Insights from human NF-κB knockouts

Abstract

The well-studied NF-κB signaling system is a key mediator of the inflammatory response. Large-scale sequencing studies in humans now allow initial insights into non-essential human genes in which both alleles carry mutations that prevent protein expression or function. Here, we compiled the non-essential genes identified in various sequencing studies and analyzed the occurrence of knockouts in the human NF-κB signaling system. This revealed a lower knockout frequency in the NF-κB system compared to the entire genome. Since drugs inhibiting NF-κB pathway components were unsuccessful in clinical trials so far, the naturally occurring knockouts of NF-κB and its upstream regulators could provide new candidates for therapeutic intervention. To investigate the potential functional importance of posttranslational modifications (PTMs) occurring on NF-κB components, we analyzed not only their evolutionary conservation but also, as a second criterion, their genetic constraint in the sequenced individuals. This approach revealed the absence of missense mutations at key modification sites involved in NF-κB activation and identified additional candidate sites for future studies.

Keywords: Gene Essentiality; Knockout; Loss-of-function; NF-κB; Signaling.

Figure 1. Summary of knockouts identified in the different sequencing studies.

(A) Knockout genes identified in the sequencing studies listed in Table 1 were compared with each other. The number of knockout genes identified in only one study (blue), two studies (light yellow), or three or more studies (yellow) is shown. (B) Venn diagram displaying the overlap of knockout genes in the five largest sequencing studies.

Figure 2. Essentiality scores of the core and upstream NF-κB pathways.

Based on the S_het value as modeled by Cassa et al (2017), we utilized S_het annotations from Sun et al (2024) to assess gene-specific constraints. Specifically, we mapped S_het scores to individual genes from defined gene lists (Dataset EV2) and to gene sets from MSigDB (R version 7.5.1), including Splicing (GO:0048024) and Olfactory Signaling (hsa04740). The analysis shows the subgroups forming the NF-κB core system and the various activation pathways, as well as the overall group derived from them, referred to as “All”.

Figure 3. Schematic representation of regulatory proteins contributing to the up- or downregulation of different NF-κB pathways.

Here we only considered proteinous regulators of the NF-κB pathways in healthy cells. Excluded from the analysis were: The membrane receptors with their directly associated cofactors, as they also feed into other pathways. Factors that were identified only based on pharmacological inhibitors or through large screens without confirmatory follow-up experiments. Cofactors of (post)transcriptional processes, as they lack of specificity for NF-κB. The schematic representation of the complexes was created based on literature data compiled in Dataset EV2, which also provides the references and respective gene names. In most cases, the stoichiometry and details of the protein–protein interactions are not known. The genes encoding the proteins highlighted in red have been described as human knockout genes. The canonical (A) and non-canonical (B) NF-κB pathways are shown.

Figure 4. Schematic representation of regulatory proteins contributing to the up- or downregulation of different NF-κB pathways.

NF-κB activation pathways triggered by antigen receptors (A) and DNA damage (B) are shown. The criteria for NF-κB-regulating proteins correspond to those in Fig. 3, the proteins highlighted in red have been described as human knockout genes.

Figure 5. Analysis of the genetic constraint for the amino acids modified by PTMs in the NF-κB core system.

The PTMs occurring in the NF-κB core system were downloaded from the PhosphoSitePlus database and analyzed for the occurrence of mutations as well as their cumulative allele frequency (as a measure for the general mutation rate) from the gnomAD database. The code to generate the figures is available on Zenodo and can be used for any protein. The IκBα and IκBβ proteins are shown as examples, the ankyrin-repeat region is shown in blue, the PEST domain in yellow. The upper part illustrates the frequency of missense mutations, while the post-translationally modified amino acids listed in the PhosphoSitePlus database are indicated by short lines below. Amino acids under constraint (i.e., not altered by missense mutations) and conserved in lemurs were identified and are displayed as red lines. These analyses were performed for all NF-κB proteins and are given in Dataset EV5.