inoculated into the skin of BALB/c mice by the gene gun method induced both humoral and cytotoxic lymphocyte class I-restrict- ed immune responses. Although intramuscular immunization induces the immunoglobulin G2a (IgG2a) antibody isotype for both antigens, with gene gun immunization, the NP still generated mainly IgG2a and the major isotype induced by the HA was IgG1. Interestingly, gene gun coimmunization of HA and NP plasmids resulted in a dominant IgG1 HA response and the switching of antibodies generated against the NP to the IgG1 isotype.The initial studies showing that injection of DNA into muscle induces an immune response to the encoded protein opened a new approach to vaccination (for reviews, see references 19 and 22). Recent studies suggest that inoculated muscle cells probably act only as a source of antigen and that immune priming takes place elsewhere in the body (14). For example, excision of an injected muscle a few minutes after DNA inoculation did not affect antibody and cytotoxic T-lymphocyte (CTL) responses (21). Thus, it may be interesting to examine other DNA delivery systems to study how the immune system responds to DNA vaccination. One alternative system involves precipitating DNA onto gold beads which are then propelled into the skin by means of pressurized helium gas (12). When such a system is used, less DNA is required, but unlike the case with intramuscular inoculations, the response is Th2-like, generating immunoglobulin G1 (IgG1) antibodies (17). More recent observations suggest that this is probably due to the mode of inoculation rather than the route (10). We have been studying DNA vaccination against the paramyxovirus measles virus (MV). This disease is one of the primary causes of infant mortality in developing countries, and there is an urgent need for an effective vaccine in infants, as the present live attenuated vaccine is inefficient in the presence of maternal antibodies. Our previous studies established that in a mouse model at least three MV proteins play a role in protection (23). Both glycoproteins, hemagglutinin (HA) and fusion, induce neutralizing antibodies (9, 11), and HA and nucleoprotein (NP) induce CTLs (3, 4), which do not protect against infection but help in recovery (5). In our previous study on DNA vaccination, we showed that intramuscular inoculation of DNAs coding for the MV HA and NP (pV1J-HA and pV1J-NP [6]) induced class I-restricted CTLs and a humoral response corresponding to a Th1 response (6). In the present study, we have extended our observations to compare the same plasmids’ ability to induce an immune response when they are delivered into the skin by a gene gun (Bio-Rad, Ivry sur Seine, France). Gold beads were coated with DNA as follows: approximately 30 mg of gold powder (1.0-μm gold beads; Bio-Rad) was mixed with 100 μl of 0.1 M spermidine (Sigma, L’Isle D’Abeau, France). After sonication, 0.5, 2, or 5 μg of plasmid DNA was added per mg of gold powder, and then 200 μl of 2.5 M CaCl₂ was added to the mixture, with gentle vortexing. Pellets were washed three times and suspended in cold 100% ethanol. Tubes containing dried DNA-coated gold beads were stored at 4°C.

Immune response to MV HA DNA.

Six- to eight-week-old female BALB/c mice (Iffa-Credo, Domaine des Oncins, France) were immunized via the shaved abdominal epidermis one to three times at 21-day intervals with 0.5, 2, or 5 μg of pV1J-HA DNA/mg of gold beads. Two gene gun inoculations (each containing 0.5 mg of gold beads) were given for each dose. The antibody levels measured by enzyme-linked immunosorbent assay, as previously described (6), reached a plateau after two inoculations and did not significantly increase with a third inoculation (result not shown).Our previous studies with intramuscular inoculation established that pV1J-HA induced IgG2a antibodies which are associated with a Th1-type response. When we studied the antibody isotype induced in BALB/c by the gene gun immunization, we observed that it was mainly IgG1 (Fig. ​(Fig.1).1). These data are similar to those described for influenza hemagglutinin by Feltquate et al. (10). The antibody isotype did not vary with time after immunization, number of immunizations, or the amount of plasmid used (data not shown) and was not influenced by genetic background, as pV1J-HA-immunized DBA/2 (H-2^d), C3H (H-2^k), and C57/Black (H-2^b) mice induced mainly the IgG1 isotype (Fig. ​(Fig.11).Open in a separate windowFIG. 1Anti-MV HA isotype of antibodies induced in BALB/c, DBA/2 (H-2^d), C3H (H-2^k), and C57/Black (H-2^b) mice immunized with 0.5, 2, or 5 μg of pV1J-HA by epidermal gene gun. Sera were collected 3 weeks after the immunization. Sera from mice immunized with a control pV1J had means ± standard deviations of 158 ± 198 ng/ml for IgG1 anti-HA antibodies (n = 11) and 10 ± 18 ng/ml for IgG2a anti-HA antibodies (n = 11). Data represent individual animals. To study CTL activity, spleen cells from the immunized mice were stimulated in vitro and analyzed in a cytolytic assay as previously described (6). Despite the apparent Th2-type response, good memory CTL responses were obtained with all protocols used, even when responses were measured just 8 days after a single immunization (Fig. ​(Fig.2),2), and persisted for several months. Open in a separate windowFIG. 2Anti-MV HA and NP CTL response after immunization with pV1J-HA or -NP, respectively. BALB/c mice were immunized with 0.5 (circle), 2 (triangle), or 5 (square) μg of pV1J-HA by epidermal gene gun one (a, d), two (b, e), or three (c, f) times at 3-week intervals. The spleen cells were removed 3 weeks (continuous line) or 8 days (dotted line) after the last immunization. After in vitro stimulation with P815-HA or -NP cells, respectively, lysis was measured on P815-HA or -NP cells, and P815 cells were used as a negative control. The results show the specific lysis of targets at graded effector/target ratios. Each curve represents an individual animal.

Immune response to MV NP DNA.

BALB/c mice were immunized with pV1J-NP with the gene gun and a similar schedule of immunizations. The antibody response with the different number of doses and different plasmid concentrations was similar to that observed for HA, i.e., increased levels after one boost. Similar antibody levels were induced in the range of 0.5 to 5 μg of DNA (data not shown). As was previously shown by intramuscular inoculation (6), the antibody isotype induced was mainly IgG2a (Fig. ​(Fig.3),3), in contrast to the HA results. One explanation for this could be that as the NP is a cytosolic protein and the HA is membrane bound, the potential processing and presentation of the two proteins may be different. However, the same argument would be valid for intramuscular inoculation. Furthermore, it has been reported that gene gun immunization with influenza NP induces a Th2 response (17), so clearly the directed differentiation of T cells is more complicated than a simple distinction between cytosol and membrane-bound proteins. The two methods of immunization (intramuscular versus gene gun) target different cell types, possibly influencing the T-cell response. Furthermore, 9 weeks after immunization, one-third of the 18 mice analyzed showed increased levels of anti-NP IgG1 over IgG2a, regardless of the quantity of DNA injected or the number of inoculations (data not shown). CTL responses were also high, even after a single inoculation (Fig. ​(Fig.2).2). Open in a separate windowFIG. 3Anti-MV NP antibody response as measured by enzyme-linked immunosorbent assay in BALB/c mice immunized with 5 μg of pV1J-NP by epidermal gene gun. Sera were collected 3 weeks after immunization. Each pair of bars represents an individual animal. Sera from mice immunized with a control pV1J had means ± standard deviations of 13 ± 45 ng/ml of IgG1 anti-NP antibodies (n = 11) and 83 ± 276 ng/ml of IgG2a anti-NP antibodies (n = 11).

Coimmunization of HA and NP DNA.

Our results show that when injected by the gene gun, the different MV proteins induce different antibody isotypes. This phenomenon has been suggested to parallel the induction of Th1 and Th2 pathways (1). The pathway taken has been shown to be influenced by the induction of certain cytokines. To determine if coimmunization of these two plasmids would influence the isotype of the antibody response, BALB/c mice were immunized with a mixture of pV1J-HA and pV1J-NP in ratios of 1:1, 4:1, or 1:4 while the total amount of DNA was kept constant (5 μg).Measurement of the anti-HA isotype antibody in mice vaccinated with the different mixtures showed it to be mainly IgG1, similar to that for HA alone (data not shown). In contrast, the anti-NP antibodies switched from the IgG2a to the IgG1 isotype after coimmunization (Fig. ​(Fig.4).4). The proximity of expression of the two antigens was not important in this switching effect, as when pV1J-HA and -NP were inoculated separately in different areas of the skin, the antibody response induced 3 weeks later was the same as that induced when the mixture was inoculated (Fig. ​(Fig.4).4). When analyzed 6 weeks later, only one of six mice showed IgG2a predominance. Open in a separate windowFIG. 4Relationship between the isotype of anti-NP antibodies in sera from mice immunized with 5 μg of pV1J-NP or mixtures of pV1J-HA and pV1J-NP at ratios of 1:1, 4:1, and 1:4 so that the total quantity of DNA/mg gold beads was 5 μg, or pV1J-HA and pV1J-NP injected in different skin area. BALB/c mice were immunized by epidermal gene gun. Sera were collected 3 weeks after immunization. Data are results for individual animals.Cytokines have been used to direct the immune response in several studies. Expression of interleukin-12 either alone or with immunizing antigens can increase protection against microbial pathogens (2) or tumors in animal models (13, 18), in parallel with a Th1 response. Expression or addition of interleukin-4 with the immunogen induces a Th2 response (16, 20). The local concentrations of the cytokines in the initial priming of the immune response are probably critical, as once the T cells have been committed, they cannot be modified. Although some studies have suggested the possibility of Th1 and Th2 switching, a more recent study has shown that once differentiated, T cells cannot switch (15). In agreement with this, Feltquate et al. (10) have shown that initial immunization establishes the Th-cell type of the immune response and that this is not modified by subsequent alternative methods of immunization.Acute viral infections induce a Th1 response, whereas soluble proteins favor a Th2 response (7). When tetanus toxoid was administered 1 day after viral infection, the response to this soluble protein changed from Th2 to Th1 (8). Presumably, this change is due to the domination by the cytokines induced by the viral infection of those produced by the tetanus toxoid. In our studies, we observed that after the coexpression of MV HA and NP, the HA-induced Th2 response was dominant. These observations will obviously have an impact on DNA vaccination, as DNAs coding for several pathogens should ideally be administered concomitantly.     

A Critical Period for Postnatal Adaptive Plasticity in a Model of Motor Axon Miswiring

The correct wiring of neuronal circuits is of crucial importance for precise neuromuscular functionality. Therefore, guidance cues provide tight spatiotemporal control of axon growth and guidance. Mice lacking the guidance cue Semaphorin 3F (Sema3F) display very specific axon wiring deficits of motor neurons in the medial aspect of the lateral motor column (LMCm). While these deficits have been investigated extensively during embryonic development, it remained unclear how Sema3F mutant mice cope with these errors postnatally. We therefore investigated whether these animals provide a suitable model for the exploration of adaptive plasticity in a system of miswired neuronal circuitry. We show that the embryonically developed wiring deficits in Sema3F mutants persist until adulthood. As a consequence, these mutants display impairments in motor coordination that improve during normal postnatal development, but never reach wildtype levels. These improvements in motor coordination were boosted to wildtype levels by housing the animals in an enriched environment starting at birth. In contrast, a delayed start of enriched environment housing, at 4 weeks after birth, did not similarly affect motor performance of Sema3F mutants. These results, which are corroborated by neuroanatomical analyses, suggest a critical period for adaptive plasticity in neuromuscular circuitry. Interestingly, the formation of perineuronal nets, which are known to close the critical period for plastic changes in other systems, was not altered between the different housing groups. However, we found significant changes in the number of excitatory synapses on limb innervating motor neurons. Thus, we propose that during the early postnatal phase, when perineuronal nets have not yet been formed around spinal motor neurons, housing in enriched environment conditions induces adaptive plasticity in the motor system by the formation of additional synaptic contacts, in order to compensate for coordination deficits.