Fig 1.
The TNFR1 signal transduction pathway.
Upon engagement of TNFR1, complex I is rapidly formed and mediates the signaling to NF-κB activation. The ubiquitination mediated by E3 ligases, like cellular inhibitor of apoptosis protein 1 (cIAP1) or cellular inhibitor of apoptosis protein 2 (cIAP2) and linear ubiquitin chain assembly complex (LUBAC), promotes the association of complex I. The Ub modification is required for full activation of the inhibitor of NF-κB (IκB) and subsequent NF-κB activation. Activated NF-κB in the nucleus initiates the expression of target genes like IκB, A20, cellular FLICE-inhibitory protein (cFLIPL), B-cell lymphoma 2 (BCL-2), and X-linked inhibitor of apoptosis protein (XIAP). A20 is a deubiquitinating enzyme (DUB), which is reported to cleave lysine 63 (K63) chains while protecting methionine 1 (M1) chains from cleavage. The deubiquitination by CYLD (cylindromatosis) destabilizes the complex and promotes the formation of complex II in the cytosol. Complex IIa associates caspase 8 (CASP8), while complex IIb additionally binds RIP1. cFLIPL reduces, but does not fully inhibit, caspase activity, which leads to RIP1 and RIP3 cleavage and inhibits apoptosis and necroptosis. cFLIPS fully inhibits caspase activity and promotes the formation of the necrosome. Autophosphorylation of RIP3 allows the recruitment and phosphorylation of MLKL, which subsequently forms active oligomers and translocates to the plasma membrane to induce necroptosis.
Fig 2.
The PN model of TNFR1 signal transduction.
The PN consists of 118 places drawn as white or gray circles, 130 transitions drawn as black squares, and 299 directed edges. Logical places were colored gray, describing vertices with equal names that represent one vertex in the underlying data structure of the PN. The essential processes of NF-κB activation, apoptosis, and necroptosis were shaded blue, green, and red, respectively. The initial marking was represented by one token (black dot) assigned to the places IκB_g, A20_g, XIAP_g, cFLIP_g, and BCL-2_g (g stands for gene) for each place invariant (PI).
Table 1.
The place invariants with their places and the biological meaning.
Fig 3.
Model of the regulation of NF-κB activity.
A part of the PN model in Fig 2 was colored black and blue. Places of the place invariant PI1, IkB_g and NF-kB:IkB_g, were colored blue. White circles represent places and black squares transitions. A dot on a place illustrates a token, here on place IkB_g. The token may move to place NF-kB_n:IkB_g and back but will never vanish. The invariant PI1 reflects the conversation of the gene for the transcription of IkB.
Fig 4.
The TI2-induced subnetwork was highlighted green in the PN model of Fig 2. The subnetwork covers the formation of complex IIb (place CIIb indicated by a red arrow) and induction of apoptosis (transition Apoptosis indicated by a red arrow) via the activation of CASP3 (places CASP3 as logical places indicated by a red arrows) in the extrinsic pathway. Additional logical places that connected graphical subnetworks were Procaspase3 and TRADD:RIP1, each marked by a red arrow.
Fig 5.
The MI7-induced subnetwork of the PN model in Fig 2.
The MI7-induced subnetwork consists of TI9 in blue, TI15 in red, and TI18 in green.
Fig 6.
Pathway classification of MIs.
Pie chart of the classification of MIs according to survival, apoptosis, and necroptosis pathways. Ambiguous MIs were neglected in the chart. In total, 214 MIs were taken into account.
Fig 7.
Complete knockout matrix the PN of TNFR1 signal transduction in Fig 2.
The color-coded effects of the knockout of all syntheses of proteins (31 transitions) are displayed for all places of the PN except for the places belonging to a PI (108 places). Red denotes affected and gray non-affected places.
Fig 8.
Ranking of proteins of the TNFR1 signaling pathway.
The bar chart displays the percentage of affected MIs for the knocked out proteins, see Fig 7.
Fig 9.
The places in the PN were clustered based on the matrix in Fig 7. The hierarchical clustering was performed, using the software tool NOVA [65] with UPGMA (Unweighted Pair Group Method with Arithmetic mean) [66] with Pearson correlation distance [67]. Each leaf of the tree is a protein. Some nodes in the cluster tree were marked blue, green, and red, referring to processes of NF-κB activation, apoptosis induction, and necroptosis induction, respectively.
Fig 10.
In silico knockout submatrix of the PN in Fig 2.
The rows list the proteins, which were knocked out, and the columns give the protein complexes in the network, which could be affected by the knockout. A red (gray) entry indicates that the respective complex was (was not) affected. We performed a single knockout analysis for twenty proteins and displayed the effect for a part of 21 pathway entities. The last two rows represent multiple knockouts and display the effect of SMAC mimetic, i.e., the knockout of XIAP and cIAP, and the impairment of the translation of upregulated genes by cycloheximide, i.e., the knockout of IκB, A20, XIAP, cFLIPL, and BCL-2.
Fig 11.
PN example for firing of a P/T net.
A) The PN consists of five places depicted by circles, two transitions depicted by squares, and six directed edges. The edge from p1 to t1 has a weight of two. For all other edges with a weight equals one, no label is drawn. p3 and t2 are connected via a read edge (read arc), which is bidirectional. Tokens are depicted as dots on the places p1, p2, and p4, defining the initial marking m0 = (2, 1, 0, 1, 0). In this marking, transition t1 is activated. B) The PN after the firing of t1. The marking has changed by removing tokens from the pre-places, p1 and p2, and producing a token on the post-place, p3. Then, the new marking is m‘ = (0, 0, 1, 1, 0).