Notch family members follow stringent requirements for intracellular domain dimerization at sequence-paired sites

Jacob J Crow; Allan R Albig

doi:10.1371/journal.pone.0234101

. 2020 Nov 24;15(11):e0234101. doi: 10.1371/journal.pone.0234101

Notch family members follow stringent requirements for intracellular domain dimerization at sequence-paired sites

Jacob J Crow ¹, Allan R Albig ^1,^2,^*

Editor: George Mosialos³

PMCID: PMC7685452 PMID: 33232323

Abstract

Notch signaling is essential for multicellular life, regulating core functions such as cellular identity, differentiation, and fate. These processes require highly sensitive systems to avoid going awry, and one such regulatory mechanism is through Notch intracellular domain dimerization. Select Notch target genes contain sequence-paired sites (SPS); motifs in which two Notch transcriptional activation complexes can bind and interact through Notch’s ankyrin domain, resulting in enhanced transcriptional activation. This mechanism has been mostly studied through Notch1, and to date, the abilities of the other Notch family members have been left unexplored. Through the utilization of minimalized, SPS-driven luciferase assays, we were able to test the functional capacity of Notch dimers. Here we show that the Notch 2 and 3 NICDs also exhibit dimerization-induced signaling, following the same stringent requirements as seen with Notch1. Furthermore, our data suggested that Notch4 may also exhibit dimerization-induced signaling, although the amino acids required for Notch4 NICD dimerization appear to be different than those required for Notch 1, 2, and 3 NICD dimerization. Interestingly, we identified a mechanical difference between canonical and cryptic SPSs, leading to differences in their dimerization-induced regulation. Finally, we profiled the Notch family members’ SPS gap distance preferences and found that they all prefer a 16-nucleotide gap, with little room for variation. In summary, this work highlights the potent and highly specific nature of Notch dimerization and refines the scope of this regulatory function.

Introduction

Notch signaling is a cornerstone of multicellularity and dictates cellular fate and identity. Notch signaling is heavily influenced by microenvironmental cues [1], including adjacent “sending cells” which present any of five Notch ligands to up to four Notch receptors expressed on so called “receiving cells”. Ligand bound and activated Notch receptors undergo a series of proteolytic cleavages which release an active intracellular domain (NICD) [2–4]. This transcriptionally active fragment translocates to the nucleus to act as a co-transcription factor. Common Notch signaling targets are transcription factors themselves, which have their own broader implications and cascades, culminating in a system that requires a fine-tuned, highly sensitive signaling network. Disruption of Notch signaling, both through over- and under-activation, leads to a variety of developmental abnormalities and cancers [5]. Understanding mechanisms behind this precise level of internal control may pave the way for treatments of many of its resulting disorders.

The mammalian Notch signaling system consists of four mostly homologous receptors (Notch1-4) which are all activated through this manner. Each NICD molecule can be readily split into three sections; the N-terminus which contains the RBPJ associated module (RAM) domain, the central ankyrin domain, and a variable C-terminus which houses the Pro-Glu-Ser-Thr (PEST) domain used in protein turnover and in some Notch proteins, a transactivation domain (TAD). Through the RAM domain, all Notch proteins bind to the same transcription factor, recombination signal binding protein for immunoglobulin kappa J region (RBPJ, also commonly called CSL, CBF-1/Suppressor of Hairless/Lag-1) [6]. Upon NICD binding to RBPJ, a new NICD/RBPJ interface is formed which recruits another co-activator, a member of the Mastermind-like (MAML) family [7]. This new tri-protein complex recruits a cascade of other transcriptional machinery to drive transcription of its target genes [8]. While each Notch protein contains the conserved RAM and ankyrin domains, their transcriptional activation profiles are not identical and are largely dependent on context within promoter elements [9].

The DNA target sites which the Notch transcriptional activation complex (NTC) binds to have been the subject of thorough analysis. The consensus binding site was originally defined as a “TP1 element” with the sequence 5’ CGTGGGAAAAT 3’ that recruits RBPJ to Notch responsive promoters [10,11]. TP1 elements are found in a variety of configurations within promoters. Perhaps most importantly, TP1 elements orientated in a head-to-head directionality and separated by 16 base pairs (bp), also known as sequence-paired sites (SPS), enable cooperative binding of two NICD molecules [12,13]. This cooperation results in better repression in the absence of NICD, and enhanced activation in its presence [9,14]. Upon modeling of two N1ICD transcriptional cores on a SPS, it was proposed that complex dimerization occurs through the N1ICD ankyrin domain [15] and this was further supported through crystallization of the interface [16]. Importantly, theses SPS-driven promoters appear to be dimer-dependent. When dimerization was interrupted, N1ICD’s transcriptional potential was substantially reduced on promoters containing SPSs [15,16] and could no longer induce T-cell acute lymphoblastic leukemia [17]. Together, sequence-paired sites and Notch dimerization appear to be potent regulators of Notch signaling and warrant a closer investigation into their mechanics.

In the search for new Notch responsive genes, ChIP-Seq approaches have recently been adopted to identify new SPS sites based on DNA interaction with RBPJ. While the NTC-dimer crystal structure dictates a 16-nucleotide spacer region, ChIP-Seq analysis by Castel, et al. identified a variety of potential SPS-driven genes with spacer regions from 11 to 21 base pairs [18]. These possible targets are opposed by in vitro analysis which observed a more limited spacer region of 15 to 17 base pairs [15,19]. This discrepancy in spacer length is further complicated due to ChIP-Seq approaches that experimentally identified individual RBPJ binding sites then computationally screened for nearby secondary sites [18–20]. Screening for secondary sites however is not straightforward since loading of a NTC onto a high-affinity site directly enables cooperativity on cryptic, low affinity sites which may not even resemble traditional RBPJ binding sequences [16,17,19]. While the transcriptional outcomes seem to be clear, the mechanisms dictating this SPS-response within promoters and enhancers are not clearly understood.

While dimerization-induced signaling of Notch1 has been previously explored, the ability, specifications, and limitations for the other members of the Notch family remain unknown. To compare dimer-dependent signaling of the various NICDs, we generated luciferase reporter constructs driven by either isolated sequence-paired sites from known dimer-dependent promoters or an artificial/optimized SPS site. We observed that all NICDs activate these promoters with varying efficiency. We also observed that Notch1, 2, and 3 functions through dimerization dependent mechanisms, while Notch4 appeared dimer independent. Finally, we compared the optimal gap length within SPS sites and found that all NICD molecules prefer promoters with 16bp between RBPJ binding sites, with little room for deviation. These results should help us to understand how the various NICD molecules interact in cells and potentially diversify Notch signaling outputs in cells that express multiple Notch proteins.

Materials and methods

Cell culture

HEK293T cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) (Mediatech, Inc.) supplemented with 10% EquaFetal Bovine Serum (FBS) (Atlas Biologicals) and 1x penicillin-streptomycin solution (Mediatech, Inc.). Cells were grown in 10 cm culture plates and subcultured at 70–80% confluency.

Expression and reporter plasmids

Protein expression constructs were obtained through the following: FLAG-N1ICD (AddGene #20183), N2ICD (#20184), N3ICD (#20185), and N4ICD (#20186) were all gifted by Raphael Kopan [9] and acquired through AddGene.org. All constructs code for the intracellular domain of the mouse Notch proteins and have a 3xFLAG peptide tag on the N-terminus. The N1ICD-MYC ΔS2184 construct, also a gift from Raphael Kopan [2] (#41730), includes a substantial C-terminal truncation, encoding mouse N1ICD V1744 to S2184 with a MYC tag located at the C-terminus. The NICD coding regions were subcloned into pKH3 (#12555), a gift from Ian Macara [21] to add a C-terminal 3xHA tag. N1ICD (R1974A), N2ICD (R1934A), N3ICD (R1896A), and N4ICD (R1685A) mutants were all created through site-directed mutagenesis of the NICDs based on sequence alignment to identify amino acids (S2 Fig) in mouse NICDs homologous to the human N1CID R1984 site previously shown to be essential for NICD dimerization [16]. The empty coding vector pcDNA3.1/MYC-His was obtained from Invitrogen and pCMV-β-Galactosidase was obtained from Clontech/Takara Bio.

Transcriptional reporter constructs were obtained or created as the following: Full-length mouse promoters for Hes1 (#41723) and Hes5 (#41724) were a gift from Ryoichiro Kageyama and Raphael Kopan [22]. 4xTP1 (#41726), a synthetic promoter containing four high-affinity RBPJ binding sites in tandem, was a gift from Raphael Kopan [23]. These promoter sequences were designed and cloned into pGL2-Basic (Promega), a luciferase reporter plasmid, which upon promoter activation drives expression of firefly luciferase.

Construct creation and mutagenesis

To create luciferase reporters that activate specifically upon Notch dimerization, we isolated the sequence-paired sites from the native mouse and human Hes1 and Hes5 genes and cloned these fragments into the pGL3-Basic vector (Promega) which contains a minimal promoter that is incapable of transcriptional initiation without additional enhancer elements. Our synthetic promoters, the 2xTP1 constructs, were designed using the TP1 response element as originally isolated from the Epstein-Barr virus, which contains two high-affinity binding site for RBPJ [10,11]. TP1’s ‘complete’ RBPJ consensus sequence (5’-CGTGGGAAAAT-3’) and a ubiquitous “core” sequence (5’-GTGGGAA-3’) were taken from the response element, its secondary site reversed and placed in the complementary strand to result in a head-to-head arrangement, and two nucleotides were inserted into TP1’s spacer region to result in a sixteen-base pair, sequence-paired site.

These fragments were synthesized as oligonucleotides (Integrated DNA Technologies, IDT) and were designed to be partially complementary so that when annealed, the ends were left overhanging, matching cuts from the restriction enzymes KpnI and SacI (New England Biolabs, NEB). pGL3-Basic was cut with these two enzymes, dephosphorylated with shrimp alkaline phosphatase (NEB), inserts phosphorylated with T4 Polynucleotide Kinase (NEB), and ligated together with T4 DNA Ligase (NEB).

A variety of sequence-paired site gap distances were created through blunt-end ligation. PCR primers to mutate Hes1, Hes5, and 2xTP1 were designed to align with the desired base excisions or to include base extensions. Base extensions were designed to keep the gap distance nucleotide composition (G/C vs A/T) approximately consistent with the native promoters’. PCR products were phosphorylated, ligated, and reaction template digested with the restriction enzyme DpnI (NEB). A statistical comparison of the modified constructs’ basal activity levels did not indicate the addition or subtraction of any other regulatory elements (data not shown).

All constructs were sequenced verified before experimental use. SPS sequence information can be found in S1 Table.

Western blotting

For western blotting analyses, HEK293T cells were plated into 6-well plates at a density of 300,000 cells/well. The following day, cells were transfected with polyethylenimine MW 25,000 (PEI, Polysciences) at a ratio of 5 μg PEI to 1 μg DNA. Wells were transfected with 1000 ng of plasmid DNA for the various FLAG-NICD constructs, allowed to grow for two days, and cells collected and prepared in 1x SDS-page lysis buffer. Western blotting was performed as described previously [24]. Membranes were incubated with primary antibodies against FLAG tag (Cell Signaling Technology, #14793) or GAPDH (Santa Cruz Biotechnology, sc-25778) and detected through horseradish peroxidase conjugated α-rabbit antibodies (GE Healthcare Life Sciences, NA934V). Experiments were repeated independently three times, where the figure displayed uses the best representative exposures.

Luciferase assays

For all luciferase assays, HEK293T cells were plated into 24 well plates at a density of 50,000 cells/well where the experimental conditions were treated as triplicates or duplicates. The following day, cells were transfected. When analyzing the full-length promoters, 100 ng of luciferase construct and 10 ng of NICD expression plasmid were used, whereas in experiments of sequence-paired site constructs, 200 ng and 100 ng were used, respectively. In all experimental variations, 10 ng of a β-Galactosidase expression plasmid was used per well to normalize data for transfection efficiency and cell growth/death. To equate amounts of DNA between experimental conditions, the empty coding plasmid pcDNA3.1/MYC-His was utilized. Forty-eight hours post transfection, cells were collected and analyzed as previously described [24]. All samples were treated in triplicates, except for the single-base change experiment, which was in duplicates. Independent experiments were performed at least four times.

Statistical analysis

Statistical significance was determined through a student’s two-tailed t test, comparing two-samples with homoscedastic variance. Significance is determined as *** is p ≤ 0.001, ** is p ≤ 0.01, and * is p ≤ 0.05.

Results

Activation of Notch target genes containing sequence-paired sites requires ankyrin-dependent dimerization

Notch target genes often have multiple RBPJ binding sites within their promoter sequences and a fraction of these are orientated in head-to-head, paired sites [18,19]. This arrangement allows for NTC dimerization through NICD ankyrin domains, resulting in potent transcription of SPS containing genes. The canonical Notch target genes Hes1 and Hes5 have previously shown to be activated by Notch1 in a dimer-dependent manner [16,19]. Using luciferase reporter assays, we first sought to confirm if other members of the mammalian Notch family also activate Hes1 and Hes5 in a dimer-dependent manner.

HEK293T cells were co-transfected with commercially available luciferase reporter plasmids containing large fragments of the Hes1 or Hes5 promoters, and either wild-type or dimerization incompetent Notch1–4 NICD expression plasmids to observe dimerization dependence of these promoters. In agreement with previous work, the full-length Hes1 (Fig 1A) and Hes5 (Fig 1B) constructs were significantly activated by wild-type N1ICD and transcriptional activation was significantly less for dimer incompetent N1ICD. Likewise, wild-type N2ICD and N3ICD also displayed activation on both promoters but dimer-dependent activation only on the Hes1 promoter. Indeed, mutant N2ICD and N3ICD dimer incompetent constructs demonstrated a similar ability to activate the Hes5 promoter suggesting the N2ICD and N3ICDs were functioning in a dimer-independent manner on the Hes5 promoter. Finally, N4ICD activated the Hes1 promoter to a similar degree as the other NICDs but was significantly weaker on the Hes5 promoter compared to the other NICDs. Moreover, N4ICD failed to demonstrate dimer dependence on either the Hes1 or Hes5 promoters. In agreement with previous observations [15], western blot analysis of the various NICDs demonstrated that the decreased transcriptional activation of the dimer incompetent ankyrin mutant NICD molecules was not associated with decreased protein expression of these mutant NICDs (Fig 1C). Moreover, mutation of NICD ankyrin domains did not impinge the basal transcriptional activity of NICD molecules since WT and ankyrin mutant NICDs exhibited nearly identical transcriptional activity on the non-dimerizing 4xTP1 promoter (Fig 1D).

Fig 1 — HEK293T cells were transfected with portions of the (A) *Hes1* (-467 to +46) (B) *Hes5* (-800 to +73) promoters for luciferase assays to monitor promoter activation. Cells were co-transfected with wild-type (WT) or ankyrin mutated (R→A) NICDs to compare their dimer dependence. Shown are the average +/- SE of 5 independent experiments. * indicates P < .05, ** < .01, *** < .001 (student’s t-test). (C) Western blots to compare the protein abundancies of WT or dimer incompetent ankyrin mutated NICDs. HEK293T cells were transfected with NICD expression constructs and NICD proteins were detected with anti-Flag antibodies. Endogenous GAPDH was also detected from the same lysates for use as a loading control. The image displayed is a representative blot from an experiment that was replicated three independent times. (D) 293T cells were transfected with the dimer independent 4XTP1 luciferase construct and WT or ankyrin mutated NICDs as above. Shown are the average +/- SE of three independent experiments.

Collectively, these results largely supported the previously reported dimer-dependent nature of the Hes1 and Hes5 promoters. However, we also noted several weaknesses with these experiments. In particular, we noticed significant background with modest activation of the Hes1 promoter and reduced dimer dependence for N2-4ICD on the Hes5 promoter. Based on these weaknesses, we more closely examined these promoter sequences (S1 Fig). We searched for high affinity binding sites that were defined by the RBPJ consensus sequence TGTGGGAA, and low affinity sites that were defined as having up to two nucleotide differences compared to the high-affinity sites. We found a multitude of possible low affinity RBPJ binding sites in addition to the high-affinity SPS sites previously described [9]. Based on this result, we hypothesized that the overall promoter complexity and number of potential low affinity RBPJ binding sites within these promoter sequences might have been responsible for the high background, low relative activation, and dimer independence observed in our experiments. Moreover, beyond Notch signaling, Hes genes are also controlled by a variety of other transcription factor families [25,26], including self-regulation [27,28], suggesting that these relatively large promoter fragments might have been responding to both Notch specific and non-Notch transcriptional activity. Collectively, these observations prompted us to perform further experiments on the Hes1 and Hes5 promoters with the goal of refining these promoters into more specific tools to monitor dimer-dependent NICD activation.

The isolated Hes5 sequence-paired site does not respond to dimerization

Our results in Fig 1 show that the Hes5 promoter demonstrated both stronger activation than the Hes1 promoter and N1ICD dimer dependence. We therefore rationalized that isolation of the Hes5 SPS site should yield a promoter element that would specifically and robustly respond to NICD dimerization. We isolated the mouse and human Hes5 SPS elements containing two head-to-head orientated RBPJ binding sites separated by a 16-nucleotide gap, and the 4 to 5 surrounding nucleotides (Fig 2A) as previously described [16] and cloned these sequences into a luciferase reporter vector containing a minimal promoter that was incapable of transcriptional initiation without additional enhance elements. Interestingly, this cloned region includes one canonical RBPJ binding site and a partnered “cryptic site” which doesn’t match standard RBPJ binding sequences, but instead was hypothesized to form the partner site for the high-affinity site [16]. In addition, there are two nucleotide differences between the mouse and human genes in this region (Fig 2A), one of which is inside the cryptic RBPJ binding site. As shown in Fig 2B, we found that both mouse and human isolated Hes5 SPS promoters were indeed responsive to N1ICD, however, neither of these isolated elements appeared to be dependent on NICD dimerization since they were equivalently activated by WT and dimer-incompetent versions of N1ICD. In addition, the isolated Hes5 SPS responded to Notch activation almost exactly the same as to promoters with RBPJ binding sites in a non-dimerizing head-to-tail orientation (i.e. 2xTP1(H-T) and 4xTP1(H-T) constructs). Moreover, N1ICD activated the Hes5 SPS construct less than the 2xTP1 construct despite these promoters both containing two RBPJ binding sites. Interestingly, N4ICD was unable to activate transcription from the isolated Hes5 SPS sites despite successfully activating the full length Hes5 promoter in Fig 1. In stark contrast to this result, the full-length Hes5 promoter did display dimer dependence to N1ICD (Fig 1B). This difference between the full-length and SPS versions of the Hes5 promoter suggested that the Hes5 promoter might have been functioning differently than previously thought. Given the weakness of the cryptic RBPJ binding site, the presence of several other potential RBPJ binding sites within the full-length promoter, and the dimer dependence of the full-length Hes5 promoter, we hypothesized that the high affinity RBPJ binding site might be making a long-distance interaction with another high-affinity site within the promoter. This possibility would allow cooperation between previously thought independent NTCs within gene promoters and enhancers and perhaps explain the transcriptional differences between full-length and SPS versions of the Hes5 promoter. To test this possibility, we designed a series of Hes5 SPS mutants based on the helical nature of DNA. Since a DNA helix completes approximately one rotation every 10 bp, we hypothesized that an insertion or deletion of five extra base pairs would place the secondary NTC on the opposite side of the DNA, breaking any cooperativity. Further, an addition of an extra five nucleotides may restore cooperativity, despite the NTC being further away, due to a long range NTC dimerization. However, no obvious cooperative binding, or loss thereof, was observed in these constructs (Fig 2C).

Fig 2 — (A) Promoter structure of the Hes5, 2xTP1(H-T), and 4xTP1(H-T) luciferase constructs. Red dots indicate nucleotide differences between the human and mouse Hes5 SPS sites. Suspected RBPJ binding sites are highlighted in yellow and arrows depict directionality of RBPJ binding sites. (B) HEK293T cells were transfected with either the mouse (mHes5(SPS)) or human (hHes5(SPS)) sequence-paired site constructs or with the 2xTP1(H-T) or 4xTP1(H-T) head-to-tail constructs and WT or dimer-incompetent versions of NICDs. Shown is the average +/- SE of n = 4 independent experiments. * <0.05, ** <0.01, *** <0.001, student’s t-test. (C) HEK293T cells were transfected with various NICDs and Hes5 SPS constructs with 11–36 bp spacer regions (5 bp increments), n = 6.

Collectively, these results indicated that the Hes5 SPS does not function in a dimer-dependent manner in cells and instead behaves as a monomeric RBPJ binding site. Due to this unexpected complication, to further explore SPS capabilities we moved on to the Hes1 promoter, which contains a more canonical SPS site.

The Hes1 SPS acts through traditional NICD dimerization mechanisms

Hes1 is another Notch target gene which demonstrates dimer-dependence. Its promoter contains a single canonical RBPJ binding site and secondary site 16 nucleotides away in the reverse orientation (Fig 1A). This secondary site is slightly non-consensus, but only displays a minor decrease in RBPJ affinity compared to the high affinity consensus site [19,29]. Therefore, we took the same reductionist approach used with Hes5 and isolated the SPS element out of the Hes1 promoter.

Since the human and mouse sequences in this promoter region are identical, a generic Hes1(SPS) construct was cloned into a minimal promoter luciferase vector to quantitatively monitor its activation by Notch signaling (Fig 3A, Top). As before, transfections of this construct into HEK293T cells, along with WT or ankyrin mutated NICDs, allowed us to analyze dimer dependence of the Hes1 SPS element. The Hes1(SPS) construct responded to NICD activation (Fig 3A, Bottom) and interestingly, the isolated Hes1 SPS demonstrated approximately 5 times greater sensitivity to N1ICD compared to the full-length Hes1 promoter, most likely due to reduced background of the isolated SPS construct. Both N1ICD and N2ICD ankyrin mutants showed a significant decrease in activation compared to their wild-type counterparts. In comparison, N4ICD was again dimer independent, matching trends observed on the counterpart full-length promoter (Fig 1A). Finally, N3ICD only weakly activated the isolated Hes1 SPS. Indeed, compared to the full-length Hes1 promoter, N3ICD showed no increased activity on the isolated Hes1 SPS. Moreover, N3ICD did not demonstrate dimer dependence on the isolated Hes1 SPS but did demonstrate dimer dependence on the full-length Hes1 promoter (Fig 1A). This is potentially unsurprising since Notch3 has been shown to synergistically utilize other nearby transcription factors for its own signaling responses [9], which would be lacking in this minimalized promoter.

Fig 3 — (**A, Top**) Sequence information for the human and mouse isolated *Hes1* SPS element. (**A, Bottom**) HEK293T cells were transfected with the Hes1(SPS) construct and WT or dimer-incompetent versions of NICD proteins. Shown is the average +/- SE of n = 4 independent experiments. Student’s t-test was performed to compare the activation between WT and mutant NICDs where * <0.05, ** <0.01, *** <0.001. (B) HEK293T cells were transfected with various NICDs and Hes1(SPS) constructs that had been modified to contain variable gap distances between RBPJ binding sites in 5 bp interments. Shown is the average +/- SE of 4 independent experiments.

With a properly responding SPS-driven promoter, we again wanted to determine if long-range interactions were possible between RBPJ binding sites. We again followed the same logic employed when mutating the Hes5 SPS and we created gap distances in steps of five nucleotides to take advantage of the periodicity of the DNA helix. We found that the wild-type, 16-nucleodtide gap, was the only distance all NICDs were capable of cooperatively binding and eliciting activation on (Fig 3B). This observation further limits the options of dimerization dependent signaling since independent NTC sites within this promoter are unlikely to cooperate over long distances by methods of kinking, looping, or untwisting the gap DNA.

Establishment of a high activity, NICD dimer-specific reporter construct

Having found that the SPS element isolated from the Hes1 promoter functions in a dimer-dependent manner, we next sought to optimize an SPS-driven luciferase construct as a tool to specifically study the transcriptional activity of NICD dimers. To accomplish this, we constructed a synthetic SPS site that contained two high-affinity RBPJ binding sites in a head-to-head configuration separated by 16 bp. We utilized the complete TP1 consensus sequence (5’-CGTGGGAAAAT-3’) originally described by Meitinger et al., [10] thus forming the 2xTP1(SPS)-Complete construct. RBPJ shows high affinity towards this site and multiple copies of this RBPJ binding site have previously been arranged in a head-to-tail orientation to measure the activation of Notch signaling [11,30,31]. These sequences however have not previously been orientated in a head-to-head orientation to measure transcriptional activation by dimerized NICD molecules.

As shown in Fig 4A, the 2xTP1(SPS)-Complete construct successfully responded to Notch signaling. Unexpectedly however, dimer-incompetent N1ICD mutants increased activation even better than wild-type proteins indicating this promoter was not dimer-dependent. We hypothesized that since the last three nucleotides in the consensus sequence were not necessary for RBPJ-DNA interactions [10], this effectively resulted in a 22 nucleotide gap distance between RBPJ binding sites, and therefore the promoter lost its dimerization dependence. In response to this possibility, we created another SPS construct, this time with a core, essential sequence for RBPJ responsiveness (5’-GTGGGAA-3’) [10,32] separated by 16 bp. We named this new construct the 2xTP1(SPS)-Core construct. Cloning of this fragment resulted in a ‘T’ on the 5’ side of each core sequence. This addition coincidentally matches the RBPJ consensus sequence found in Hes1/5, though this position within the consensus is variable [33]. As shown in Fig 4B, wild-type N1, N2, and N3ICD all strongly activated this promoter while the corresponding dimer-incompetent ankyrin domain mutants demonstrated significantly reduced transcriptional activity. N4ICD also activated the promoter, but remained dimerization independent, which has persisted across all SPS promoter variations tested. Nonetheless, the 2xTP1(SPS)-Core construct appeared to be an optimized promoter for evaluating dimer-dependent Notch signaling.

Non-optimal SPS sites select against transcriptional activation by NICD dimers

The results in Fig 4A revealed an interesting phenomenon wherein non-dimerizing ankyrin mutant NICDs performed better than their WT counterparts on the 2xTP1(SPS)-Complete construct which had a slightly longer gap than the 16 bp preferred by NICD molecules. This observation suggested that two RBPJ binding sites gapped slightly more or less than 16 bp within a promoter might actually suppress promoter responsiveness to NICD dimer-dependent Notch signaling and favor NICD dimer-independent Notch signaling, a phenomenon which has not been previously described. To test this hypothesis, we compared WT and ankyrin mutant N1ICD transcriptional activation from the Hes1(SPS) and 2xTP1(SPS)-Core constructs which we modified to contain head-to-head RBPJ binding sites separated by 11, 16, or 21 bp. As shown in Fig 5A and 5B, WT N1ICD strongly activated reporter transcription on the 16 bp gap promoters and was significantly weaker on the 11 bp and 21 bp gap promoters. Moreover, the ankyrin mutant N1ICD also performed as expected, showing no synergistic activity across the various gap distances. Importantly however, the ankyrin mutant N1ICD slightly outperformed the WT N1ICD on the 11 bp and 21 bp gap promoters (Fig 5A and 5B right panels). This observation suggested that head-to-head orientated RBPJ binding sites with non-optimal gap widths are more likely to be activated by non-dimerizing NICD molecules. Whether or not there is a condition that actively manipulates NICD dimerization is currently unknown.

Fig 5 — (**A and B)** 293T cells were transfected with 11, 16, or 21 bp gap versions of the (A) Hes1(SPS) or (B) 2xTP1(SPS)-Core reporters and WT or ankyrin mutated N1ICD. Shown is the average +/- SE of n = 6 experiments. In all panels, a student’s t-test was performed to compare the activation between WT and ankyrin mutated NICDs, where P-values are reported as * <0.05, ** <0.01, *** <0.001.

A restrictive spacer range dictates the signaling capabilities of NICD dimerization

Having established a highly active and NICD dimer-specific promoter, we set out to compare the promoter specificity of the various NICD molecules. To accomplish this, we established a series of 2xTP1(SPS)-Core promoters with modified SPS gap distances ranging from 11 to 21 bp in one bp increments. Combinations of each NICD protein coupled with each spacer variation were then transfected into HEK293T cells and assayed for their transcriptional activity. Our rational for this approach was the previous finding that ChIP-Seq screening identified SPS-driven genes with various gap distances, from 11–21 base spacers [18]. Our data in Fig 3 did not support the idea that gap distances greater or less than 16 bp (in 5 bp increments) can support dimer-dependent transcription on the Hes1 promoter. Nonetheless, we wanted to use our optimized promoter to investigate the possibility that smaller variation in gap distances may be tolerated during NICD dimerization, or that different Notch family members may exhibit slightly different preferences in SPS gap width. As shown in Fig 6, all mouse NICDs have a strong preference for the SPS sites with 16 bp gaps, with little room for variation. For N1ICD, there was some flexibility observed with 15 and 17 bp spacers, though anything outside of this range did not demonstrate cooperative signaling (Fig 6A). In contrast, N2ICD showed even more specificity, where there was only slight cooperativity at the 15 bp gap, and none observed at 17 (Fig 6B). N3ICD showed cooperativity at 16 bases (Fig 6C), and N4ICD at 16 and 17 bases (Fig 6D), however it’s difficult to make a judgement call about their flexibility due to their low activation and their apparent dimerization independent signaling. Also, it’s worth noting is that N4ICD preferred a 16 bp gap suggesting that N4ICD was functioning as a dimer which is in conflict with our earlier observations that N4ICD functioned dimer-independently. Moreover, N4ICD demonstrated slight inhibition on promoters with 11 or 14 bp gaps, which was similarly observed in the Hes5 (Fig 2C) and Hes1 (Fig 3B) SPS promoter constructs. This might indicate that N4ICD proteins generally act as transcription inhibitors on monomeric NTC sites, or perhaps occupy RBPJ binding sites and prevent other NICDs from binding, though we did not explore these possibilities. Further we also observed a slight increase in activation of the 12 or 13 bp gap constructs for all four NICD proteins, which could be attributed to some form of cooperativity, though we did not explore this thread any further. Finally, as with the Hes5(SPS) and Hes1(SPS), we extended the gap distances of the 2xTP1(SPS)-Core out to 36 nucleotides and again observe little to no long-range cooperativity.

Fig 6 — (**A-D**) HEK293T cells were co-transfected with various NICD expression plasmids and 2xTP1(SPS)-Core promoter with varying SPS gap lengths from 11–21 bp, in one bp increments, or 21–36 bp, in 5 bp increments. For all luciferase studies, an empty expression vector was used to equate the amount of DNA transfected across samples. Fold differences are represented as the NICD-induced activity compared to the basal activity levels. Shown are the average +/- SE of n = 5 independent experiments. (E) Overlay of data in panels A-D to emphasize different transcriptional strengths of the various NICDs. Statistical analysis was performed by using a one-way analysis of variance (ANOVA), followed by a Tukey-Kramer's HSD post-hoc test. Only points that were statistically different from other points are indicated in each graph where capital letters indicate P values < 0.001, whereas small case letters indicate P < 0.05 on a point-by-point basis.

Discussion

Notch signaling is an important cellular communication mechanism that is required for multicellular organisms. Ongoing research continues to reveal how Notch functions and how Notch signaling is integrated into many facets of cell biology. Despite our growing understanding of Notch function, however, there are basic questions about Notch signaling that remains to be addressed. In this study, we sought to address some of these basic questions about Notch signaling that have been overlooked in the quest to dig into the deeper questions of Notch function. In particular, we felt that a head-to-head transcriptional comparison of the full-length Notch NICD molecules on a variety of promoter elements should be performed. Most of what we know about Notch dimerization, including how the proteins interact and on which promoters, comes from studies with Notch1. While the four mammalian Notch proteins are mostly homologous within the N-terminal and ankyrin domains, there are substantial differences within the C-terminal regions. Notably, this domain is absent in much of the field’s previous characterization work. Indeed, molecular modeling [15] and crystallization [16] were performed with just the ankyrin domain, and these studies laid the original groundwork for the NTC’s spatial conformations, interacting amino acids, and DNA preferences. Similarly, further in vitro work with EMSAs and FRET assays supported NTC cooperative loading and SPS gap preferences, though these were performed using N1ICDs with just the N-terminal RAM and ankyrin domains [15,16,19]. Importantly, the C-terminus has known transcriptional effects [34], regulatory capacity [35], and contains a multimerization site [36]. Therefore, it was important to ensure 1) the presence of these domains don’t interrupt dimerization responses in a cellular context, and 2) if the lessons learned by studying Notch1 can be broadly applied to the other mammalian Notch proteins.

To address these concerns, we first revisited full-length, Notch-responsive promoters, which contain canonical sequence-paired sites to drive transcription. While these promoters did require NICD dimerization for full activation and therefore allowed for preliminary conclusions, our data collection suffered from low activation levels (Hes1) or high experimental variability (Hes5). This is a logical byproduct from using full-length promoters since these genes are regulated through mechanisms beyond just Notch signaling and overexpression of constitutively active NICDs would have wide, overarching, signaling outcomes. For example, as in the case of Hes1, this gene is understood to be self-regulated, resulting in its oscillatory nature [37]. For our purposes, this implies that endogenous Hes1 may be activated by NICD overexpression and circles back to our Hes1-luciferase promoter to negatively-regulate it, resulting in its poor overall activation. For reasons like this, we sought to create minimalized promoters to more specifically monitor dimerization-induced signaling.

Since synthetic promoters with tandemly arranged RBPJ sites had been previously used to monitor Notch activity (e.g. 4xCSL-Luc, [23]), we reasoned that the SPS alone should also be sufficient for Notch activation. Therefore, we isolated the known sequence-paired sites out of Hes1 and Hes5 and created the Hes1/5(SPS)-luciferase vectors. Since the reverse sites within Hes1 and 5 do not perfectly match the canonical RBPJ binding site, we also crafted the synthetic 2xTP1(SPS)-Core promoter to make a “perfect” SPS element. Upon testing we found that the Hes1(SPS) and the 2xTP1(SPS)-Core promoters operated as expected in that they were robustly activated by N1ICD and inefficiently activated by dimer-incompetent NICD ankyrin mutants. Surprisingly however, our results did not indicate that the isolated Hes5 SPS site functions in a dimer-dependent manner, although the full-length Hes5 promoter was dimer-dependent. We return to this point later in this discussion.

Having established two reporters with high activity and specificity for dimerized NICD signaling we were able to specifically ask questions about dimerization of full-length NICD proteins in cells. We found that full-length N1ICD and N2ICD activated sequence-paired sites in a dimerization-dependent mechanism. In contrast, N3ICD’s activation of these minimal promoters was poor, which may be partially attributed to its lack of a TAD domain in its C-terminal [34] or usage of other transcription factors for its activation [9]. It also differed in its dimer-dependence; where it was not dimer-dependent on the Hes1(SPS) but was on the 2xTP1(SPS)-Core. This implies there is some difference located within these two minimalized constructs that can affect dimerization outcomes. The primary source of variation can be found within the 16 bp gap between RBPJ binding sites. The original crystallography studies found that the Hes1 gap distance DNA must go through substantial untwisting to bring the ankyrin domains into contact with each other for dimerization [16]. Gap composition of the 2xTP1(SPS)-Core construct has even higher G/C content than Hes1’s, so while N1ICD and N2ICD can utilize variable gap compositions, perhaps N3ICD utilizes a more inflexible gap sequence. Whether or not this would enable differential transcriptional activity of Notch remains to be tested. In contrast, however, N4ICD did not appear to have dimer-dependent activation on the Hes1(SPS) or 2xTP1(SPS)-Core constructs despite still displaying a preference for a 16 bp gap between RBPJ binding sites on these promoters. Based on these results, we believe that the single R1685A mutation (which ablates dimerization of the other NICDs) on N4ICD was insufficient to abolish its dimerization activity suggesting that perhaps N4ICD has an alternative dimerization interface. Indeed, crystallography work on human N1ICD homodimers indicated two other amino acids (K1945 and E1949) beyond the R1984 that are involved in ankyrin-mediated dimerization [16]. Interestingly the N4ICD is the only NICD with an “R” located at the K1945 site, suggesting this position may be especially important for N4ICD dimerization (S2 Fig).

In addition to comparing the activation parameters of the various NICDs on these optimized promoters, we were also curious about the standard 16-nucleotide gap distance established by the N1ICD containing tripartite complex. Two hypotheses presented themselves to us. First, given the contortion of DNA evident in the crystalized N1ICD trimeric complex, we hypothesized that RBPJ sites outside of the optimal 16 bp gap might be utilized through further contortions of DNA between RBPJ binding sites. Second, we further hypothesized that given the variable sequences and sizes of the NICD proteins, different NICDs might have preferences for slightly longer or shorter gap distances between RBPJ binding sites. To test the first hypothesis, we created SPS sites with variable gap distances in 5 bp increments to take advantage of the periodicity of helical DNA. We found that none of these long-range alternative gap distances, for any of the NICDs, had significantly enhanced signaling above monomeric signaling. This indicated that long-range interactions between NTC complexes are unlikely to occur. Our analysis however only extended to measuring 36 bp gaps and therefore, longer-range interactions between RBPJ binding sites cannot be ruled out. To address the second hypothesis (that different sized NICDs might have subtle differences in SPS preference), we compared NICD activity on SPS sites with 11–21 bp gap distances in one bp increments. Similar to previous results [16,19,38], we found that all full-length NICDs also prefer a 16 bp gap distance, with little room for variation. Based on this result, we conclude that the C-terminal regions of NICDs do not appear to change the SPS dimensions preferred by NICDs nor impact NICD dimerization. It should also be stated, however, that deletion of the N1ICD C-terminal tail (ΔS2184) which contains the trans-activation domain (TAD) significantly reduced transcriptional activity (S3 Fig). Minor differences in the NICDs however were observed. For example, N2ICD appeared slightly more restrictive when choosing SPS sites than Notch1, since N2ICD activated 15/16 bp spacers but not the 17 bp gap. This poses an interesting possibility where N1ICD might dimer-dependently activate CUL1 or TXN2, which have SPSs with a 17-nucleotide gap distance [16,18], whereas binding of N2ICD on this promoter would only cause monomeric levels of transcription, though this idea was left untested. While N3ICD and N4ICD had questionable dimerization activities, they did seem to have cooperative binding and higher activity on the 16/17 bp gap SPSs. Making firm conclusions about their preferences is difficult however due to their low overall activity and high variability. Together, these experiments highlight the stringent requirements for dimerization-induced transcription.

Our results have also shed some light on the nature of NICD dimerization. First, there has been some debate as to the order of events leading up to NICD activated transcription. It is currently unclear whether individual NICD molecules first bind to RBPJ/DNA then form dimers, or alternatively if two NICDs first form a dimer, then bind to RBPJ/DNA. In Figs 4 and 5, we demonstrated that dimer-incompetent N1ICD performed slightly better than WT N1ICD on SPS elements with non-optimal 11 or 21 bp gaps. This suggests that the attempt to dimerize may impede NICD molecules from binding to these sites and might be a clue as to the mechanics of NICD function. Our result indicates that non-optimal SPS elements discourage NICD dimer formation and we believe this is evidence that NICD molecules are pre-forming dimers before binding to RBPJ/DNA. Whether or not this has an actual impact on how promoters with non-optimal SPS sites are utilized by Notch signaling remains to be seen but as shown by Castel et al., [18], several Notch responsive promoters with non-optimal SPS elements have been identified. Another outstanding question about NICD dimerization is whether or not NICD molecules can engage in heterodimerization. Given the conservation of sequence in the ankyrin domains and the importance of ankyrin domains for NICD dimerization it has been hypothesized that NICD heterodimers may exist. While further research on this topic is certainly warranted, our results in Fig 6 showing similar SPS element preferences and in Fig 4 showing ChIP recovery of N1/N4 complexes suggests that heterodimers between N1ICD and N4ICD can form in transfected 293T cells and therefore possibly under more physiological conditions. Here again, the biological implications of this observation are unknown but given the strong differences in transcriptional activation between N1ICD and N4ICD, we hypothesize that N1/N4ICD heterodimers would have intermediate activity compared to N1ICD or N4ICD homodimers. Thus, heterodimerization of NICD molecules may offer a new mechanism to regulate outputs from Notch signaling.

Finally, our minimized Hes5(SPS) promoter was not sufficient to elicit dimerization-dependent activation. The Hes5 promoter does not contain a canonical sequence-paired site and instead has been described as ‘cryptic’ [16], with a standard forward RBPJ binding site but an abnormal reverse site. When arranged as a palindrome in a SPS, this abnormal reverse site does not support dimerization, yet it facilitated dimerization when paired with a strong RBPJ binding site [19]. Work with EMSAs of this cryptic SPS showed that N1ICD homodimers can form, and dimerization-dependent activation through this site was supported through luciferase assays [16], but distinctly, these luciferase assays were still performed with the full-length promoter. In our analyses we isolated out this SPS, which should be sufficient for dimerization, yet this construct did not demonstrate dimerization-dependent activation.

This inconsistency poses two thought-provoking problems. First, and as originally described by Severson et al. [19], if cryptic sequence-paired sites are capable of forming NICD dimers, then searching for SPSs by ‘sequence-gazing’ becomes far more difficult. For example, previous ChIP work isolated out RBPJ-bound DNA targets, and these sequences were screened for nearby RBPJ binding motifs located in tandem [18]. The issue here lies in the partner sequence, wherein any non-conforming RBPJ sequences would be missed through a simple screening approach. To further identify other cryptic sequence-paired sites, like those in Hes5 or pTα [16,17], we propose that a logical course of action is to perform ChIP-Seq on a double-selected pool of DNA that specifically identifies two dimerizing NICD partners. The second problem concerns the activation of Hes5 through its SPS. While it appears that NTC complexes dimerize on this promoter segment in vitro, do they still form in living cells and if so, what’s the missing link for transcriptional activation? Other than the cryptic RBPJ binding site, the Hes5(SPS) construct is nearly identical to the Hes1(SPS) and 2xTP1(SPS)-Core constructs which respond as expected. Since the full-length Hes5 promoter is apparently dimer-responsive and the other minimalized SPS constructs are sufficient for activation, we predict that there are other promoter elements involved which enable these low-affinity dimers to form and signal in a cellular context. Comparing the promoters and enhancers of multiple, cryptic, SPS genes may identify other sequence motifs in common and identify signaling or regulatory pathways involved.

In conclusion, since much of the work on NICD dimerization has been performed studying C-terminally truncated N1ICD, we felt it was important to 1.) examine full-length NICD molecules to insure the C-terminal domain does not affect NICD promoter preference, and 2.) compare promoter preferences of the other NICD molecules, which have been largely overlooked. In so doing, we confirmed that while the C-terminal domain of the various NICDs has a transcriptional role, this domain does not appear to play a role in promoter preference. In addition, we also found that all the mammalian NICDs have remarkably similar SPS gap length preferences with only minor (+/- one bp) flexibility. Overall, these results both support previous work but also help fill in missing gaps in our understanding of Notch transcriptional activity.

Supporting information

S1 Fig. Identification of low and high affinity SPS sites in Hes1 and Hes5 promoters.

DNA sequences are as reported for Addgene products Hes1-luciferase (#41723) and Hes5-luciferase (41724) vectors. Possible high and low affinity sites (as described in material and methods) are indicated by forward or reverse arrows. High affinity sites confirmed to serve as RBPj binding sites are indicated with red shading. Nucleotides that diverge from the core TP1 element as originally defined (C/tGTGGGAA) are indicated by bolded letters.

(TIF)

Click here for additional data file.^{(2.6MB, tif)}

S2 Fig. Sequence comparison of human N1ICD and mouse NICD dimerization domains.

Amino acids previously identified as important for dimerization of N1ICD ankyrin domains are indicated [16]. Amino acid numbers are based on equivalent positions in human N1ICD. R1984A equivalents were used throughout this work to attempt to make generate dimer-incompetent NICD variants.

(TIF)

Click here for additional data file.^{(3MB, tif)}

S3 Fig. N1ICD C-terminal does not impact NICD dimerization.

HEK293T cells were transfected with 11, 16, or 21 bp versions of Hes1(SPS) or 2xTP1(SPS)-Core luciferase reporters and either full-length N1ICD or ΔS2184 N1ICD (which deletes residues C-terminal of the N1ICD ankyrin domain). Shown is the average +/- SE of four independent experiments.

(TIF)

Click here for additional data file.^{(601.1KB, tif)}

S1 Table. Plasmid and sequence information for the NICD-activated sequence-paired sites.

(DOCX)

Click here for additional data file.^{(16.4KB, docx)}

S1 Raw images

(PDF)

Click here for additional data file.^{(3.8MB, pdf)}

Data Availability

All relevant data are within the manuscript and its Supporting Information files.

Funding Statement

This work was supported by funding from the National Institute of General Medical Sciences (https://www.nigms.nih.gov) to A. Albig (2R15GM102852-02 and 1R15GM134501-01) and from NIGMS grants P20GM103408 and P20GM109095.

References

1.LaFoya B, Munroe JA, Mia MM, Detweiler MA, Crow JJ, Wood T, et al. Notch: A multi-functional integrating system of microenvironmental signals. Dev Biol. 2016;418: 227–241. 10.1016/j.ydbio.2016.08.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Schroeter EH, Kisslinger JA, Kopan R. Notch-1 signalling requires ligand-induced proteolytic release of intracellular domain. Nature. 1998;393: 382–386. 10.1038/30756 [DOI] [PubMed] [Google Scholar]
3.Mumm JS, Schroeter EH, Saxena MT, Griesemer A, Tian X, Pan DJ, et al. A ligand-induced extracellular cleavage regulates gamma-secretase-like proteolytic activation of Notch1. Mol Cell. 2000;5: 197–206. 10.1016/s1097-2765(00)80416-5 [DOI] [PubMed] [Google Scholar]
4.Kopan R, Schroeter EH, Weintraub H, Nye JS. Signal transduction by activated mNotch: Importance of proteolytic processing and its regulation by the extracellular domain. Proc Natl Acad Sci U S A. 1996;93: 1683–1688. 10.1073/pnas.93.4.1683 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Siebel C, Lendahl U. Notch signaling in development, tissue homeostasis, and disease. Physiol Rev. 2017;97: 1235–1294. 10.1152/physrev.00005.2017 [DOI] [PubMed] [Google Scholar]
6.Hsieh JJ, Henkel T, Salmon P, Robey E, Peterson MG, Hayward SD. Truncated mammalian Notch1 activates CBF1/RBPJk-repressed genes by a mechanism resembling that of Epstein-Barr virus EBNA2. Mol Cell Biol. 1996;16: 952–959. 10.1128/mcb.16.3.952 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Wu L, Aster JC, Blacklow SC, Lake R, Artavanis-Tsakonas S, Griffin JD. MAML1, a human homologue of Drosophila mastermind, is a transcriptional co-activator for NOTCH receptors. Nat Genet. 2000;26: 484–489. 10.1038/82644 [DOI] [PubMed] [Google Scholar]
8.Borggrefe T, Oswald F. The Notch signaling pathway: Transcriptional regulation at Notch target genes. Cell Mol Life Sci. 2009;66: 1631–1646. 10.1007/s00018-009-8668-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Ong CT, Cheng HT, Chang LW, Ohtsuka T, Kageyama R, Stormo GD, et al. Target selectivity of vertebrate notch proteins: Collaboration between discrete domains and CSL-binding site architecture determines activation probability. J Biol Chem. 2006;281: 5106–5119. 10.1074/jbc.M506108200 [DOI] [PubMed] [Google Scholar]
10.Meitinger C, Strobl LJ, Marschall G, Bornkamm GW, Zimber-Strobl U. Crucial Sequences within the Epstein-Barr Virus TP1 Promoter for EBNA2-Mediated Transactivation and Interaction of EBNA2 with Its Responsive Element. J Virol. 1994;68: 7497–7506. 10.1128/JVI.68.11.7497-7506.1994 [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Strobl LJ, Höfelmayr H, Stein C, Marschall G, Brielmeier M, Laux G, et al. Both Epstein-Barr viral nuclear antigen 2 (EBNA2) and activated notch 1 transactivate genes by interacting with the cellular protein RBP-Jκ. Immunobiology. 1997;198: 299–306. 10.1016/s0171-2985(97)80050-2 [DOI] [PubMed] [Google Scholar]
12.Bailey AM, Posakony JW. Suppressor of hairless directly activates transcription of Enhancer of split complex genes in response to Notch receptor activity. Genes Dev. 1995;9: 2609–2622. 10.1101/gad.9.21.2609 [DOI] [PubMed] [Google Scholar]
13.Nellesen DT, Lai EC, Posakony JW. Discrete enhancer elements mediate selective responsiveness of Enhancer of split complex genes to common transcriptional activators. Dev Biol. 1999;213: 33–53. 10.1006/dbio.1999.9324 [DOI] [PubMed] [Google Scholar]
14.Cave JW, Loh F, Surpris JW, Xia L, Caudy MA. A DNA Transcription Code for Cell-Specific Gene Activation by Notch Signaling. Curr Biol. 2005;15: 94–104. 10.1016/j.cub.2004.12.070 [DOI] [PubMed] [Google Scholar]
15.Nam Y, Sliz P, Pear WS, Aster JC, Blacklow SC. Cooperative assembly of higher-order Notch complexes functions as a switch to induce transcription. Proc Natl Acad Sci U S A. 2007;104: 2103–8. 10.1073/pnas.0611092104 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Arnett KL, Hass M, McArthur DG, Ilagan MXG, Aster JC, Kopan R, et al. Structural and mechanistic insights into cooperative assembly of dimeric Notch transcription complexes. Nat Struct Mol Biol. 2010;17: 1312–1317. 10.1038/nsmb.1938 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Liu H, Chi AWS, Arnett KL, Chiang MY, Xu L, Shestova O, et al. Notch dimerization is required for leukemogenesis and T-cell development. Genes Dev. 2010;24: 2395–2407. 10.1101/gad.1975210 [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Castel D, Mourikis P, Bartels SJJ, Brinkman AB, Tajbakhsh S, Stunnenberg HG. Dynamic binding of RBPJ is determined by notch signaling status. Genes Dev. 2013;27: 1059–1071. 10.1101/gad.211912.112 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Severson E, Arnett KL, Wang H, Zang C, Taing L, Liu H, et al. Genome-wide identification and characterization of Notch transcription complex–binding sequence-paired sites in leukemia cells. Sci Signal. 2017;10 10.1126/scisignal.aag1598 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Wang H, Zang C, Taing L, Arnett KL, Wong YJ, Pear WS, et al. NOTCH1-RBPJ complexes drive target gene expression through dynamic interactions with superenhancers. Proc Natl Acad Sci U S A. 2014;111: 705–710. 10.1073/pnas.1315023111 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Mattingly RR, Macara IG. Phosphorylation-dependent activation of the Ras-GRF-CDC25Mm exchange factor by muscarinic receptors and G-protein By subunits. Nature. 1996;382: 268–272. 10.1038/382268a0 [DOI] [PubMed] [Google Scholar]
22.Nishimura M, Isaka F, Ishibashi M, Tomita K, Tsuda H, Nakanishi S, et al. Structure, Chromosomal Locus, and Promoter of Mouse Hes2 Gene, a Homologue of Drosophila hairy and Enhancer of split. Genomics. 1998;49: 69–75. 10.1006/geno.1998.5213 [DOI] [PubMed] [Google Scholar]
23.Saxena MT, Schroeter EH, Mumm JS, Kopan R. Murine Notch Homologs (N1-4) Undergo Presenilin-dependent Proteolysis. J Biol Chem. 2001;276: 40268–40273. 10.1074/jbc.M107234200 [DOI] [PubMed] [Google Scholar]
24.Lafoya B, Munroe JA, Pu X, Albig AR. Src kinase phosphorylates Notch1 to inhibit MAML binding. Sci Rep. 2018;8: 1–12. 10.1038/s41598-017-17765-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Dhanesh SB, Subashini C, James J. Hes1: the maestro in neurogenesis. Cell Mol Life Sci. 2016;73: 4019–4042. 10.1007/s00018-016-2277-z [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Takebayashi K, Akazawa C, Nakanaishi S, Kageyama R. Structure and Promoter Analysis of the Gene Encoding the Mouse Helix-Loop-Helix Factor Hes-5. J Biol Chem. 1995;270: 1342–1349. 10.1074/jbc.270.3.1342 [DOI] [PubMed] [Google Scholar]
27.Takebayashi K, Sasai Y, Sakai Y, Watanabe T, Nakanishi S, Kageyama R. Structure, chromosomal locus, and promoter analysis of the gene encoding the mouse helix-loop-helix factor HES-1. J Biol Chem. 1994;269: 5150–5156. [PubMed] [Google Scholar]
28.Akazawa C, Sasai Y, Nakanishi S, Kageyama R. Molecular characterization of a rat negative regulator with a basic helix- loop-helix structure predominantly expressed in the developing nervous system. J Biol Chem. 1992;267: 21879–21885. [PubMed] [Google Scholar]
29.Friedmann DR, Kovall RA. Thermodynamic and structural insights into CSL-DNA complexes. Protein Sci. 2010;19: 34–46. 10.1002/pro.280 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Raffetseder U, Rauen T, Boor P, Ostendorf T, Hanssen L, Floege J, et al. Extracellular YB-1 blockade in experimental nephritis upregulates Notch-3 receptor expression and signaling. Nephron—Exp Nephrol. 2011;118: 100–108. 10.1159/000324209 [DOI] [PubMed] [Google Scholar]
31.Hansson EM, Teixeira AI, Gustafsson M V, Dohda T, Chapman G, Meletis K, et al. Recording Notch Signaling in Real Time. Dev Neurosci. 2006;28: 118–127. 10.1159/000090758 [DOI] [PubMed] [Google Scholar]
32.Tun T, Hamaguchi Y, Matsunami N, Furukawa T, Honjo T, Kawaichi M. Recognition sequence of a highly conserved DNA binding protein RBP-Jk. Nucleic Acids Res. 1994;22: 965–971. 10.1093/nar/22.6.965 [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Del Bianco C, Vedenko A, Choi SH, Berger MF, Shokri L, Bulyk ML, et al. Notch and MAML-1 complexation do not detectably alter the DNA binding specificity of the transcription factor CSL. PLoS One. 2010;5 10.1371/journal.pone.0015034 [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Kurooka H, Kuroda K, Honjo T. Roles of the ankyrin repeats and C-terminal region of the mouse Notch1 intracellular region. Nucleic Acids Res. 1998;26: 5448–5455. 10.1093/nar/26.23.5448 [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Carrieri FA, Dale JK. Turn It Down a Notch. Front Cell Dev Biol. 2017;4: 1–9. 10.3389/fcell.2016.00151 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Vasquez-Del Carpio R, Kaplan FM, Weaver KL, VanWye JD, Alves-Guerra M-C, Robbins DJ, et al. Assembly of a Notch Transcriptional Activation Complex Requires Multimerization. Mol Cell Biol. 2011;31: 1396–1408. 10.1128/MCB.00360-10 [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Hirata H, Yoshiura S, Ohtsuka T, Bessho Y, Harada T, Yoshikawa K, et al. Oscillatory expression of the bHLH factor Hes1 regulated by a negative feedback loop. Science (80-). 2002;298: 840–843. 10.1126/science.1074560 [DOI] [PubMed] [Google Scholar]
38.Nam Y, Weng AP, Aster JC, Blacklow SC. Structural requirements for assembly of the CSL-intracellular Notch1-Mastermind-like 1 transcriptional activation complex. J Biol Chem. 2003;278: 21232–21239. 10.1074/jbc.M301567200 [DOI] [PubMed] [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0234101.r001

Decision Letter 0

George Mosialos

5 Jul 2020

PONE-D-20-14835

Notch family members follow stringent requirements for intracellular domain dimerization at sequence-paired sites

PLOS ONE

Dear Dr. Albig,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

I recommend that you pay particular attention and address the points raised by reviewer 1.

Please submit your revised manuscript by Aug 17 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

George Mosialos

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2.PLOS ONE now requires that authors provide the original uncropped and unadjusted images underlying all blot or gel results reported in a submission’s figures or Supporting Information files. This policy and the journal’s other requirements for blot/gel reporting and figure preparation are described in detail at https://journals.plos.org/plosone/s/figures#loc-blot-and-gel-reporting-requirements and https://journals.plos.org/plosone/s/figures#loc-preparing-figures-from-image-files. When you submit your revised manuscript, please ensure that your figures adhere fully to these guidelines and provide the original underlying images for all blot or gel data reported in your submission. See the following link for instructions on providing the original image data: https://journals.plos.org/plosone/s/figures#loc-original-images-for-blots-and-gels.

In your cover letter, please note whether your blot/gel image data are in Supporting Information or posted at a public data repository, provide the repository URL if relevant, and provide specific details as to which raw blot/gel images, if any, are not available. Email us at plosone@plos.org if you have any questions.

3. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Review of: “Notch family members follow stringent requirements for intracellular domain dimerization at sequence paired sites.”

Summary: The focus of this paper is to compare the transcriptional activity of Notch1, 2, 3, and 4 wild type and mutants containing a specific amino acid change in an analogous residue of each molecule that was previously shown to disrupt NICD-1 dimer formation on SPS sites. The general strategy is to use cell culture and over-expression assays coupled with an array of different luciferase vectors containing Rbpj/CSL binding sites in different orientations and with different spacers to assess transcriptional changes. The authors further use Western Blot analysis to compare NICD protein levels, and perform a ChIP-PCR study to assess for protein binding to DNA.

The authors make several conclusions based on their data including, A) That all Notch ICD molecules can form dimers on the SPS sites; B) Each Notch dimer requires similar spacing and site orientation as that published for Notch1. C) The authors claim to have found a “mechanical difference between canonical and cryptic SPSs, leading to differences in their dimerization-induced regulation.” D) Notch1 and Notch4 can form heterodimers on SPS sites. Overall, the authors have provided evidence for some of these conclusions, but not all. Hence, I have the following concerns in regards to the experimental conclusions/interpretations and the experimental design of the study.

1) An assumption made in the paper is that the same mutation that disrupts Notch1-ICD dimer formation will also disrupt N2, N3, and N4 dimer formation. Past publications used assays like EMSAs or FRET assays to show that Notch1-ICD forms dimers, but the authors provided no direct evidence that the mutations tested in the luciferase assays actually abolishes dimer formation for the other Notch molecules. Hence, any change in transcriptional activity is assumed to be caused by a decrease in dimer formation without any evidence it actually does cause that change. They need to provide such evidence or they need to dramatically change the way they write the paper about whether these mutations actually impact cooperative DNA binding.

2) A better description of the plasmid design of the synthetic luciferase reporters is needed. In particular, I did not understand how these plasmids work given the authors wrote the following: “We isolated the mouse and human Hes5 SPS elements containing two head-to-head orientated RBPJ binding sites separated by a 16-nucleotide gap, and the 4 to 5 surrounding nucleotides (Figure 2A) as previously described and cloned these sequences into promoterless luciferase reporter vectors.” The authors similarly stated that they used a promoterless pGL3-basic vector in the methods. If the luciferase vector does not have a promoter and the synthetic insert only has an SPS site – then how does RNA polymerase initiate transcription? Is the assumption here that the constructs initiate transcription in the complete absence of a promoter or does it contain a minimal promoter that is insufficient to activate gene expression on its own? If it completely lacks a promoter – the authors should provide a clear rationale for why it provides a good model to use to study Notch-dependent transcription, as all Notch target genes presumably have a promoter. Given this discrepancy, the relevance of the results can be called into question.

3) The authors claim in the abstract that “each family member is capable of dimerization-induced signaling”, but the authors provide no evidence for Notch4. The only evidence the authors show is that N4 induces optimal induction of a 2xTP1(SPS)core construct with a 16bp spacer in Figure 6. But the authors used that same luciferase reporter to test the N4 wild type and mutant protein and found no significant change in activation using the mutant N4 molecule in Figure 4B. Hence, no N4 mutant protein was shown to significantly change transcription via SPS sites or any other site. Hence, the authors do not have the data to support this claim.

4) The authors also did double-selection ChIP-PCR experiments in Figure 4 to make the following claim: “Our analysis determined that both Notch1 and Notch4 homodimers cooperate on the 2xTP1(SPS)-Core construct and are therefore likely actively signaling as a duplex (Figure 4C). Importantly, we also observed Notch1-Notch4 heterodimers, though at a lesser abundance. This implies that two different Notch proteins can come together at the same promoter site and interact, possibly modulating the other’s transcriptional output.” I have several concerns about the interpretation of this experiment. First, the authors stated that these homodimers bind in a “cooperative” manner. However, no proper positive or negative controls are included to demonstrate cooperativity. For example, the ChIP-PCR experiments should also be done on a region containing two high affinity binding sites that do not mediate cooperative binding (i.e. 21bp spacer) and show that there is a dramatic decrease in binding – which would be consistent with cooperativity. Second, the authors need a negative control by testing a reporter with both binding sites mutated. Third, it is standard practice to quantify this data using a fold enrichment over input chromatin. Fourth, the authors should also test the dimer mutations – especially to make the claim that N4/N1 heterodimers form – given that the authors have not convincingly shown that N4 even forms homodimers (see point 3 above).

5) The authors made the following statement in the results section: “The results in figure 4A revealed an interesting phenomenon wherein non-dimerizing ankyrin mutant NICDs performed better than their WT counterparts on the 2xTP1(SPS)-Complete construct which had a slightly longer gap than the 16 bp preferred by NICD molecules. This observation suggested that two RBPJ binding sites gapped slightly more or less than 16 bp within a promoter might actually suppress promoter responsiveness to NICD dimer-dependent Notch signaling and favor NICD dimer-independent Notch signaling, a phenomenon which has not been previously described.” However the change was only from ~5-fold for the wild type protein to ~7.5-fold for the RA mutation. Couldn’t such a small relative change in luciferase activity between the wild type vs RA mutant protein actually be explained by differences in the levels of the NICD wild type and RA proteins? In looking at the Western blots in Figure 1C, it appears that the NICD-RA proteins are consistently higher than the Wild type protein – although no quantitation is provided. This increase in protein levels could explain why the RA protein is better on the 11bp and 21bp spacer luciferase constructs. Note, the increase in levels of the RA protein relative to wild type could be due to increased protein stability as has been recently suggested in Drosophila by Kuang et al eLife 2020. However, another recent paper (Kobia et al 2020 on bioarchives) found that the N2RA mutation did not dramatically alter N2ICD levels – and thus such changes in levels/stability could be cell type specific.

6) The authors should better describe how they selected the sequences used in the different spacers. Were they selected to exclude additional TF binding sites? Some previously described SPS sequences were found to often contain a common hexamer sequence (typically something like GAAAGT, which is found in Hes1, see Posakony papers). Did they purposely avoid or include such sequences in their constructs? Did they scan the sequences to exclude the possibility of making additional low affinity Rbpj sites?

7) Minor comment: Figure 6 needs statistical tests.

Reviewer #2: In this interesting study, Crow and Albig dissect the requirements and activity for the dimerization of all 4 different NOTCH proteins using reporter assays in 293T cells. Authors show that all members can form dimers and can even form heterodimers, while the sequence GAP for dimer binding-sequences is mostly limited to the originally described 16bp gap. I think this study will positively contribute to the understanding of the NOTCH field.

I only have very minor comments:

1. Authors shouldn't refer to "dimerization null constructs", but rather "dimerization incompetent constructs" throughout the text.

2. In Fig 1D, it would be good if authors could exactly specify the high/low affinity DNA sequences identified in a Supplementary Fig.

3. In the discussion, authors hypothesize that R1945 might be relevant in N4ID for dimerization. Can authors check this? This should be fairly easy to do using their reporter system, upon site-directed mutagenesis for that position.

4. Please correct grammar in the Discussion in the sentence "This indicated

that long-range interactions between NTC complexes are unlikely TO occur"

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2020 Nov 24;15(11):e0234101. doi: 10.1371/journal.pone.0234101.r002

Author response to Decision Letter 0

20 Aug 2020

Thank you for serving as editor for our submission to PLOS One. We thank the reviewers for their thoughtful review of our manuscript and we were honored by the overall enthusiasm for our work. We have addressed the reviewers concerns on a point-by-point basis below. In each case, we repeat the reviewers concern and follow with our response below each point.

Reviewer #1:

- The reviewer is concerned that the amino acid mutation that renders N1ICD dimer incompetent, may not also cause the other NICDs to be dimer incompetent. The reviewer is also concerned that reduced transcriptional activity of NICD mutants might be due to altered protein expression or stability. This is a fair concern and one that we would have loved to address with EMSA assay (as suggested). However, despite our best attempts, we have been unable to make the NICD EMSA assay work in our hands. As an alternative, we have resorted to our luciferase promoter assay to address the reviewer's concerns. In our new data, we compared the transcriptional activity of both WT and mutated NICD molecules on a dimer-independent 4XTP1 luciferase construct. We rationalized that if WT and mutant NICD performed equivalently on this 4XTP1 that this would eliminate any concerns about protein expression, stability, or interactions with CSL and leave the only explanation for reduced activity on dimer-dependent promoters that the mutation blocks dimerization activity. Our new data in figure 1D shows that all the WT and mutant NICD molecules in fact do have equivalent activities on this dimer-independent construct suggesting that changes in transcriptional activity are not related to protein expression or stability. As has been noted elsewhere in this rebuttal letter, N4ICD is the exception since it does not appear to function in a dimer dependent manner based on mutation of the ankyrin domain. That said, our data does demonstrate what appears to be dimerization activity by ChIP and by the preference of N4ICD for a 16 bp template. Based on these observations, we do feel that N4ICD engages in dimerization, but that the amino acid mutated in our present experiments "missed the mark". Instead, future research is going to be focused on identifying the critical amino acids in N4ICD that are critical for dimerization activities.

2) A better description of the plasmid design of the synthetic luciferase reporters is needed. In particular, I did not understand how these plasmids work given the authors wrote the following: "We isolated the mouse and human Hes5 SPS elements containing two head-to-head orientated RBPJ binding sites separated by a 16-nucleotide gap, and the 4 to 5 surrounding nucleotides (Figure 2A) as previously described and cloned these sequences into promoterless luciferase reporter vectors." The authors similarly stated that they used a promoterless pGL3-basic vector in the methods. If the luciferase vector does not have a promoter and the synthetic insert only has an SPS site - then how does RNA polymerase initiate transcription? Is the assumption here that the constructs initiate transcription in the complete absence of a promoter or does it contain a minimal promoter that is insufficient to activate gene expression on its own? If it completely lacks a promoter - the authors should provide a clear rationale for why it provides a good model to use to study Notch-dependent transcription, as all Notch target genes presumably have a promoter. Given this discrepancy, the relevance of the results can be called into question.

- The reviewer is absolutely correct. The term "promoterless" is a sort of shorthand commonly used to describe a plasmid that has a minimal promoter, incapable of initiating transcription without additional enhancer elements (such as the SPS sites) used in our paper. We have clarified this throughout the manuscript.

3) The authors claim in the abstract that "each family member is capable of dimerization-induced signaling", but the authors provide no evidence for Notch4. The only evidence the authors show is that N4 induces optimal induction of a 2xTP1(SPS)core construct with a 16bp spacer in Figure 6. But the authors used that same luciferase reporter to test the N4 wild type and mutant protein and found no significant change in activation using the mutant N4 molecule in Figure 4B. Hence, no N4 mutant protein was shown to significantly change transcription via SPS sites or any other site. Hence, the authors do not have the data to support this claim.

-We agree with the reviewer that our evidence for N4ICD dimerization is very minimal. Unfortunately, I believe that the amino acid mutation that seems to "break" dimerization in the other NICD molecules does not affect N4ICD in the same way. We have enhanced our previous figure 1 (now figure S2) to show an expanded sequence alignment that provides a better understanding into how the assumed dimerization domain is fairly different in N4ICD compared to the other NICDs. Regardless, we cannot say for sure that N4ICD does (or does not) dimerize. As the reviewer notes, our only evidence is a preference for the 16bp gap and the CHiP data in panel 4C wherein we detect N4ICD dimers by Co-IP and PCR. We attempted to emphasize this throughout the text but overlooked this comment in the abstract. This has been amended. This comment is also similar to one of reviewer two's comments. We are currently performing experiments to determine which amino acids in N4ICD are necessary for dimerization.

4) The authors also did double-selection ChIP-PCR experiments in Figure 4 to make the following claim: "Our analysis determined that both Notch1 and Notch4 homodimers cooperate on the 2xTP1(SPS)-Core construct and are therefore likely actively signaling as a duplex (Figure 4C). Importantly, we also observed Notch1-Notch4 heterodimers, though at a lesser abundance. This implies that two different Notch proteins can come together at the same promoter site and interact, possibly modulating the other's transcriptional output." I have several concerns about the interpretation of this experiment.

First, the authors stated that these homodimers bind in a "cooperative" manner. However, no proper positive or negative controls are included to demonstrate cooperativity. For example, the ChIP-PCR experiments should also be done on a region containing two high affinity binding sites that do not mediate cooperative binding (i.e. 21bp spacer) and show that there is a dramatic decrease in binding - which would be consistent with cooperativity.

-Thank you, reviewer, for pointing this out. We were mistaken to suggest that we were interested in demonstrating cooperativity in this experiment. Instead, we were only interested in observing NICD dimerization. We have extensively rewritten this section in the revised manuscript. Demonstrating cooperativity would be extremely challenging by this approach since as we understand the assembly of the Notch tripartite complex, the synergy of the 16bp template is not associated with NICD/RBPj binding to DNA, but rather to the ability to recruit RNA polymerase. Therefore, since the 21 bp promoter can still theoretically bind to two differently tagged (but undimerized) NICD molecules, we would probably still detect ChIP on the 21 bp promoter. A better readout for cooperative function is in fact strong transcriptional activity on the 16bp template which is demonstrated in figure 6.

Second, the authors need a negative control by testing a reporter with both binding sites mutated.

-We also originally thought about using mutated SPS sites as a negative control in this experiment. However given the sensitivity of PCR as a readout, we were concerned that any residual binding (even on mutated sites) would be detected in our PCR analysis. Since we were also not attempting a quantitative approach to our analysis, this would have been difficult to control for. In the end, we feel that the two negative controls we did use are excellent. The most basic negative control we used was just the luciferase plasmid transfected with no NICD. The more robust negative control we used was the sample transfected with HA-tagged NICD and lacking Flag-tagged NICD, but still subjected to the two-step Co-IP procedure. We felt (and still do) that the absence of PCR signal in this more robust control provides very good evidence for the specificity and lack of background in our experiment.

Third, it is standard practice to quantify this data using a fold enrichment over input chromatin.

-It was never our intention (at this point anyway) to use this ChIP assay for anything more than a demonstration of another technique to detect dimerized NICD molecules on DNA.

Fourth, the authors should also test the dimer mutations - especially to make the claim that N4/N1 heterodimers form - given that the authors have not convincingly shown that N4 even forms homodimers (see point 3 above).

-On the surface, this sounds like an excellent control for our experiment however it is necessary to remember that any two NICD molecules might be able to bind to the DNA regardless if they are dimer capable or incapable. Therefore, this experiment would probably give false positives, suggesting dimerization even though it was not happening. As stated above, we really feel the controls we chose to use in this study are a good compromise to show specificity of the results and to avoid the potentially sticky false positives we sought to avoid.

5) The authors made the following statement in the results section: "The results in figure 4A revealed an interesting phenomenon wherein non-dimerizing ankyrin mutant NICDs performed better than their WT counterparts on the 2xTP1(SPS)-Complete construct which had a slightly longer gap than the 16 bp preferred by NICD molecules. This observation suggested that two RBPJ binding sites gapped slightly more or less than 16 bp within a promoter might actually suppress promoter responsiveness to NICD dimer-dependent Notch signaling and favor NICD dimer-independent Notch signaling, a phenomenon which has not been previously described." However the change was only from ~5-fold for the wild type protein to ~7.5-fold for the RA mutation. Couldn't such a small relative change in luciferase activity between the wild type vs RA mutant protein actually be explained by differences in the levels of the NICD wild type and RA proteins? In looking at the Western blots in Figure 1C, it appears that the NICD-RA proteins are consistently higher than the Wild type protein - although no quantitation is provided. This increase in protein levels could explain why the RA protein is better on the 11bp and 21bp spacer luciferase constructs. Note, the increase in levels of the RA protein relative to wild type could be due to increased protein stability as has been recently suggested in Drosophila by Kuang et al eLife 2020. However, another recent paper (Kobia et al 2020 on bioarchives) found that the N2RA mutation did not dramatically alter N2ICD levels - and thus such changes in levels/stability could be cell type specific.

-The reviewer is correct that this small change could be caused by properties of the mutant N1ICD other than dimerization potential. This concern however is addressed in the revised figure 1D wherein we compare WT and ankyrin mutant NICD transcriptional activity on dimer independent template 4xTP1. We found no significant difference between WT and mutant on this promoter suggesting that any difference between WT and mutant NICDs is based on dimerization potential and the synergistic transcriptional activity associated with NICD dimerization.

-We have revised the section in the material and methods that describes plasmid construction to more precisely address this concern.

7) Minor comment: Figure 6 needs statistical tests.

-At the reviewers request we did perform ANOVA and Tukey-Kramer's HSD post-hoc test on this data. We have reported these results in the revised figure legend for figure 6.

Reviewer #2:

1. Authors shouldn't refer to "dimerization null constructs", but rather "dimerization incompetent constructs" throughout the text.

-As suggested, we have made this change throughout the manuscript.

2. In Fig 1D, it would be good if authors could exactly specify the high/low affinity DNA sequences identified in a Supplementary Fig.

-As suggested, we have prepared a new more detailed figure S1 that shows the DNA sequences of all of the high/low affinity sites.

-We agree with the reviewer that this is an experiment that needs to be done. At this time however, Mr. Crow is ready to leave the lab and the project has been passed onto a new student. This will be the subject of a future study.

4. Please correct grammar in the Discussion in the sentence "This indicated

that long-range interactions between NTC complexes are unlikely TO occur"

-done.

Attachment

Submitted filename: Rebuttal points.docx

Click here for additional data file.^{(146.1KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0234101.r003

Decision Letter 1

George Mosialos

16 Sep 2020

PONE-D-20-14835R1

Notch family members follow stringent requirements for intracellular domain dimerization at sequence-paired sites

PLOS ONE

Dear Dr. Albig,

Please submit your revised manuscript by Oct 31 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

We look forward to receiving your revised manuscript.

Kind regards,

George Mosialos

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

**********

6. Review Comments to the Author

Reviewer #1: Review of: “Notch family members follow stringent requirements for intracellular domain dimerization at sequence paired sites.”

Summary: In the revised version of this paper, the authors addressed several of my comments – in particular the fact that the reporter contained a minimal promoter (not promoterless), some additional data showing that the NICD-RA mutants could still activate gene expression, and re-wording of the abstract to not state that N4ICD forms dimers on SPS sites. However, one point was not addressed in a very convincing way and I also raise a few other minor issues the authors should consider.

1) I still do not believe the ChIP study demonstrates that N1ICD and N4ICD form heterodimers. The luciferase 2xTP1 DNA used in this assay has two Rbpj binding sites. Thus, when the ChiP-reChiP assay is done and the authors find that both N1ICD and N4ICD are bound to the same DNA – how do they know it forms a heterodimer (which is what they claim) versus simply one site bound by N4 and the other site independently bound by N1? That is why I recommended doing additional control experiments with Rbpj binding sites separated by 21 nts – which are non- cooperative sites. In fact, the authors themselves stated in the rebuttal to my comments the following: “since the 21bp promoter can still theoretically bind to two different tagged (but undimerized) NICD molecules, we would probably still detect ChIP on the 21bp promoter.” But that argument is also true of a 16bp spacer as well. I also recommended testing dimer-deficient NICD molecules in this assay – to which they stated in their rebuttal: “on the surface, this sounds like an excellent control for our experiment, however, it is necessary to remember that any two NICD molecules might be able to bind to the DNA regardless if they are dimer capable or incapable.” And yet, in the manuscript, the authors state that their data supports the conclusion that N1 and N4 form heterodimers on SPS sites. In my opinion, this data shows that N4 and N1 can both bind to the same DNA that has two Rbpj binding sites at the same time. It does not show that these molecules form heterodimers.

2) Figure 1B needs significance tests between the WT and RA mutants for N2ICD, N3ICD, and N4ICD.

3) A point of clarity – In Fig 4A and 4B the sequence is provided for the SPS site tested and it is labeled TP1 (complete) and TP1 (core). However in all the luciferase data it is called 2xTP1(SPS)-complete or 2xTP1(SPS)-core. Does that mean it contains two copies of the core and complete sequence? Or should the above sequence be re-labeled as 2xTP1?

4) A minor comment for the authors to consider – but it is their choice to leave it or change it. I still personally find the small differences in luciferase activity between the WT and RA mutant NICD molecules to be overstated. Hence, I don't really understand why they make such a big deal about these small differences (luciferase assays are really sensitive and it is unclear what a less than a 2-fold difference means). Hence, I find their argument in the section on “Non-optimal SPS sites select against transcriptional activation by NICD dimers” to be less than compelling. But again, I am a firm supporter of the authors telling the story how they want to – I would just state that as a person that has studied transcription for over 25 years and performed luciferase assays throughout that time – it is really hard to determine the importance of such small changes, even when statistically different.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

PLoS One. 2020 Nov 24;15(11):e0234101. doi: 10.1371/journal.pone.0234101.r004

Author response to Decision Letter 1

27 Oct 2020

10-26-2020

Dear Dr. Mosialos,

Thank you for continuing serving as editor for our submission to PLOS One. We again thank reviewer #1 for his/her thoughtful review of our manuscript. I must say that I admire reviewer #1 for the dedication shown in reviewing our manuscript. In the true sense of the peer review process, our manuscript has been greatly improved due to the reviewer's diligence. I hope our new revisions will be satisfactory. We have addressed the concerns below.

1) I still do not believe the ChIP study demonstrates that N1ICD and N4ICD form heterodimers. The luciferase 2xTP1 DNA used in this assay has two Rbpj binding sites. Thus, when the ChiP-reChiP assay is done and the authors find that both N1ICD and N4ICD are bound to the same DNA - how do they know it forms a heterodimer (which is what they claim) versus simply one site bound by N4 and the other site independently bound by N1? That is why I recommended doing additional control experiments with Rbpj binding sites separated by 21 nts - which are non- cooperative sites. In fact, the authors themselves stated in the rebuttal to my comments the following: "since the 21bp promoter can still theoretically bind to two different tagged (but undimerized) NICD molecules, we would probably still detect ChIP on the 21bp promoter." But that argument is also true of a 16bp spacer as well. I also recommended testing dimer-deficient NICD molecules in this assay - to which they stated in their rebuttal: "on the surface, this sounds like an excellent control for our experiment, however, it is necessary to remember that any two NICD molecules might be able to bind to the DNA regardless if they are dimer capable or incapable." And yet, in the manuscript, the authors state that their data supports the conclusion that N1 and N4 form heterodimers on SPS sites. In my opinion, this data shows that N4 and N1 can both bind to the same DNA that has two Rbpj binding sites at the same time. It does not show that these molecules form heterodimers.

Originally, this figure was only intended to show that the 2xTP1(core) construct was in fact able to bind NICD molecules, but it has (for the better) taken on a whole new "life of its own". The reviewer is 100% correct about the concerns with the experiment and we have performed the suggested experiments. I have included the new data in this letter. The result is very interesting and raises new questions about how Notch functions. However, as is so often the case, this new data raises more questions than it answers. And so, given that the manuscript does not hinge on this data, I have opted to remove panel 4C from the final manuscript. That said, I plan to pursue this ChiP approach to see what it can teach us (if anything) about how Notch functions. This data can be found in the submitted response to reviewers letter.

The new data shows (as the reviewer correctly guessed) that NICD molecules can associate with the DNA even if the gap between RBPj binding sites is non-optimal (21 bp) or the NICD molecules are dimer-null (Ank Mut). Keep in mind however that the N4ICD ankyrin mutation used here is the same mutation we used in the manuscript that was shown to NOT break N4ICD dimerization activity (I have a student currently trying to solve that little N4ICD mystery). However, the N1ICD ankyrin mutation does certainly suppress transcription from SPS sites with 16bp gaps as shown in our manuscript. Regardless, we are forced to agree with the reviewer that this data cannot distinguish between 1.) NICDs engaging in dimerization on the DNA, or 2.) two NICD molecules binding to DNA independently of dimerization.

That said, the new ChiP data does raise other questions and I would welcome the reviewers input about what this data might mean. If the reviewer feels comfortable, I would welcome a discussion in any format he/she feels appropriate with. Even as an anonymous note through the official reply to this resubmission would be welcome. In particular, the new data is particularly interesting since we consistently recover the most DNA in the 4/4 sample ChiP sample, followed by the 1/4 sample, followed by 1/1 sample. Almost always in that order. And, it is not just the transfected target DNA since ChiP analysis of the chromosomal Hes1 and Hes4 promoters returned the same pattern of DNA recovery (not shown).

If this data does not illustrate dimerization, what does it mean? Does this data suggest that N4ICD simply binds to RBPj and DNA better than N1ICD? Does it suggest an activity of NICD binding to DNA that we are not aware of? Or, maybe the data is screaming the obvious at us … maybe NICDs bind just fine to 21 bp gapped SPS elements and they don't need ankyrin domains to bind (this is what the data says). Maybe the only way to get the synergistic amplification of transcription that comes with NICD dimerization is to have dimerization on the 16bp gap so that MAML and p300 can bind to the a NICD complex with the correct dimensions? My student's and I are currently scratching our heads over this one. In the end, since the data is not critical for the overall paper, I have decided to simply remove the data and save it for another time when we can more robustly analyze the experiment, perform follow up experiments, and give these questions the full attention they deserve.

2) Figure 1B needs significance tests between the WT and RA mutants for N2ICD, N3ICD, and N4ICD.

This has been updated as requested.

3) A point of clarity - In Fig 4A and 4B the sequence is provided for the SPS site tested and it is labeled TP1 (complete) and TP1 (core). However in all the luciferase data it is called 2xTP1(SPS)-complete or 2xTP1(SPS)-core. Does that mean it contains two copies of the core and complete sequence? Or should the above sequence be re-labeled as 2xTP1?

Thank you for pointing this out. These are both 2xTP1 elements and we have modified the description of each DNA sequence appropriately.

4) A minor comment for the authors to consider - but it is their choice to leave it or change it. I still personally find the small differences in luciferase activity between the WT and RA mutant NICD molecules to be overstated. Hence, I don't really understand why they make such a big deal about these small differences (luciferase assays are really sensitive and it is unclear what a less than a 2-fold difference means). Hence, I find their argument in the section on "Non-optimal SPS sites select against transcriptional activation by NICD dimers" to be less than compelling. But again, I am a firm supporter of the authors telling the story how they want to - I would just state that as a person that has studied transcription for over 25 years and performed luciferase assays throughout that time - it is really hard to determine the importance of such small changes, even when statistically different.

I completely understand your point here and I hesitate to fly in the face of experience. It is not a major point in our paper, however I think that I would like to keep this data in the paper. My reason for doing this is that there is not much known about how the NICD complexes actually bind DNA and initiate transcription. My gut instinct is telling me that this data might be able to teach us something about Notch. Hopefully, it will be useful for others.

Attachment

Submitted filename: rebuttal letter #2.pdf

Click here for additional data file.^{(432.4KB, pdf)}

PLoS One. doi: 10.1371/journal.pone.0234101.r005

Decision Letter 2

George Mosialos

10 Nov 2020

Notch family members follow stringent requirements for intracellular domain dimerization at sequence-paired sites

PONE-D-20-14835R2

Dear Dr. Albig,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

George Mosialos

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

Reviewer #1: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

**********

6. Review Comments to the Author

Reviewer #1: The authors have done a great job addressing the remaining concerns and I strongly support publication.

Regarding the request for my thoughts on the new unpublished ChIP-reChIP data - I don't have anything profound to say! One thing that I also struggle with is the association of amount of DNA binding with the amount of transcriptional activation. In some cases, there has been a clear link between transcription factor turnover and the amount of transcription activation with higher rates of protein turnover linked with increased transcription. So, one possibility is that NICD1 may be more rapidly phosphorylated and ubiquitylated when bound to SPS sites and thus turnover quickly whereas NICD4 may have a slow turnover kinetics but not activate transcription as well...... Again, totally my guess but there has been a recent paper in eLife with some evidence that NICD binding in flies leads to faster turnover when bound to SPS sites than CSL sites - however, this phenomenon was not linked directly to transcriptional activation.

Best of luck on your future work!

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Brian Gebelein

PLoS One. doi: 10.1371/journal.pone.0234101.r006

Acceptance letter

George Mosialos

13 Nov 2020

PONE-D-20-14835R2

Notch family members follow stringent requirements for intracellular domain dimerization at sequence-paired sites

Dear Dr. Albig:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. George Mosialos

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Fig. Identification of low and high affinity SPS sites in Hes1 and Hes5 promoters.

(TIF)

Click here for additional data file.^{(2.6MB, tif)}

S2 Fig. Sequence comparison of human N1ICD and mouse NICD dimerization domains.

(TIF)

Click here for additional data file.^{(3MB, tif)}

S3 Fig. N1ICD C-terminal does not impact NICD dimerization.

(TIF)

Click here for additional data file.^{(601.1KB, tif)}

S1 Table. Plasmid and sequence information for the NICD-activated sequence-paired sites.

(DOCX)

Click here for additional data file.^{(16.4KB, docx)}

S1 Raw images

(PDF)

Click here for additional data file.^{(3.8MB, pdf)}

Attachment

Submitted filename: Rebuttal points.docx

Click here for additional data file.^{(146.1KB, docx)}

Attachment

Submitted filename: rebuttal letter #2.pdf

Click here for additional data file.^{(432.4KB, pdf)}

Data Availability Statement

All relevant data are within the manuscript and its Supporting Information files.

[pone.0234101.ref001] 1.LaFoya B, Munroe JA, Mia MM, Detweiler MA, Crow JJ, Wood T, et al. Notch: A multi-functional integrating system of microenvironmental signals. Dev Biol. 2016;418: 227–241. 10.1016/j.ydbio.2016.08.023 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref002] 2.Schroeter EH, Kisslinger JA, Kopan R. Notch-1 signalling requires ligand-induced proteolytic release of intracellular domain. Nature. 1998;393: 382–386. 10.1038/30756 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref003] 3.Mumm JS, Schroeter EH, Saxena MT, Griesemer A, Tian X, Pan DJ, et al. A ligand-induced extracellular cleavage regulates gamma-secretase-like proteolytic activation of Notch1. Mol Cell. 2000;5: 197–206. 10.1016/s1097-2765(00)80416-5 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref004] 4.Kopan R, Schroeter EH, Weintraub H, Nye JS. Signal transduction by activated mNotch: Importance of proteolytic processing and its regulation by the extracellular domain. Proc Natl Acad Sci U S A. 1996;93: 1683–1688. 10.1073/pnas.93.4.1683 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref005] 5.Siebel C, Lendahl U. Notch signaling in development, tissue homeostasis, and disease. Physiol Rev. 2017;97: 1235–1294. 10.1152/physrev.00005.2017 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref006] 6.Hsieh JJ, Henkel T, Salmon P, Robey E, Peterson MG, Hayward SD. Truncated mammalian Notch1 activates CBF1/RBPJk-repressed genes by a mechanism resembling that of Epstein-Barr virus EBNA2. Mol Cell Biol. 1996;16: 952–959. 10.1128/mcb.16.3.952 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref007] 7.Wu L, Aster JC, Blacklow SC, Lake R, Artavanis-Tsakonas S, Griffin JD. MAML1, a human homologue of Drosophila mastermind, is a transcriptional co-activator for NOTCH receptors. Nat Genet. 2000;26: 484–489. 10.1038/82644 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref008] 8.Borggrefe T, Oswald F. The Notch signaling pathway: Transcriptional regulation at Notch target genes. Cell Mol Life Sci. 2009;66: 1631–1646. 10.1007/s00018-009-8668-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref009] 9.Ong CT, Cheng HT, Chang LW, Ohtsuka T, Kageyama R, Stormo GD, et al. Target selectivity of vertebrate notch proteins: Collaboration between discrete domains and CSL-binding site architecture determines activation probability. J Biol Chem. 2006;281: 5106–5119. 10.1074/jbc.M506108200 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref010] 10.Meitinger C, Strobl LJ, Marschall G, Bornkamm GW, Zimber-Strobl U. Crucial Sequences within the Epstein-Barr Virus TP1 Promoter for EBNA2-Mediated Transactivation and Interaction of EBNA2 with Its Responsive Element. J Virol. 1994;68: 7497–7506. 10.1128/JVI.68.11.7497-7506.1994 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref011] 11.Strobl LJ, Höfelmayr H, Stein C, Marschall G, Brielmeier M, Laux G, et al. Both Epstein-Barr viral nuclear antigen 2 (EBNA2) and activated notch 1 transactivate genes by interacting with the cellular protein RBP-Jκ. Immunobiology. 1997;198: 299–306. 10.1016/s0171-2985(97)80050-2 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref012] 12.Bailey AM, Posakony JW. Suppressor of hairless directly activates transcription of Enhancer of split complex genes in response to Notch receptor activity. Genes Dev. 1995;9: 2609–2622. 10.1101/gad.9.21.2609 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref013] 13.Nellesen DT, Lai EC, Posakony JW. Discrete enhancer elements mediate selective responsiveness of Enhancer of split complex genes to common transcriptional activators. Dev Biol. 1999;213: 33–53. 10.1006/dbio.1999.9324 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref014] 14.Cave JW, Loh F, Surpris JW, Xia L, Caudy MA. A DNA Transcription Code for Cell-Specific Gene Activation by Notch Signaling. Curr Biol. 2005;15: 94–104. 10.1016/j.cub.2004.12.070 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref015] 15.Nam Y, Sliz P, Pear WS, Aster JC, Blacklow SC. Cooperative assembly of higher-order Notch complexes functions as a switch to induce transcription. Proc Natl Acad Sci U S A. 2007;104: 2103–8. 10.1073/pnas.0611092104 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref016] 16.Arnett KL, Hass M, McArthur DG, Ilagan MXG, Aster JC, Kopan R, et al. Structural and mechanistic insights into cooperative assembly of dimeric Notch transcription complexes. Nat Struct Mol Biol. 2010;17: 1312–1317. 10.1038/nsmb.1938 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref017] 17.Liu H, Chi AWS, Arnett KL, Chiang MY, Xu L, Shestova O, et al. Notch dimerization is required for leukemogenesis and T-cell development. Genes Dev. 2010;24: 2395–2407. 10.1101/gad.1975210 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref018] 18.Castel D, Mourikis P, Bartels SJJ, Brinkman AB, Tajbakhsh S, Stunnenberg HG. Dynamic binding of RBPJ is determined by notch signaling status. Genes Dev. 2013;27: 1059–1071. 10.1101/gad.211912.112 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref019] 19.Severson E, Arnett KL, Wang H, Zang C, Taing L, Liu H, et al. Genome-wide identification and characterization of Notch transcription complex–binding sequence-paired sites in leukemia cells. Sci Signal. 2017;10 10.1126/scisignal.aag1598 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref020] 20.Wang H, Zang C, Taing L, Arnett KL, Wong YJ, Pear WS, et al. NOTCH1-RBPJ complexes drive target gene expression through dynamic interactions with superenhancers. Proc Natl Acad Sci U S A. 2014;111: 705–710. 10.1073/pnas.1315023111 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref021] 21.Mattingly RR, Macara IG. Phosphorylation-dependent activation of the Ras-GRF-CDC25Mm exchange factor by muscarinic receptors and G-protein By subunits. Nature. 1996;382: 268–272. 10.1038/382268a0 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref022] 22.Nishimura M, Isaka F, Ishibashi M, Tomita K, Tsuda H, Nakanishi S, et al. Structure, Chromosomal Locus, and Promoter of Mouse Hes2 Gene, a Homologue of Drosophila hairy and Enhancer of split. Genomics. 1998;49: 69–75. 10.1006/geno.1998.5213 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref023] 23.Saxena MT, Schroeter EH, Mumm JS, Kopan R. Murine Notch Homologs (N1-4) Undergo Presenilin-dependent Proteolysis. J Biol Chem. 2001;276: 40268–40273. 10.1074/jbc.M107234200 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref024] 24.Lafoya B, Munroe JA, Pu X, Albig AR. Src kinase phosphorylates Notch1 to inhibit MAML binding. Sci Rep. 2018;8: 1–12. 10.1038/s41598-017-17765-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref025] 25.Dhanesh SB, Subashini C, James J. Hes1: the maestro in neurogenesis. Cell Mol Life Sci. 2016;73: 4019–4042. 10.1007/s00018-016-2277-z [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref026] 26.Takebayashi K, Akazawa C, Nakanaishi S, Kageyama R. Structure and Promoter Analysis of the Gene Encoding the Mouse Helix-Loop-Helix Factor Hes-5. J Biol Chem. 1995;270: 1342–1349. 10.1074/jbc.270.3.1342 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref027] 27.Takebayashi K, Sasai Y, Sakai Y, Watanabe T, Nakanishi S, Kageyama R. Structure, chromosomal locus, and promoter analysis of the gene encoding the mouse helix-loop-helix factor HES-1. J Biol Chem. 1994;269: 5150–5156. [PubMed] [Google Scholar]

[pone.0234101.ref028] 28.Akazawa C, Sasai Y, Nakanishi S, Kageyama R. Molecular characterization of a rat negative regulator with a basic helix- loop-helix structure predominantly expressed in the developing nervous system. J Biol Chem. 1992;267: 21879–21885. [PubMed] [Google Scholar]

[pone.0234101.ref029] 29.Friedmann DR, Kovall RA. Thermodynamic and structural insights into CSL-DNA complexes. Protein Sci. 2010;19: 34–46. 10.1002/pro.280 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref030] 30.Raffetseder U, Rauen T, Boor P, Ostendorf T, Hanssen L, Floege J, et al. Extracellular YB-1 blockade in experimental nephritis upregulates Notch-3 receptor expression and signaling. Nephron—Exp Nephrol. 2011;118: 100–108. 10.1159/000324209 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref031] 31.Hansson EM, Teixeira AI, Gustafsson M V, Dohda T, Chapman G, Meletis K, et al. Recording Notch Signaling in Real Time. Dev Neurosci. 2006;28: 118–127. 10.1159/000090758 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref032] 32.Tun T, Hamaguchi Y, Matsunami N, Furukawa T, Honjo T, Kawaichi M. Recognition sequence of a highly conserved DNA binding protein RBP-Jk. Nucleic Acids Res. 1994;22: 965–971. 10.1093/nar/22.6.965 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref033] 33.Del Bianco C, Vedenko A, Choi SH, Berger MF, Shokri L, Bulyk ML, et al. Notch and MAML-1 complexation do not detectably alter the DNA binding specificity of the transcription factor CSL. PLoS One. 2010;5 10.1371/journal.pone.0015034 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref034] 34.Kurooka H, Kuroda K, Honjo T. Roles of the ankyrin repeats and C-terminal region of the mouse Notch1 intracellular region. Nucleic Acids Res. 1998;26: 5448–5455. 10.1093/nar/26.23.5448 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref035] 35.Carrieri FA, Dale JK. Turn It Down a Notch. Front Cell Dev Biol. 2017;4: 1–9. 10.3389/fcell.2016.00151 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref036] 36.Vasquez-Del Carpio R, Kaplan FM, Weaver KL, VanWye JD, Alves-Guerra M-C, Robbins DJ, et al. Assembly of a Notch Transcriptional Activation Complex Requires Multimerization. Mol Cell Biol. 2011;31: 1396–1408. 10.1128/MCB.00360-10 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0234101.ref037] 37.Hirata H, Yoshiura S, Ohtsuka T, Bessho Y, Harada T, Yoshikawa K, et al. Oscillatory expression of the bHLH factor Hes1 regulated by a negative feedback loop. Science (80-). 2002;298: 840–843. 10.1126/science.1074560 [DOI] [PubMed] [Google Scholar]

[pone.0234101.ref038] 38.Nam Y, Weng AP, Aster JC, Blacklow SC. Structural requirements for assembly of the CSL-intracellular Notch1-Mastermind-like 1 transcriptional activation complex. J Biol Chem. 2003;278: 21232–21239. 10.1074/jbc.M301567200 [DOI] [PubMed] [Google Scholar]

PERMALINK

Notch family members follow stringent requirements for intracellular domain dimerization at sequence-paired sites

Jacob J Crow

Allan R Albig

Roles

Abstract

Introduction

Materials and methods

Cell culture

Expression and reporter plasmids

Construct creation and mutagenesis

Western blotting

Luciferase assays

Statistical analysis

Results

Activation of Notch target genes containing sequence-paired sites requires ankyrin-dependent dimerization

Fig 1. Dimerization dependence of full-length Hes1 and Hes5 promoters.

The isolated Hes5 sequence-paired site does not respond to dimerization

Fig 2. Lack of dimerization dependence of the Hes5 sequence-paired site.

The Hes1 SPS acts through traditional NICD dimerization mechanisms

Fig 3. Dimer dependence of the Hes1 sequence-paired site.

Establishment of a high activity, NICD dimer-specific reporter construct

Fig 4. An optimized Notch responsive SPS element.

Non-optimal SPS sites select against transcriptional activation by NICD dimers

Fig 5. Non-optimal SPS sites select against dimer-dependent NICD transcriptional activation.

A restrictive spacer range dictates the signaling capabilities of NICD dimerization

Fig 6. Notch homodimers do not tolerate alternative gap distances.

Discussion

Supporting information

Data Availability

Funding Statement

References

Decision Letter 0

George Mosialos

Roles

Author response to Decision Letter 0

Decision Letter 1

George Mosialos

Roles

Author response to Decision Letter 1

Decision Letter 2

George Mosialos

Roles

Acceptance letter

George Mosialos

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases