Dependence receptors display dual signaling that is dependent on ligand availability: in the presence of their trophic ligands, they transduce various signals, whereas in the absence of their ligands, they are not inactive but rather actively trigger apoptosis(10-12). We recently showed that the dependencereceptor Patched triggers apoptosis in the absence of its ligand Sonic Hedgehog through the formation of a caspase-activating complex – i.e., the dependosome- that includes DRAL, TUCAN, and caspase-9 (9). We investigated here the mechanism by which this Ptc- complex triggers caspase-9 activation and we observed the recruitment of Nedd4 in this complex.
Moreover, apoptosis induction by Ptc requires the caspase-cleavage site at Asp1392 5 . However the mechanism by which the withdrawal of trophic ligands from their respective dependence receptors leads to apoptosis is in large part unknown. Previous studies on DCC have suggested that some of these dependence receptors may recruit a caspase-activating complex 13 . This complex comprising the initiator/apical caspase-9 is independent of the two main apoptosis-related-caspase-activating complexes named, the DISC and the apoptosome 14 . Yet, the precise identification of such an apoptosis- induced dependencereceptor complex has been unsuccessful so far.
ALK (anaplastic lymphoma kinase) is a transmembrane receptor tyrosine kinase, initially discovered as part of the NPM-ALK fusion protein, resulting from a chromosomal rearrangement frequently associated with anaplastic large cell lymphomas. The native ALK protein is normally expressed in the developing and, at a weaker level, adult nervous system. We recently demonstrated that ALK is a novel dependencereceptor. As such, in the absence of ligand, the ALK receptor is kinase inactive and its expression results in enhanced apoptosis, whereas kinase activation, due to a ligand or constitutive as in NPM-ALK, decreases apoptosis. Unligated/kinase unactivated ALK receptor facilitates apoptosis via its own cleavage by caspases, a phenomenon allowing the exposure of a proapoptotic juxta- membrane intra-cellular domain. This review summarizes the biological significance of the ALK receptor in cancer and development, in perspective with its dependencereceptor function. The dual function of ALK in the physiology of development is illustrated in the visual system of Drosophila. In this part of the nervous system, ALK in the presence of ligand appears essential for axonal guidance, whereas in the absence of ligand, ALK expression can lead to developmental neuronal apoptosis. ALK is also found expressed in neural crest- derived tumors such as human neuroblastomas or glioblastomas but its role is not fully elucidated. However, an excessive or constitutive ALK tyrosine kinase activation can lead to deregulation of cell proliferation and survival, therefore to human cancers such as lymphomas and inflammatory myofibroblastic tumors. Our observations could have important implications in the therapy of ALK-positive tumors harboring the chimeric or wild type ALK protein.
In multicellular organisms, long-distance communication between cells is typically achieved by secreted ligands that diffuse through the extracellular medium and bind transmembrane receptors on target cells. Signal propagation through the plasma membrane is then achieved by receptor conformational changes upon ligand binding and classically involves modulation of enzymatic activity, interaction with intracellular partners or ion permeability. The classical view that transmembrane receptors only signal when bound to their ligand is now outdated, especially since the emergence of the dependencereceptor concept. Dependence receptors do not form a family per se, but rather regroup a variety of receptors (transmembrane but also nuclear) that share the ability to trigger cell death unless bound to their respective ligand(s) 1 . Dependence receptors therefore display two signaling activities: a “positive” signaling in the presence of a ligand that can modulate various cellular processes (proliferation, differentiation, migration…), and a “negative” signaling in the absence of ligand consisting in the activation of a pro-apoptotic cascade downstream the receptor. To date, more than a dozen dependence receptors have been identified among which, DCC, Unc5A-D, Patched, TrkA/C, PlexinD1 and others 1-7 , all of which play major roles during embryonic development, especially in the nervous system. Dependence receptors are also involved in cancer: since their apoptotic activity may allow to restrain cells within a ligand-rich environment and eliminate those that migrate away (such as metastatic cells), they were proposed to act as conditional tumor suppressors 1 . Accordingly, loss of function of a dependencereceptor or overexpression of a ligand can confer a selective advantage to tumor cells.
We present here data supporting that in the absence of ligand, the ALK receptor kinase is not tyrosine phosphorylated, and actively enhances or triggers apoptosis, whereas kinase activation, induced by a ligand or constitutive as in NPM-ALK, decreases apoptosis. Unligated/ non-activated ALK receptor facilitates apoptosis via its own cleavage by caspases at D1160, a phenomenon allowing the exposure of a proapoptotic juxta-membrane intra- cellular domain (located between residues 1058 and 1160), most likely through the association with proapoptotic effectors. Caspases have been shown to associate to death receptors such as Fas/CD95, TNFR1 or TRAIL via intracellular adaptor proteins through different motifs within the death-like domain superfamily: death domain (DD), death-effector domain (DED) or caspase-recruitment domain (CARD) (18, 36). We searched databases but were unable to find any such domain within the juxta-membrane intra-cytoplasmic sequence (1058-1160) of ALK. Finer mapping allowed us to restrict the proapoptotic domain to a sequence lying between aminoacids 1090 and 1125 of ALK. The direct or indirect interaction of ALK with proapoptotic effector molecules will be the object of further studies. Yet preliminary data show that expression of several Bcl-2 family members is modulated in ALK- expressing Jurkat cells (data not shown).
Recent data from our laboratory shown that, in the absence of its ligand, Ptc interacts with the adaptor protein DRAL/FHL2 to trigger cell death. DRAL was identified by a two-hybrid screen taimed at finding proteins involved in cell death control that would interact with the propaoptotic domain of Ptc (i.e., its 7th intracellular domain). We demonstrated that, in the absence of Shh, Ptc actually recruits a protein complex that includes DRAL, one of the caspase recruitment domain-CARD containing proteins TUCAN or NALP1, and the apical caspase-9 32 . All these proteins were identified as the partners required by Ptc to induce apoptosis both in immortalized cells and during neural tube development in chick embryos. Moreover, Ptc triggers caspase-9 activation and enhances cell death through a caspase-9-dependent mechanism. In view of these findings, we proposed that in the absence of its ligand Shh, the dependencereceptor Ptc serves as the anchor for a caspase-activating complex that includes DRAL and caspase-9.
SUMO-1, the overall SRC-1 form was decreased in the presence of ubiquitin.
Steroid hormone receptors undergo a variety of covalent modifications which can modify their functional properties (phosphorylation, ubiquitinylation, acetylation). In this manu- script, we describe SUMO-1 conjugation to the progesterone receptor at lysine 388. The non-sumoylated PR mutant was still transcriptionally active and the gain of activity observed in this mutant suggested that the sumoylation of the receptor may have a repression effect. Indeed, it has recently been shown that the sumoylation of the PR is implicated in the regulation of its autoinhibition and transrepression activities (48). Sumoylation has different effects on different steroid re- ceptors: the estrogen receptor ␣ is not sumoylated (49) whereas the androgen receptor (AR) is sumoylated in its N-terminal domain in an androgen-enhanced fashion (49). SUMO-1 conju- gation of AR apparently decreases its transcriptional activity as observed after substitution of the sumoylated lysines. On the contrary this substitution has no effect on AR-mediated tran- srepression (49). Ubc9, the conjugating enzyme for SUMO-1, interacts with both AR and GR (49, 67– 68). In transiently transfected cells, coexpression of Ubc9 enhances AR-depend- ent transcription. However, mutated Ubc9 having lost its SUMO-1-ligating activity retains its effect on AR-mediated transactivation (68). Thus, the role of Ubc9 in sumoylation reactions does not explain its binding to AR and its effect on transactivation.
However, some severe problems with these dependence-based operators were pointed out by Herzig et al. . They stem from the fact that dependence is not di- rected: as we shall explain below, in many cases this makes us forget too much. In this article we show how dependence-based update operators can be generalized to take into account the truth value of fluents (here, the propositional variables) on which the update formula depends (and not only the variables themselves). This leads to a better handling of the frame problem and is specifically important when persistent literals are considered: literals that remain true once they become true. The idea is that in order to update by µ one should forget the negations of all those literals µ depends on. Let us take up the second above example where forgetting the variables a and b in a∧b resulted in the trivial belief base ⊤. It is natural to consider that a (positively) depends on the truth of a, but that ¬a does not (positively) depend on the truth of a. Therefore in order to update by a ∨ b we should forget ¬a and ¬b from a ∧ b. As both a and b “survive” the forgetting of ¬a and of ¬b, updating a ∧ b by a ∨ b should result in a ∧ b. In order to overcome these problems we introduce a family of update operators that is based on formula/literal dependence. We explore several possible definitions of de- pendence functions: they may be defined syntactically or semantically, and they may exploit further information such as knowledge about persistent literals and pre-set dependencies. Different notions of formula/literal dependence will lead to different up- date operators. We show that these belief update operators allow to grasp the effects of actions in a way better than the previous dependence-based operators, while avoid- ing a complexity shift. We also evaluate the operators with respect to their logical and computational properties: as to the former, we check the status of the Katsuno- Mendelzon (KM) postulates for update; as to the latter, we identify the complexity of a number of decision problems (including model checking, consistency, and inference), as well as compactability issues. We identify some operators from the family that provide interesting trade-offs, compared to previous belief update operators.
algorithm to return exactly the independence copula (by setting weights of each leaf equal to its volume). Indeed, while splitting, two splits in adjacent leaves are not synchronized: since the optimization routines are independent of each other, it is unlikely that they return the same breakpoint value for a given dimension, meaning that weights will not be transferable between the two zones: in our case, the constraints forced the weights back to independence. We see that the forest tried to correct this behavior, but the result is quite bad. The dependence measures (τ and ρ) are here structurally 0, and every model respected this correctly. Table 3 shows the same results as for the previous model.
In this paper, we consider the bootstrap for factor models where the idiosyncratic errors are correlated in the cross-section. We show that some natural approaches fail in this context because they lead to a singular bootstrap covariance matrix for the estimated factors, inducing a zero bias in the bootstrap distribution of the estimated coe¢ cients. Instead, we propose a solution based on a consistent esti- mator of the covariance matrix of the idiosyncratic errors. We show that if we use a hard threshold estimator, we can obtain bootstrap consistency for inference on the parameters in a factor-augmented regression. It would be interesting to see whether this approach can be generalized to other estimators of the covariance matrix of the idiosyncratic errors. We also think that it would be interesting to extend this approach to more general models of cross-sectional dependence and with time heterogeneity.
dependence functions, conform to the desiderata: they satisfy (U1), (U3), and (U8), and those who are syntax-independent moreover satisfy (U4); none of them satisfies
(U2), and (U5) and (U6) are not always satisfied.
This article calls for some perspectives for further research. First of all, it would be interesting to investigate other logical settings where update operators based on the “forget-then-conjoin” scheme could be defined. A key property for it is the ability of expressing in the language of the logic the result of forgetting some atoms in a formula. Indeed, while this is feasible in many logical settings, it is not always possible (for instance, this cannot be achieved in modal logic S4). On the other hand, in this article we have supposed that dependence functions are given. In future work we plan to investigate how they can be built automatically.
distribution, where in the heterogeneous model
E h Y k (c) i = e D T k γ , k = 1, . . ., n (5.1) with Dk a (q × 1) set of covariates which can be different from the one used for frequency, and γk a (q×1) set of predictors optimised through maximum likelihood estimation. The parameters obtained for the homogeneous case are in Appendix C while the predictors obtained for the heterogeneous model are in Appendix D. Then, to determine which pairs of coverages may have dependent loss amounts, we look at Kendall’s tau and Spearman’s rho (see (Kendall, 1948)), listed in table 5.2, and calculated with the help of the VGAM (Yee et al., 2010) package in R. We determine that there is weak dependence between Accident Benefits and DCPD, tail dependence between Accident Benefits and Liability, and strong dependence between Collision and DCPD, as we can observe in figures 5.2 to 5.4. The line observed near 0.44 for accident benefits in figure 5.2 stems from policy limits and is not abnormal.
Conditional ablation of the bone morphogenetic protein 2 receptor (BMPR2) in the mouse uterus leads to female infertility, primarily due to a defect in stromal decidualization, caused by the disregulation of the Cyclin D3 (Ccnd3), a key mediator of G 2 phase arrest in
decidualizing stromal cells (Chiang 2003). Bmpr2 conditional knock out (cKO) mice exhibit decidual growth restriction and vascularization anomalies (Nagashima, Li et al. 2013). Both, cKO mice for the estrogen receptor-α (Esr1) and progesterone receptor A (Prg) present multiple reproductive abnormalities, including implantation and decidualization defects (Lydon, DeMayo et al. 1995) (Lubahn, Moyer et al. 1993). Androgen receptors also seem to play an important role in the regulation of a distinct set of genes during in vitro
3. Under the assumption of positive or arbitrary dependence of the p-values, we introduce new two-stage adaptive versions of known step-up procedures (namely, of the LSU under positive dependences, and of the family of procedures introduced by Blanchard and Fleuret, 2007, under unspecified dependences). These adaptive versions provably control the FDR and result in an improvement of power over the non-adaptive versions in some situations (namely, when the number of hypotheses rejected in the first stage is large, typically more than 60%). A second goal of this work is to present a review of several existing adaptive step-up procedures with provable FDR control under independence. For this, we present the theoretical FDR control as a consequence of a single general theorem for plug-in procedures, which was first established by Benjamini et al. (2006). Here, we give a short self-contained proof of this result that is of independent interest. The latter is based on some tools introduced earlier (Blanchard and Roquain, 2008; Roquain, 2007), that aim to unify FDR control proofs. Related results and tools also appear independently in Finner et al. (2009); Sarkar (2008b).
Figure 23. Electric field interface response E-IR dependence on the thickness of the contaminated layer h for: a) PM-
Model and b) Al-Model (meaning Perrier-Morat model and All`egre model respectively, see Table 1). In the former, S(S w ) is
a monotonic function of water saturation, in the latter a non monotonic one. Notice that for both models the responses related
to the absence of contamination (h = 0 m) and h = 0.1 m are coincident, as expected.
When actin assembly from the micropatterned area reached a steady state, we added 2- mM Alexa-488-labeled yeast ADF/co- filin to the reaction mixture ( Figure 1 D). Although ADF/cofilin in- teracted with all actin networks, it bound more sparsely on the Arp2/3 networks than on the different populations of cables ( Fig- ures 1 D and S1 A). Binding of ADF/cofilin was correlated with a decrease of the actin fluorescence signal, corresponding to actin disassembly ( Figure S1 B). The level of disassembly was archi- tecture dependent. Indeed, the actin fluorescence signal after 450 s had decreased more on branched networks than on paral- lel cables or on cables of mixed polarity ( Figures 1 D, S1 B, and S1C). Experiments performed at different concentrations of ADF/cofilin revealed that even for equivalent binding densities of ADF/cofilin, branched actin networks disassembled better than all the actin cables ( Figure S1 C). Together, these results indicate a strong dependence of the actin architectures in ADF/cofilin-driven actin network disassembly. ADF/cofilin bind- ing is able to destabilize branched networks but is not sufficient to disassemble bare parallel or antiparallel bundles.
measure of external …nance dependence used by Klingebiel, Kroszner and Laeven (2002). They recompute Rajan and Zingales measure for 3 digit ISIC level. Using concordance between ISIC and hs6, we are able to include this variable into the set of our regressors. Given that hs6 product classi…cation is more disaggregated than the ISIC one, several hs6 product categories share the same value for external …nance dependence. The data on legal origin used in the …rst step of the instrumental regression procedure, comes from La Porta et al.(1998).