Research Article

Increasing the Unneddylated Cullin1 Portion Rescues the csn Phenotypes by Stabilizing Adaptor Modules To Drive SCF Assembly

Qingqing Liu, Yike Zhou, Ruiqi Tang, Xuehong Wang, Qiwen Hu, Ying Wang, Qun He
DOI: 10.1128/MCB.00109-17

This article has a correction. Please see:

ABSTRACT

The dynamic SCF (Skp1–cullin1–F-box protein) assembly is controlled by cycles of cullin neddylation/deneddylation based on the deneddylation activity of the COP9 signalosome (CSN) and global sequestration of cullins by CAND1. However, acceptance of this prediction was hampered by the results of recent studies, and the regulatory mechanism and key players remain to be identified. We found that maintaining a proper Cul1Nedd8/Cul1 ratio is crucial to ensure SCF functions. Reducing the high Cul1Nedd8/Cul1 ratios in csn mutants through ectopic expression of the nonneddylatable Cul1K722R proteins or introducing the endogenous cul1K722R point mutation significantly rescues their defective phenotypes. In vivo protein degradation assays revealed that the large portion of the unneddylated Cul1 contributes to F-box protein stabilization. Moreover, the unneddylated Cul1 tends to associate with adaptor modules, and disruption of Cul1–Skp-1 binding fails to restore the csn phenotypes. Taking the data together, we propose that unneddylated Cul1 is another central player involved in maintenance of the adaptor module pool through the formation of Cul1–Skp-1–F-box complexes and promotion of rapid SCF assembly.

INTRODUCTION

Maintaining an integrated and functional proteome is crucial for cell survival and development (1). Eukaryotes have evolved two major systems to remove aberrant proteins: the ubiquitin-proteasome system (UPS) and lysosomal degradation system (autophagy) (2). The UPS pathway is a major clearance system for misfolded or unwanted proteins in which ubiquitin is first activated by E1 (ubiquitin activating enzyme), subsequently transferred to E2 (ubiquitin conjugating enzyme), and then covalently attached to substrates supported by E3 (ubiquitin ligase) (3–5). Among the three classes of enzymes, E3 ubiquitin ligase binds to both substrate and E2 thioesterified with ubiquitin and plays a central role in providing the specificity for the substrates to be ubiquitinated. The cullin-RING ligases (CRLs) represent the largest family of E3s, which consist of, in general, four components: cullin scaffolds, RING-finger proteins, adaptor proteins, and substrate recognition receptors (SR) (6–8). The SCF (Skp1–cullin1–F-box protein) complexes are the best-studied members of CRLs. As a molecular scaffold, cullin1 N terminus recruits the Skp1–F-box complex, and its C terminus interacts with Rbx1 RING protein to form an SCF E3 ligase that promotes ubiquitin transfers from Rbx1-bound E2 to a substrate, which is recognized through the WD40 or leucine-rich repeat (LRR) domain of an F-box protein. Through promoting ubiquitination and subsequent degradation of large amounts of substrates, the Cul1-mediated SCF pathway plays significant roles in cancer, cell cycle regulation, signal transduction, and development (7).

The key feature of CRLs is that the cullins can be covalently modified by Nedd8 (neural precursor cell expressed developmentally downregulated protein 8) molecules on a conserved lysine residue at its C-terminal winged-helix motif (9, 10). Like ubiquitination, conjugation of Nedd8 to target proteins is executed by an E1 (Nedd8-activating enzyme 1 [NAE1])-E2 (Nedd8-conjugating enzyme [Ubc12])-E3 (defective in cullin neddylation 1 [DCN1] and Rbx1) multienzyme cascade (11). Genetic studies indicated that neddylation is important for the activities of CRLs in vivo (12–14). Structural studies showed that there is a gap of about 50 Å between the substrate docking site and the E2 active site, which is too great a distance for E2 to transfer the ubiquitin to the substrate. Neddylation modification of cullin1 induces a conformation change to bridge the gap and to help position the bound E2-S-Ub for transferring ubiquitin to the sequestered substrate (15–17). Reversibly, the Nedd8 conjugation can be removed through the process of deneddylation, which is achieved by the activity of the COP9 signalosome (CSN) complex (18, 19). CSN is an evolutionarily conserved protein complex in eukaryotes and was first discovered in Arabidopsis thaliana as a repressor of constitutive photomorphogenesis (20). Recently, the 3.8-Å resolution CSN crystal structure was determined, providing new detailed insight into the organization of this holoenzyme (21).

Whereas CSN negatively regulates the in vitro activity of CRLs through the removal of Nedd8, genetic data indicate a positive role of CSN with respect to the optimal function of CRLs in vivo (8, 22–25). This apparent paradox can be resolved by a model in which dynamic cycles of neddylation and deneddylation are required to regulate the assembly and activity of cullin-based ubiquitin E3 ligases. The notion of this model is that the cullin-Rbx1 core complex transitions between active and inactive states. The active complex, containing a substrate adaptor protein with its bound substrate, is modified by Nedd8 on a conserved lysine residue at the C terminus of cullin, which facilitates the transfer of ubiquitin from Rbx1-recruited E2 to the substrate. After substrate degradation, the active complex is converted to an inactive complex with deneddylation of cullin by CSN; CAND1 then binds the deneddylated cullin protein and blocks the binding of the substrate adaptor protein. Subsequent neddylation of cullin by Ubc12 weakens the grip of CAND1 on cullin and enables another substrate adaptor protein to displace CAND1 and to form a new active CRL (24). This prediction rationalized the CSN paradox and has been prevailing in the past years. Nevertheless, the cycling model is challenged by several factors. First, loss of CAND1 orthologs did not affect the neddylation states of cullins in plants, human cells, or yeast, suggesting that the neddylation cycle may function independently of CAND1 (26–29). Second, CSN can also inhibit the activity of CRLs independently of its deneddylase activity and this regulation can be influenced by the levels of substrates (30, 31). Third, quantitative proteomic studies showed that the large portion of unneddylated cullins are not converted into cullin-CAND1 complexes (32).

There are three conserved cullin proteins (cullin1, cullin3, and cullin4) and a CSN complex (composed of CSN-1 to CSN-7) in Neurospora crassa. In order to investigate the regulatory mechanism associated with the Nedd8 modification with respect to the activity of CRLs, we generated Cul1 neddylation-deficient mutant cul1K722R. Unexpectedly, the phenotype of the cul1K722R strain mimicked the phenotype of CSN deletion strains on race tubes, suggesting that a proper Cul1Nedd8/Cul1 ratio (the unneddylated Cul1 makes up a large portion of the total Cul1 amount) is important for maintaining the normal activity of SCF. To change the proportion of neddylated Cul1 or unneddylated Cul1, we ectopically expressed Myc-tagged Cul1 or Cul1K722R proteins in csn mutants. Surprisingly, the defective phenotypes of csn, including irregular conidiation rhythm, impaired light regulation processes, and temperature-sensitive growth, were shown to be significantly rescued by the inducible expression of unneddylated Myc-Cul1K722R proteins rather than of Myc-Cul1 proteins, as the ectopically expressed Cul1K722R proteins sustainably reduced the high Cul1Nedd8/Cul1 ratios in csn mutants. In addition, we created a csn-1KO cul1K722R double mutant to decrease the level of neddylated Cul1 in the csn-1KO strain; as expected, the phenotypic defects were restored as a consequence of the neddylation site mutation. A protein degradation assay showed that the unneddylated Cul1K722R proteins contribute to the stabilization of F-box proteins and promote degradation of substrates in csn mutants. Furthermore, abolishment of the Cul1–Skp-1 interaction failed to restore the csn phenotypes and the stability of the F-box proteins seen with the Myc-Cul1K722R proteins. Therefore, a properly large portion of unneddylated Cul1 is required to stabilize the F-box proteins through the formation of Cul1–Skp-1–F-box complexes to rapidly initiate SCF assembly and activation, which is a new function of the unneddylated Cul1 protein in cells and indicates an alternative mechanism for the dynamic SCF organization.

RESULTS

Cullin1 neddylation-deficient mutant cul1K722R mimics the phenotypes of CSN deletion strains.Attachment of ubiquitin-like protein Nedd8 to the Cul1 C terminus is essential for the function of SCF-type ubiquitin E3 ligases. Alignments performed with homologous sequences of cullin1 in different species showed that the neddylation site was conserved from fungi to human and that the Nedd8 molecule was conjugated to Neurospora Cul1 at the lysine722 residue (Fig. 1A). In order to evaluate the effect of Nedd8 modification on Cul1-based ubiquitin ligases in Neurospora, we generated a heterozygous neddylation site mutation strain by homolog recombination. Western blotting and sequencing results showed that lysine722 of endogenous Cul1 was largely changed into arginine in the heterozygous cul1K722R knock-in strain (here named strain cul1K722R) compared to the wild-type (WT) strain (Fig. 1B and C). We did not succeed in generating a homozygous cul1K722R knock-in strain, indicating that the neddylation modification of Cul1 is essential for cell survival. To determine whether the neddylation point mutation of Cul1 affects the activity of SCF, we tested the stability of FRQ (FREQUENCY; the circadian clock component involved in the generation of biological rhythms in Neurospora), one of the ubiquitination substrates of SCFFWD-1 ubiquitin ligase, in the WT and cul1K722R strains. As shown in Fig. 1D and E, after addition of the protein synthesis inhibitor cycloheximide (CHX), the FRQ protein levels in both strains decreased, but the degradation rate of FRQ was much lower in the cul1K722R strain, indicating that the FRQ degradation pathway was severely blocked in the cul1K722R strain.

FIG 1
FIG 1

Cullin1 neddylation-deficient mutant cul1K722R mimics the phenotypes of CSN deletion strains. (A) Alignment with homologous sequences of cullin1 in seven species: Neurospora crassa (Nc), Homo sapiens (Hs), Mus musculus (Mm), Arabidopsis thaliana (At), Drosophila melanogaster (Dm), Caenorhabditis elegans (Ce), and Schizosaccharomyces pombe (Sp). The lysine722 residue (K722) indicated by the asterisk is the neddylation site of Cul1 in Neurospora crassa. (B) Western blot analysis showing the levels of Cul1, Cul1Nedd8, Cul3, and Cul3Nedd8 in the WT, cul1K722R, and csn-1KO strains. (C) Sequencing results from the WT and cul1K722R strains. The lysine residue (AAG) at position 722 in the WT strain is largely mutated into arginine (CGT) (marked with an asterisk) in the cul1K722R heterozygous strain. (D) Western blot analysis showing the degradation of FRQ in the WT and cul1K722R strains after cycloheximide (CHX) treatment (10 μg/ml). Cultures were grown under LL (constant light) conditions. (E) Densitometric analysis of the degradation of FRQ in the WT and cul1K722R strains after addition of cycloheximide (CHX). Error bars represent standard deviations of the results of three experiments. (F) Race tube assays of the conidiation rhythm in the WT, csn, and cul1K722R strains in minimal medium under DD conditions (constant darkness, 25°C; the black bar represents the dark treatment), LD conditions (light-dark cycles [12 h light/12 h darkness], 25°C; the white bars represent the light treatment, and the black bars represent the dark treatment), and Tm conditions (temperature cycles [constant darkness, 12 h 25°C/12 h 30°C]; the white bars represent 30°C, and the black bars represent 25°C) after 1 day of light treatment at 25°C. The black lines mark the growth fronts determined every 24 h.

We next examined the circadian conidiation rhythm of the cul1K722R strain by race tube assays. Surprisingly, the race tube assays performed under conditions of cycles of constant darkness (DD; 25°C), 12 h light/12 h darkness (LD; 25°C), and two melting temperatures (Tm; 12 h at 25°C/12 h at 30°C in DD) revealed that the growth patterns and conidiation rhythms of the cul1K722R strain were very similar to those of csn deletion mutants, with the exception of the csn-3KO strain (Fig. 1F). The cul1K722R and csn deletion strains but not the csn-3KO mutant grew more slowly and produced fewer vegetative hyphae and conidia than the WT strain. It is notable that the conidiation rhythm of cul1K722R strain was irregular and interrupted and could not be entrained under LD or Tm cycle conditions (Fig. 1F), which is the typical phenotype of csn mutants (except mutant csn-3KO) reported in a previous study (33).

However, the cul3KO and cul4KO strains exhibited growth patterns and conidiation rhythms that were totally different from those of the csn mutants (see Fig. S1 in the supplemental material), suggesting that the specific phenotype of csn mutants mainly results from the presence of dysfunctional Cul1 in the cells. On the one hand, the hyperneddylation of Cul1 in CSN deletion strains could not be removed and SCF ligases remained at a persistently active state, resulting in instability of the F-box proteins. On the other hand, the neddylation site mutation of Cul1 in the cul1K722R strains resulted in hyponeddylation of most Cul1 and the blockage of ubiquitin transfer to substrates. Both ends reached the same goal: the reduction of SCF activity and the accumulation of substrates. Thus, the neddylation status of Cul1 plays key roles in SCF activities to control the substrate ubiquitination and degradation responsible for the phenotypic changes seen with the csn and cul1K722R mutants in Neurospora.

The Cul1Nedd8/Cul1 ratio is a crucial indicator for Neurospora crassa physiology and development.To further determine the relationship between Cul1 neddylation status and phenotypic changes, we examined the levels of neddylated Cul1 and unneddylated Cul1 proteins in the WT, cul1K722R, and csn strains. As shown in Fig. 2A and B, only a small fraction (about 20%) of the total amount of Cul1 was modified by Nedd8 molecules in the WT strain. In contrast, Cul1 was hyperneddylated in csn-1KO, csn-2KO, csn-4KO, csn-5KO, csn-6KO, and csn-7KO mutants. But the neddylation modification of Cul1 was not much affected by loss of CSN-3, showing a slightly higher level of Cul1Nedd8 than was shown by the WT strain (Fig. 2A and B). 339-59 is a csn-4 disruption mutant that was generated by inserting a hygromycin resistance gene (hph) into the middle of its open reading frame (ORF), resulting in the disruption of its PCI (proteasome-COP9-initiation factor) domain and C-terminal region. Western blot analyses revealed that the level of Cul1Nedd8 in the 339-59 mutant was a little higher than that in the csn-3KO strain but lower than those in the other csn mutants (Fig. 2A and B). In the cul1K722R strain, about 60% of the Cul1 was mutated into the nonneddylatable form of Cul1 and the level of the Nedd8 modification of Cul1 was even lower than that seen in the WT strain (Fig. 2A and B).

FIG 2
FIG 2

The Cul1Nedd8/Cul1 ratio is a crucial indicator for Neurospora crassa physiology and development. (A) Western blot analysis of levels of unneddylated Cul1, neddylated Cul1, F-box protein GRR-1, and FRQ in the cul1K722R, csn-1KO, csn-2KO, csn-3KO, csn-4KO, csn-5KO, csn-6KO, csn-7KO, WT, and 339-59 strains. (B) The occupancies of neddylated Cul1 and unneddylated Cul1 calculated from the quantitative analysis of the Western blot results presented in panel A. (C) Race tube assays of the conidiation rhythms in the csn-1KO, csn-2KO, csn-4KO, csn-5KO, csn-6KO, csn-7KO, 339-59, csn-3KO, WT, and cul1K722R strains in minimal medium under DD conditions. The black lines mark the growth fronts determined every 24 h. (D) The ratio of neddylated Cul1 to total Cul1 calculated from the quantitative analysis of the Western blot results. Error bars represent standard deviations of the results of three experiments.

We next analyzed the phenotypes of the WT strain and each mutant to evaluate the effect of the Cul1Nedd8/Cul1 ratio on the growth and development of Neurospora. Race tube assays showed that different Cul1 neddylation states led to different phenotypic changes. The WT strain, with a Cul1Nedd8/Cul1 ratio about 0.22, exhibited the normal growth rate and the normal conidiation rhythm, with a free-running conidiation period of about 22 h at 25°C under DD conditions (Fig. 2C and D). The csn-3KO strain, with a slightly higher Cul1Nedd8/Cul1 ratio (about 0.31), exhibited a higher growth rate and an unaffected conidiation rhythm (Fig. 2C and D). Interestingly, with a Cul1Nedd8/Cul1 ratio about 0.44, the 339-59 mutant had a growth rate similar to that of the csn-3KO strain but a disrupted conidiation rhythm (Fig. 2C and D). The csn and cul1K722R strains, with ratio values (over 0.71 and under 0.10) that deviated greatly from the normal value in the WT strain, grew very slowly and showed loss of their circadian conidiation rhythms (Fig. 2C and D). These observations strongly suggest that there is a threshold level of the Cul1Nedd8/Cul1 ratio for normal growth and development, which serves as an indicator of phenotypic changes in Neurospora. Disruption of the balance between Cul1Nedd8 and Cul1 resulted in substrate accumulation, which has also been previously observed in plants (25). Therefore, we hypothesized that the Cul1Nedd8/Cul1 ratio plays a critical role in the normal function of SCF complexes.

Inducible expression of Myc-Cul1K722R rather than Myc-Cul1 rescues the phenotypic defects of the conserved CSN subunit deletion strains.To test the hypothesis presented above, we needed to change the Cul1Nedd8/Cul1 ratio by increasing or decreasing the levels of neddylated Cul1 or unneddylated Cul1 to see whether there would be any effects on the phenotypes. We tried to overexpress the neddylation enzymes (NAE1, Ubc12, and DCN1) and Nedd8 proteins in the WT strain to elevate the levels of neddylated Cul1. But no phenotypic changes were observed in these strains. We then tried to reduce the Cul1Nedd8/Cul1 ratio by expressing nonneddylatable Cul1K722R proteins in the WT strain to mimic the hyperactive CSN complex. We made a Myc-Cul1 construct with a mutation of the conserved neddylation residue lysine722 to arginine driven by the qa-2 promoter and transferred pqa-Myc-Cul1 or pqa-Myc-Cul1K722R into the WT strains. However, there were few phenotypic changes in the wild-type background with expression of Myc-Cul1K722R or Myc-Cul1 proteins, confirming that ectopic expression of Cul1 or Cul1K722R cannot break the balance between Cul1Nedd8 and Cul1 in cells with a functional CSN complex.

To test whether the expression of Cul1 or Cul1K722R protein can alter the Cul1Nedd8/Cul1 ratio in csn mutants, we transferred pqa-Myc-Cul1 or pqa-Myc-Cul1K722R into csn strains and then systematically analyzed the effect of Myc-Cul1 or Myc-Cul1K722R expression on the growth and canidiation rhythm of csn mutants by race tube assays. Under conditions of constant darkness, the WT strain exhibited a robust circadian conidiation rhythm, with a period of about 22 h in the minimal medium (Fig. 3A; Fig. S2A to E). The csn mutants as well as the csn Myc-Cul1K722R transformants grew more slowly, produced fewer conidia than the WT strain, and showed an abnormal conidiation pattern under DD conditions using race tubes with minimal medium (Fig. 3A; Fig. S2A to E). Surprisingly, when quinic acid (QA) was added into the medium to induce the expression of Myc-Cul1K722R proteins, the phenotypic defects of csn mutants were rescued in the transformants in a QA concentration-dependent manner (Fig. 3A; Fig. S2A to E). However, expression of Myc-Cul1 in csn mutants had no effect on or even aggravated the growth and development defects of transformants on race tubes with QA addition (Fig. 3A; Fig. S2A to E). These data suggest that the defects of the normal circadian conidiation rhythm, circadian clock function, and development in csn mutants can be significantly rescued by expression of nonneddylatable Cul1K722R proteins rather than the wild-type Cul1 proteins.

FIG 3
FIG 3

Inducible expression of Myc-Cul1K722R but not Myc-Cul1 rescues the phenotypic defects of the conserved CSN subunit deletion strains. (A to C) Race tube assays for conidiation rhythms of the WT strain, csn-1KO mutant, and csn-1KO qa-Myc-Cul1 and csn-1KO qa-Myc-Cul1K722R transformants in minimal medium, 10−3 M QA medium, and 10−2 M QA medium under DD, LD, and Tm conditions. Results of phenotypic analyses of the other csn mutants and transformants are shown in Fig. S2A to E. (D) Plating assays for the thermosensitive growth of the WT strain, csn mutants, and csn qa-Myc-Cul1K722R and csn qa-Myc-Cul1 transformants in minimal and 10−2 M QA medium at 25°C and 38°C. Conidia of each strain were serially diluted and spotted onto plates with minimal or 10−2 M QA medium and then incubated for 48 h at 25°C or 38°C, respectively.

Since CSN was initially discovered in Arabidopsis as an important regulator of photomorphogenesis, it has been implicated in a wide variety of light-responsive processes in plants, Aspergillus, Neurospora, and possibly other eukaryotic organisms (34, 35). Previous studies showed that the conidiation rhythm of csn mutants cannot be entrained under conditions of LD cycles (33, 36); thus, we examined the conidiation rhythm of csn mutants with ectopic expression of Myc-Cul1K722R or Myc-Cul1 under conditions of LD cycles. As with the WT strain, LD cycles could entrain the conidiation rhythms in the csn mutants with Myc-Cul1K722R expression, but csn Myc-Cul1 transformants did not produce regular conidiation bands under conditions of LD cycles (Fig. 3B; Fig. S2A to E), indicating that the impaired regulation of conidiation by light in the csn mutants was able to be rescued by expression of nonneddylatable Cul1K722R but not wild-type Cul1. Thus, CSN-mediated deneddylation of Cul1 appears to play a conserved role in regulating light-regulated developmental processes in eukaryotic orgamisms.

Previous studies showed that the conidiation rhythm of the csn mutants also cannot be synchronized by Tm cycles (33, 36). We then tested whether the conidiation rhythms of the csn Myc-Cul1K722R and csn Myc-Cul1 transformants can be entrained under conditions of different temperatures. Under the culture conditions of 12 h at 25°C/12 h at 30°C, the conidiation rhythms of the WT and csn Myc-Cul1K722R strains were synchronized on race tubes after QA addition, with conidiation occurring in the cold phase (Fig. 3C; Fig. S2A to E). Unsurprisingly, for the csn mutants and the csn Myc-Cul1 strains, the conidiation rhythm was not entrained by exposure to Tm cycles (Fig. 3C; Fig. S2A to E). These data further demonstrate that ectopically expressed Cul1K722R proteins but not wild-type Cul1 proteins restore the defects of the temperature-regulated conidiation process in csn mutants.

We found that csn mutants (except the csn-3KO mutant) are sensitive to high temperature. As shown in Fig. 3D, the WT strain grew well on the plates even at 38°C. However, the csn mutants (except the csn-3KO mutant) were extremely sensitive under this condition (Fig. 3D), indicating that the csn mutants had lost the ability to grow at 38°C. In order to determine whether QA-driven Cul1K722R expression can also rescue the thermosensitive phenotypes of csn mutants, the WT strain, csn mutants, and csn Myc-Cul1K722R transformants were grown on plates at 25°C and 38°C to test their growth rates. The csn Myc-Cul1K722R strains grew as weakly as the csn mutants on plates with minimal medium at 38°C (Fig. 3D). Once QA had been added to the medium to induce Myc-Cul1K722R expression, the csn Myc-Cul1K722R transformants grew as well as the WT strain. However, the csn Myc-Cul1 strains were still thermosensitive at 38°C with QA addition in the medium (Fig. 3D). Therefore, expression of Myc-Cul1K722R suppresses the thermosensitive phenotypes of csn mutants.

In Neurospora, there are three conserved cullin proteins, Cul1, Cul3, and Cul4. We expressed Myc-Cul3K782R or Myc-Cul4K986R in csn-2KO mutants (Fig. S2F). As shown in Fig. S2G and H, the growth, development, and circadian rhythm of csn-2KO mutant were not suppressed by the expression of Myc-Cul3K782R or Myc-Cul4K986R proteins. Together, these results demonstrate that the phenotypic defects of csn mutants, including irregular conidiation, impaired light regulation processes, and thermosensitive growth, were dramatically rescued by ectopically expressed Myc-Cul1K722R protein but not by Myc-Cul1, Myc-Cul3K782R, or Myc-Cul4K986R protein.

Reducing the high Cul1Nedd8/Cul1 ratio in csn mutants can rescue their phenotypes.Although the Nedd8 modification plays a positive role in the normal function of SCF ubiquitin ligases, the level of neddylated Cul1 is tightly controlled in strains with functional CSN complex. As shown in Fig. 4A, the pqa-Myc-Nedd8 plasmid was cotransferred into the WT Flag-Cul1 strain and the transformants, and different expression levels of Myc-Nedd8 due to different copy numbers of inserted Myc-Nedd8 genes into genome were obtained. Western blot results showed that there were two forms of Nedd8-modified Flag-Cul1 proteins, i.e., Flag-Cul1Nedd8 and Flag-Cul1Myc-Nedd8, as the Myc-Nedd8 molecules could also be conjugated onto Cul1 (Fig. 4A). On the basis of quantitative analysis of the Western blot results, we found that there was a reciprocal relationship between the two forms of Nedd8-modified Flag-Cul1 proteins: as the levels of Flag-Cul1Myc-Nedd8 dropped, the Flag-Cul1Nedd8 levels rose, and vice versa, but the total amount of neddylated Flag-Cul1 barely changed among the different transformants (Fig. 4B). These results indicate that the neddylation level of Cul1 is tightly maintained in a steady state within cells with a functional CSN complex. These results also provide an explanation for the observation that the expression of Myc-Cul1 or Myc-Cul1K722R protein had no effect on the phenotypes of the WT strain.

FIG 4
FIG 4

Expression of Myc-Cul1K722R proteins dramatically reduces the high Cul1Nedd8/Cul1 ratios in csn mutants. (A) Western blot analysis showing the levels of neddylated Flag-Cul1 and unneddylated Flag-Cul1 in the WT Flag-Cul1 and WT Flag-Cul1/Myc-Nedd8 strains. The asterisk indicates the nonspecific background bands. (B) Quantitative analysis of the occupancy of neddylated Flag-Cul1 (Flag-Cul1Nedd8 and Flag-Cul1Myc-Nedd8) in the WT Flag-Cul1 and WT Flag-Cul1/Myc-Nedd8 transformants from the Western blot results. Error bars represent standard deviations of the results of three experiments. (C) Western blot analysis of the levels of neddylated Cul1 and unneddylated Cul1 in the WT strain, csn mutants, and csn qa-Myc-Cul1K722R transformants with anti-Cul1 antibody in a reaction mixture with 0.1% glucose and 0 M, 10−3 M, or 10−2 M QA medium. (D) Western blot analysis of the levels of neddylated Cul1 and unneddylated Cul1 in the WT strain, csn mutants, and csn qa-Myc-Cul1 transformants with anti-Cul1 antibody in a reaction mixture with 0.1% glucose and 0 M, 10−3 M, or 10−2 M QA medium. (E) The ratios of neddylated Cul1 to total Cul1 in the WT strain, csn mutants, and csn qa-Myc-Cul1K722R transformants calculated from the quantitative analysis of Western blot results. Error bars represent standard deviations of the results of three experiments. (F) The ratios of neddylated Cul1 to total Cul1 in the WT strain, csn mutants, and csn qa-Myc-Cul1 transformants calculated from the quantitative analysis of Western blot results. Error bars represent standard deviations of the results of three experiments.

Suppression of csn phenotypes by Myc-Cul1K722R rather than Myc-Cul1 prompted us to examine the changes of the Cul1Nedd8/Cul1 ratio in csn qa-Myc-Cul1K722R and csn qa-Myc-Cul1 strains. We examined the levels of neddylated Cul1 and unneddylated Cul1 in the WT strain, csn mutants, and csn qa-Myc-Cul1 and csn qa-Myc-Cul1K722R transformants cultured in 0.1% glucose liquid medium without or with QA (Fig. 4C and D). The quantitative analysis of the Western blot results showed that the levels of ectopically expressed Myc-Cul1K722R were close to or modestly higher than those of Myc-Cul1 in each csn mutant (Fig. S3). We then calculated the Cul1Nedd8/Cul1 ratio under various csn deletion conditions without or with QA addition. As shown in Fig. 4E and F, without QA induction, the Cul1Nedd8/Cul1 ratios of the csn qa-Myc-Cul1 and csn qa-Myc-Cul1K722R transformants were similar to those of csn mutants, and the majority of Cul1 was modified by Nedd8. The Cul1Nedd8/Cul1 ratios in the csn qa-Myc-Cul1K722R transformants continuously fell to reach those seen with the WT strain with the expression of Myc-Cul1K722R as a consequence of increasing the concentration of QA in the medium (Fig. 4E). Meanwhile, the Cul1Nedd8/Cul1 ratios in the csn qa-Myc-Cul1 transformants also slightly declined but were still much higher than those seen in the WT strain with induced expression of Myc-Cul1 (Fig. 4F). Taken together, these results suggest that the expression of Myc-Cul1K722R proteins but not Myc-Cul1 proteins can rescue the phenotypic defects of csn mutants through reducing the high Cul1Nedd8/Cul1 ratios.

In order to further confirm that reducing the high Cul1Nedd8/Cul1 ratio in csn mutants can rescue their phenotypes, we created the csn-1KO cul1K722R double mutant with the objective of first increasing the levels of unneddylated Cul1 and then reducing the Cul1Nedd8/Cul1 ratio seen with the csn-1KO strain (Fig. S4A and B). Race tube assays performed under DD, LD, and Tm conditions showed that the growth rate and conidiation rhythm of csn-1KO strain were significantly restored by introducing an endogenous cul1K722R mutation into the csn-1KO strain (Fig. S4C), which strongly demonstrates that reducing the proportion of neddylated Cul1 rescues the phenotypic defects of csn mutants.

The phenotypic defects of csn-3KO and 339-59 mutants are suppressed by Myc-Cul1K722R but exacerbated by Myc-Cul1.The results of a previous study showed that Neurospora CSN-3 is not essential for the deneddylation activity of CSN complex and that deletion of csn-3 slightly affected the neddylation of cullin proteins and increased the growth rate of the mutant strain (33). To test whether the hypothesis of the significance of the Cul1Nedd8/Cul1 ratio is also applicable to the csn-3KO mutant, we ectopically expressed Myc-Cul1 or Myc-Cul1K722R in the csn-3KO mutant and examined the growth rates and conidiation rhythms of the transformants by race tube assay. As shown in Fig. 5A, the csn-3KO mutant grew faster on race tubes in either minimal or QA-containing medium than the WT strain. With the addition of QA into the medium, the csn-3KO qa-Myc-Cul1K722R transformants exhibited almost the same growth rate as the WT strain, whereas the growth rate of the Myc-Cul1 transformants was even higher than that of the csn-3KO mutant under DD, LD, and Tm cycle conditions (Fig. 5A). Quantitative analysis of the Western blot results showed that the Cul1Nedd8/Cul1 ratio of the csn-3KO qa-Myc-Cul1K722R transformant decreased with QA addition (Fig. 5B and C). However, the Cul1Nedd8/Cul1 ratio was still much higher in the csn-3KO qa-Myc-Cul1 transformant with QA treatment than in the WT strain (Fig. 5B and D).

FIG 5
FIG 5

The phenotypic defects of csn-3KO mutant were suppressed by Myc-Cul1K722R but exacerbated by Myc-Cul1l. (A) Race tube assays for conidiation rhythm of the WT strain, csn-3KO mutant, and csn-3KO qa-Myc-Cul1 and csn-3KO qa-Myc-Cul1K722R transformants in minimal, 10−3 M QA, and 10−2 M QA medium under DD, LD, and Tm conditions after 1 day of light treatment at 25°C. The black lines mark the growth fronts determined every 24 h. (B) Western blot analysis of the levels of neddylated Cul1 and unneddylated Cul1 of the WT strain, csn-3KO mutant, and csn-3KO qa-Myc-Cul1 and csn-3KO qa-Myc-Cul1K722R transformants with anti-Cul1 antibody in a reaction mixture with 0.1% glucose and 0 M, 10−3 M, or 10−2 M QA medium. (C) The ratios of neddylated Cul1 to total Cul1 of the WT strain, csn-3KO mutant, and csn-3KO qa-Myc-Cul1K722R transformants calculated from the quantitative analysis of Western blot results. Error bars represent standard deviations of the results of three experiments. (D) The ratios of neddylated Cul1 to total Cul1 of the WT strain, csn-3KO mutant, and csn-3KO qa-Myc-Cul1 transformants calculated from the quantitative analysis of Western blot results. Error bars represent standard deviations of the results of three experiments.

Western blot analyses revealed that the level of Cul1Nedd8 in the 339-59 mutant was higher than in the csn-3KO and WT strains (Fig. 2A and B). To test whether expression of Myc-Cul1K722R or Myc-Cul1 has an effect on the phenotype of the 339-59 mutant, we ectopically expressed Myc-Cul1 or Myc-Cul1K722R in 339-59 mutants and examined their growth rates and conidiation rhythms by race tube assay. As shown in Fig. S5, the 339-59 mutant grew faster and exhibited a disrupted conidiation rhythm on race tubes in minimal and QA-containing medium. With QA treatment, the 339-59 qa-Myc-Cul1K722R transformants exhibited regular conidiation rhythms under DD, LD, and Tm cycle conditions, although the 339-59 mutant itself was shown to be partially entrained by LD cycles (Fig. S5). However, expression of Myc-Cul1 failed to suppress the disrupted conidiation of 339-59 qa-Myc-Cul1 transformants. These results confirmed the hypothesis of the significance of the Cul1Nedd8/Cul1 ratio in csn-3KO and 339-59 mutants. Taking the data together, the defective phenotypes of all the csn mutants were shown to be suppressed by reduction of high Cul1Nedd8/Cul1 ratios.

Unneddylated Cul1 contributes to the stabilization of F-box proteins and the degradation of their substrates in csn mutants.Given that maintaining a proper Cul1Nedd8/Cul1 ratio is crucial for the activities of SCF ligases, we wanted to know the key function of ectopically expressed Myc-Cul1K722R in the rescued CSN deletion strains. First, we examined the stabilities of the SCF components (cullin1, Skp-1, Rbx-1, and F-box proteins) in the WT and csn-2KO strains. As shown in Fig. 6A and B, Cul1, Skp-1, and Myc-Rbx-1 remained extremely stable, with a half-life more than 12 h in both the WT and csn-2KO strains, whereas F-box protein GRR-1 was rapidly degraded in the csn-2KO strain, and similar results were observed in the csn-3KO strain (Fig. S6A). Due to the fact that the defects of conidiation rhythms and development of csn mutants can be rescued by expression of Myc-Cul1K722R, we asked whether unneddylated Cul1 contributes to the stabilization of F-box proteins. To address this issue, we examined the stability of GRR-1 in the WT, csn-2KO, and cul1K722R strains (Fig. 6C). The levels of GRR-1 in the csn-2KO mutant were dramatically lower than those in the WT strain. In contrast, GRR-1 became extremely stable in the cul1K722R strain compared to the WT strain (Fig. 6C). Like the GRR-1 protein, another F-box protein, FWD-1, was also much more stable in the cul1K722R strain (Fig. 6C), suggesting that the nonneddylatable Cul1K722R proteins contribute to stabilizing F-box proteins. To further confirm this conclusion, we performed protein degradation experiments in the transformants and found that expression of Myc-Cul1K722R significantly increased the half-life of GRR-1 (Fig. 6D and E). Similarly, in the csn-1KO cul1K722R double mutant, GRR-1 became more stable with the unneddylated Cul1 level increasing (Fig. S6B).

FIG 6
FIG 6

Expression of Myc-Cul1K722R but not Myc-Cul1 contributes to the stabilization of F-box proteins and degradation of their substrates in csn mutants. (A) Western blot analysis of the degradation of Skp-1, Cul1, F-box protein (GRR-1), and Myc–Rbx-1 in the WT and csn-2KO strains after cycloheximide (CHX) treatment (10 μg/ml). The asterisk indicates the nonspecific background bands. (B) The densitometric results of analysis of the Western blot data presented in panel A. Error bars represent standard deviations of the results of three experiments. (C) Western blot analysis showing the degradation of F-box proteins, GRR-1, and FWD-1 in the WT, csn-2KO, and cul1K722R strains after cycloheximide (CHX) treatment (10 μg/ml). (D) Western blot analysis showing the degradation of GRR-1 in the indicated strains after cycloheximide (CHX) treatment (10 μg/ml). (E) The densitometric results of analysis of the Western blot data presented in panel D. Error bars represent standard deviations of the results of three experiments. (F) Western blot analysis showing the degradation of FWD-1 and its substrate FRQ in the indicated strains after cycloheximide (CHX) treatment (10 μg/ml). The asterisk indicates the nonspecific background bands. (G) The densitometric results of analysis of the Western blot data presented in panel F. Error bars represent standard deviations of the results of three experiments.

In order to test whether the F-box proteins stabilized by Myc-Cul1K722R participate in SCF dynamic assembly, we examined the degradation of the ubiquitination substrate in the transformants. As shown in Fig. 6F and G, FRQ, the substrate of the SCFFWD-1 complex, accumulated in the csn-4KO mutant due to the instability of FWD-1 protein. However, the degradation rate of FRQ in the csn-4KO qa-Myc-Cul1K722R mutant (but not in the csn-4KO qa-Myc-Cul1 mutant) was much higher than that in the csn-4KO mutant and similar to that in the WT strain (Fig. 6F and G; see also the data for csn-5KO qa-Myc-Cul1/Cul1K722R and csn-6KO qa-Myc-Cul1/Cul1K722R in Fig. S6C and D), demonstrating that the degradation of SCF substrates can be rescued through the activity of Myc-Cul1K722R-stabilized F-box proteins to reorganize functional SCF ubiquitin ligases in csn mutants.

Unneddylated Cul1 directly binds to and stabilizes substrate adaptor modules to drive SCF assembly and activation.The previous model showed that the majority of unneddylated Cul1 is sequestered by CAND1 protein and then reutilized by a large number of substrate adaptors (26). One of its predictions is that inhibition of CAND1 activity should result in a reduction of SCF activity, the hyperneddylation state of Cul1, and early ubiquitination and degradation of F-box proteins (24). However, consistent with a previous observation in human cells (27), deletion of the cand1 gene from the Neurospora genome resulted in few phenotypic changes on race tubes under DD conditions (Fig. 7A), slightly increased neddylation levels of Cul1 (Fig. 7B), and no effect on the degradation rate of FRQ (Fig. 7C), suggesting that there is only a small amount of unneddylated Cul1 associated with CAND1 protein in Neurospora. Meanwhile, the systematic analysis of the human CRL regulatory network revealed that a large fraction of Cul1 is in a complex with Skp-1 and F-box proteins independently of their neddylation status (32). These results suggest that the large amount of unneddylated Cul1 may associate with the adaptor modules to protect F-box proteins from autoubiquitination and degradation.

FIG 7
FIG 7

Unneddylated Cul1 binds directly to and stabilizes substrate adaptor modules to drive SCF assembly and activation. (A) Race tube assays for conidiation rhythm of the WT and cand1KO strains under DD conditions. (B) Western blot analysis showing the levels of neddylated Cul1 in the WT and cand1KO strains with anti-Cul1 antibody. (C) (Top panel) Western blot analysis of the degradation of FRQ in the WT and cand1KO strains after cycloheximide (CHX) treatment (10 μg/ml). (Bottom panel) The densitometric results determined from the Western blot data. Error bars represent standard deviations of the results of three experiments. (D) Immunoprecipitation (IP) assays showing the association between Skp-1/GRR-1 and Myc-Cul1/Myc-Cul1K722R. IB, immunoblot. (E) Immunoprecipitation assay showing the association between Skp-1 and different versions of Myc-Cul1/Myc-Cul1K722R mutant proteins. (F) Race tube assays for the conidiation rhythms in the WT strain, csn-6KO mutant, and transformants with ectopically expressed Myc-Cul1/Cul1K722R mutant proteins in minimal, 10−3 M QA, and 10−2 M QA medium under DD conditions. (G) Western blot analysis showing the degradation of GRR-1 after cycloheximide (CHX) treatment (10 μg/ml). WT Cul1K722R, WT qa-Myc-Cul1K722R; Cul1, csn-6KO qa-Myc-Cul1; Cul1K722R, csn-6KO qa-Myc-Cul1K722R; Cul1M49S, csn-6KO qa-Myc-Cul1M49S; Cul1K722RM49S, csn-6KO qa-Myc-Cul1K722RM49S; Cul1R145D, csn-6KO qa-Myc-Cul1R145D; Cul1K722RR145D, csn-6KO qa-Myc-Cul1K722RR145D; Cul1M49S/R145D, csn-6KO qa-Myc-Cul1M49S/R145D; Cul1K722RM49SR/145D, csn-6KO qa-Myc-Cul1K722RM49S/R145D; Cul1NΔ, csn-6KO qa-Myc-Cul1NΔ; Cul1K722RNΔ, csn-6KO qa-Myc-Cul1K722RNΔ.

To determine the interaction among Myc-Cul1/Myc-Cul1K722R, GRR-1, and Skp-1, we performed an immunoprecipitation assay with anti-GRR-1 and anti-Skp-1 polyclonal antibodies in Myc-Cul1 and Myc-Cul1K722R transformants. The Western blot results showed that both Myc-Cul1 and Myc-Cul1K722R proteins were immunoprecipitated by GRR-1 and Skp-1 antibodies (Fig. 7D), confirming that unneddylated Cul1 proteins could also associate with Skp-1 and F-box proteins. To further investigate whether the binding of Myc-Cul1K722R to Skp-1 is required for the stabilization of F-box proteins in csn mutants, we mutated the adaptor binding sites (M49S, R145D, and M49S/R145D) (16) or deleted the region from M49 to R145 on the N terminus (NΔ) of Cul1 to disrupt the interaction between Cul1 and Skp-1 (Fig. S7B). The immunoprecipitation results showed that mutations of M49S, R145D, or M49S/R145D on the Cul1 and Cul1K722R proteins weakened the association of Cul1/Cul1K722R and Skp-1 (Fig. 7E, lanes 2, 3, and 4 and lanes 8, 9, and 10) and that this association was totally abolished by deletion of the residues from M49 to R145 (Fig. 7E, lanes 5 and 11), which was not due to the changes of expression levels or stabilities resulting from introducing the NΔ mutation into Cul1 or Cul1K722R protein (Fig. S7C and D). As shown in Fig. 7F, expression of the four versions of mutated Cul1 proteins with an unchanged lysine722 residue could not rescue the disrupted conidiation of csn-6KO mutants, which was similar to the results observed with the expression of Myc-Cul1 in csn mutants. As expected, the interrupted conidiation of csn-6KO mutants was rescued by the expression of Myc-tagged Cul1K722RM49S, Cul1K722RR145D, or Cul1K722RM49S/R145D (Fig. 7F). However, the defective growth rates and conidiation rhythm of csn-6KO mutants could not be rescued by the expression of Myc-Cul1K722RNΔ protein (Fig. 7F), which was due to the complete disruption of the interaction between Myc-Cul1K722RNΔ and Skp-1 protein (Fig. 7E), resulting in even rapider degradation of F-box protein GRR-1 than was seen in the csn mutants (Fig. 7G). These data demonstrate that expression of Myc-Cul1K722R but not Myc-Cul1 is required to stabilize F-box proteins through directly binding to the substrate adaptor modules in the csn mutants. Taken together, our data further indicate that the large amount of unneddylated Cul1 proteins maintained by functional CSN complex plays an important role in the neddylation/deneddylation cycles through directly binding to the substrate adaptor modules to keep the F-box protein pool and promote rapid initiation of SCF assembly.

DISCUSSION

Mutation of the Cul1 neddylation site results in phenotypes that are similar to csn mutant phenotypes.The Neurospora crassa genome encodes three cullin proteins, cullin1, cullin3, and cullin4, which are the most conserved cullin proteins among eukaryotes. In the present study, we used mutant analysis to show that Cul1, Cul3, or Cul4 mutation leads to different phenotypes (see Fig. S1 in the supplemental material) and that only the cul1K722R mutation strain phenotype mimics the phenotype of csn mutants on race tubes with interrupted conidiation bands under DD, LD, and Tm cycle conditions (Fig. 1F). To the best of our knowledge, this is the first systematic phenotypical and functional analysis of Cul1 protein. However, the molecular mechanisms underlying the similar phenotypes of the cul1K722R strain and the csn mutants are different. In the cul1K722R heterozygous strain, most (about 90%) of the Cul1 proteins are unneddylated and the large portion of the unneddylated Cul1 mutant proteins (Cul1K722R) stabilizes the adaptor modules, leading to excessive levels of F-box proteins in cells. Not enough Cul1 proteins can be neddylated to assemble the active SCF E3 ligases, resulting in the accumulation of substrates (Fig. 1D and E). In contrast, CSN deletion strains fail to remove Nedd8 from Cul1, leaving Cul1 in a hyperneddylation state and eventually resulting in ubiquitination and degradation of F-box proteins. Thus, in csn mutants, substrates of SCF E3 ligases also cannot be efficiently ubiquitinated due to the reduced activities of SCF complexes and accumulate in cells (24). The turnover of the corresponding SCF substrates is impaired in both the cul1K722R strain and the CSN deletion strains with the opposite Cul1 neddylation status; therefore, similar phenotypes are observed. These results confirm that the neddylation/deneddylation cycle of Cul1 is critical for the SCF activities. The known biochemical activity of CSN contrasts with its known in vivo functions, which gives rise to a CSN paradox (22, 24, 36). The analysis of the cul1K722R strain provides genetic and molecular evidence to reconcile the CSN paradox by showing that both hyperneddylation and hyponeddylation of Cul1 are detrimental to the SCF function.

Maintaining a proper Cul1Nedd8/Cul1 ratio is crucial for the stabilization of F-box proteins to organize functional SCF complexes.In the present study, we observed that the level of the neddylated Cul1 is highly controlled in normal cells (Fig. 4A and B) and that different Cul1Nedd8/Cul1 ratios correspond to different phenotypes of mutant strains on race tubes (Fig. 2). To explore the role of a proper Cul1Nedd8/Cul1 ratio in cellular physiology and development, we ectopically expressed Myc-tagged Cul1K722R in csn mutants to reduce the high Cul1Nedd8/Cul1 ratio. Surprisingly, the defective phenotypes of csn, including an irregular conidiation rhythm, impaired light regulation processes, and thermosensitive growth, were rescued (Fig. 3; Fig. S2A to E; Fig. 5A; Fig. S5) as the occupancy of unneddylated Cul1 increased (Fig. 4C to F and 5B to D). In addition, a csn-1KO cul1K722R double mutant was created to reduce the level of the endogenous neddylated Cul1 (Fig. S4A and B). As expected, the phenotypic defects of the csn-1KO strain were rescued upon mutation of the neddylation site (Fig. S4C). Consistent with our genetic suppression results, CUL3A or CUL3B loss of function results in the suppression of the pleiotropic developmental defects of an Arabidopsis csn5a-2 mutant (37). In addition, the effect of csn mutants on synapsis and crossover formation was shown to be due to increased neddylation in Caenorhabditis elegans, as reducing neddylation in these mutants can partially suppress their phenotypes (38). These observations indicate that the defective phenotypes of csn mutants in different organisms can be suppressed by reducing expression levels of cullin proteins or their neddylation.

Ectopic expression of the nonneddylatable Cul1K722R proteins rescues the phenotypic defects of csn mutants, suggesting that the impaired SCF complexes regain activity to ubiquitinate the accumulated substrates. To investigate the role of the large portion of unneddylated Cul1 proteins in SCF activity, we examined the stability of SCF components and found that only the F-box proteins, such as FWD-1 and GRR-1, were extremely unstable in csn mutants (Fig. 6A to C), whereas the expression of the Cul1K722R proteins increased their half-lives (Fig. 6D and E), and the degradation rates of FRQ were restored to the level in the WT strain due to the stabilization of FWD-1 (Fig. 6F and G), demonstrating that the nonneddylatable Cul1K722R proteins rescue the csn phenotypes through stabilizing the F-box proteins which participate in SCF dynamic assembly and promote substrate degradation. We have attempted to determine and compare the levels of stability of endogenous and ectopically expressed Cul1. However, we were unable to reliably measure those levels because the anti-Cul1 antibody exhibits nonspecific bands overlapping the Myc-Cul1 bands. Nevertheless, the hypothesis of the effects of ectopic expression of Myc-Cul1K722R is strongly supported by the results presented in Fig. S4 and S6B, which show that the signalosome defects can be rescued by replacing the endogenous Cul1 with the nonneddylated form (Cul1K722R). Finally, it was shown that the effect of Cul1K722R expression was dependent on the interaction between Cul1 and the Skp-1–F-box complex. Disrupting the Cul1–Skp-1 binding through deleting the putative Skp-1 binding domain (M49 to R145) on the Cul1 N terminus (Fig. S7B and 7E), the inducible expression of Myc-Cul1K722RNΔ protein failed to rescue the defective phenotype of the csn-6KO mutant (Fig. 7F), demonstrating the requirement of the interaction between nonneddylatable Cul1 proteins and substrate adaptor modules for the F-box protein stabilization (Fig. 7G). Previous studies suggested that CAND1 likely serves as an exchange factor for substrate receptor proteins on the Cul1-Rbx1 scaffold (39); however, our data showed that there were few changes in the neddylation state of Cul1 upon deletion of the cand1 gene and that FRQ can still be degraded in a cand1KO strain in a manner similar to that seen in the WT strain (Fig. 7A to C). Moreover, CAND1 deletion had little effect on the complementation of the csn deficiency by Myc-Cul1K722R (Fig. S7A). Perhaps the CAND1 proteins can have effects on other F-box proteins and their substrates, which is a possibility that needs to be further examined.

Last, we proposed a model of the SCF cycle based on previous CSN studies and the results of this work (Fig. 8). Our results presented above provide strong experimental support for this model and an explanation for the similarity of the phenotypes of the csn mutants and cul1K722R strains in Neurospora. In cells with functional CSN complexes, Nedd8 modification at the C terminus of Cul1 activates SCF ubiquitin ligases and promotes ubiquitination and degradation of substrates. Like CSN in other eukaryotic systems, the Neurospora CSN removes Nedd8 from Cul1 to avoid its hyperneddylation, leading to the autoubiquitination and degradation of F-box proteins. The large portion of unneddylated Cul1 proteins regenerated by CSN binds to Skp-1 and F-box proteins and maintains the levels of substrate adaptor modules, which drive SCF assembly and activation. There is only a small fraction of Cul1 associated with CAND1, and these sequestered Cul1 proteins can be utilized to organize SCF complexes through dissociation from CAND1 or shuttle from this pathway to the adaptor module pool depending on the concentration of Skp-1 in cells. In contrast to the WT strain, such an SCF assembly/disassembly cycle is disrupted in mutants with improper Cul1Nedd8/Cul1 ratios in csn mutants and cul1K722R strains, resulting in the accumulation of substrates. Therefore, the unneddylated Cul1 regenerated by CSN plays a key role in maintaining the levels of substrate adaptor modules to drive the dynamic assembly of SCF complexes.

FIG 8
FIG 8

A model of SCF assembly. The balance between neddylated Cul1 and unneddylated Cul1 is critical for the normal function of SCF ligases. Most of the Cul1 proteins in cells are unneddylated and stabilize the F-box proteins through the formation of Cul1–Skp-1–F-box complexes to maintain the adaptor module pool and promote rapid SCF assembly and activation. Disruption of the Cul1Nedd8/Cul1 balance may lead either to the degradation of F-box proteins (hyperneddylation in the csn mutants) or to the sequestration of F-box proteins into inactive complexes (hyponeddylation in the cul1K722R strain). There is only a small amount of Cul1 interacting with CAND1, and the shift between the Cul1–Skp-1–F-box and Cul1–CAND1 may depend on the level of Skp-1 in the cells.

MATERIALS AND METHODS

Strains and culture conditions.N. crassa strain 87-3 (bd a) was used as the wild-type strain in this study. The previously generated bd ku70RIP strain (40) was used as the host strain for creating the csn mutants (33) and cul3KO and cul4KO strains (41). The csn and csn his-3 strains used in the present study were previously created (33, 36). The 339-59 mutant strain was created by inserting the hph gene at the BamHI site in the middle of the CSN-4 ORF. The 301-6 (bd his-3 A) strain and csn his-3 strains were the host strains for the his-3 targeting construct transformation.

Liquid culture conditions were as previously described (42). For quinic acid (QA)-induced protein expression, 0.01 M (10−2 M) or 0.001 M (10−3 M) QA (pH 5.8) was added to liquid medium containing 1× Vogel's medium, 0.1% glucose, and 0.17% arginine (43). The medium for the race tube assays contained 1× Vogel's medium, 0.1% glucose, 0.17% arginine, 50 ng/ml biotin, and 1.5% agar as minimal medium. As for the race tube medium containing QA, 0.1% glucose is replaced with 0.01 M (10−2 M) or 0.001 M (10−3 M) QA. The medium for plating assays contained 1× Vogel's medium, 3% sucrose, and 1.5% agar with or without QA addition.

Generation of cul1K722R knock-in strain and csn-1KO cul1K722R double mutant in Neurospora crassa.To create the knock-in cassette with the cul1K722R mutation, the wild-type cul1 gene (ORF and 3′ untranslated region [3′UTR]) and 3′ flank (about 1.0 kb genomic DNA downstream from cul1 3′UTR) were cloned and inserted, respectively, into the sequences upstream and downstream of the hph hygromycin resistance gene in the plasmid pBluescript-sk-hph. The resulting plasmid, pBluescript-cul1-hph-cul13′flank, was then used as the template for in vitro site-directed mutagenesis (44) to mutate the Cul1 lysine722 to arginine with the following primers: primer cul1K722R5′ (TCATGCGTGCGCGCAAGAAGATG) and primer cul1K722R3′ (CGCGCACGCATGATACGGACGATAG) (the boldface characters indicate the neddylation-deficient mutation site).

Afterward, the mutated plasmid was linearized by single-enzyme digestion to generate a knock-in cassette as follows: the entire cul1 gene (with the mutation)-hph gene-cul13′flank. The cassette was then transformed into the bd ku70RIP strain to select for hygromycin-resistant transformants. The genomic DNA from the transformants was prepared, and PCR was used to identify transformants with hph integrated into the endogenous cul1 locus. PCR products containing the endogenous cul1 gene were sequenced to identify the heterozygous strains with the cul1K722R (AAG to CGT) mutation (40, 41). The heterozygous cul1K722R strain was obtained in this study.

In generating the csn-1KO cul1K722R double mutant, the cul1K722R knock-in cassette with ClonNAT resistance gene nat was transformed into the homozygous csn-1KO strain with hygromycin resistance to select for transformants with both ClonNAT and hygromycin resistance. DNA sequencing was used to identify the heterozygous strains with the cul1K722R (AAG to CGT) mutation.

Plasmids.Full-length ORFs and the 3′UTR for cullin1, Nedd8, and Rbx-1 protein were amplified from genomic DNA by PCR and cloned into the pqa-5Myc-6His and pqa-3Flag plasmids. The pqa-5Myc-6His-Cul1K722R plasmid was generated through site-directed mutagenesis using plasmid pqa-5Myc-6His-Cul1 as the PCR template and the primers as described above. By using the same method, we constructed the pqa-5Myc-6His-Cul3K782R and pqa-5Myc-6His-Cul4K986R plasmids with corresponding primers. All these plasmids were used for the his-3 targeting transformation in the 301-6 strain or csn his-3 strain.

Generation of antiserum against cullin1, Nedd8, Skp-1, GRR-1, FWD-1, and FRQ.The glutahione S-transferase (GST)–Cul1 (amino acids L91 to Y379), GST-Nedd8 (full length; amino acids M1 to W78), GST–Skp-1 (full length; amino acids M1 to R121), GST–GRR-1 (amino acids F550 to V781), and GST-FRQ (amino acids S249 to K315 and E359 to G766) fusion proteins were expressed in BL21 cells, and the recombinant proteins were purified and used as the antigens to generate rabbit polyclonal antisera as previously described (41, 45). The FWD-1 antiserum was generated as described previously (36).

Protein analyses.Protein extraction, Western blot analysis, protein degradation assays, and immunoprecipitation assays were performed as previously described (36, 41). For Western blot analyses, equal amounts (40 μg) of total protein were loaded in each protein lane. After electrophoresis, proteins were transferred onto a polyvinylidene difluoride (PVDF) membrane, and Western blot analysis was performed with the corresponding antiserums.

ACKNOWLEDGMENTS

We thank Chengcheng Zhang and Xuemei Cao for critically reading the manuscript and members of the laboratory for helpful discussion.

This work was supported by grants from the State Key Program of National Natural Science Foundation of China (grant no. 31330004 to Q. He) and the National Natural Science Foundation for Young Scholars of China (grant no. 31400075 to Q. Hu).

FOOTNOTES

    • Received 13 March 2017.
    • Returned for modification 26 April 2017.
    • Accepted 6 September 2017.
    • Accepted manuscript posted online 18 September 2017.

  • Supplemental material for this article may be found at https://doi.org/10.1128/MCB.00109-17.

REFERENCES

  1. 1.↵
    1. Wang ZV,
    2. Hill JA
    . 2015. Protein quality control and metabolism: bidirectional control in the heart. Cell Metab 21:215–226. doi:10.1016/j.cmet.2015.01.016.
    OpenUrlCrossRefPubMed
  2. 2.↵
    1. Ciechanover A,
    2. Kwon YT
    . 2015. Degradation of misfolded proteins in neurodegenerative diseases: therapeutic targets and strategies. Exp Mol Med 47:e147. doi:10.1038/emm.2014.117.
    OpenUrlCrossRefPubMed
  3. 3.↵
    1. Weissman AM
    . 2001. Themes and variations on ubiquitylation. Nat Rev Mol Cell Biol 2:169–178. doi:10.1038/35056563.
    OpenUrlCrossRefPubMedWeb of Science
  4. 4.↵
    1. Ambroggio XI,
    2. Rees DC,
    3. Deshaies RJ
    . 2004. JAMM: a metalloprotease-like zinc site in the proteasome and signalosome. PLoS Biol 2:E2. doi:10.1371/journal.pbio.0020002.
    OpenUrlCrossRefPubMed
  5. 5.↵
    1. Smalle J,
    2. Vierstra RD
    . 2004. The ubiquitin 26S proteasome proteolytic pathway. Annu Rev Plant Biol 55:555–590. doi:10.1146/annurev.arplant.55.031903.141801.
    OpenUrlCrossRefPubMedWeb of Science
  6. 6.↵
    1. Deshaies RJ
    . 1999. SCF and Cullin/Ring H2-based ubiquitin ligases. Annu Rev Cell Dev Biol 15:435–467. doi:10.1146/annurev.cellbio.15.1.435.
    OpenUrlCrossRefPubMedWeb of Science
  7. 7.↵
    1. Petroski MD,
    2. Deshaies RJ
    . 2005. Function and regulation of cullin-RING ubiquitin ligases. Nat Rev Mol Cell Biol 6:9–20. doi:10.1038/nrm1547.
    OpenUrlCrossRefPubMedWeb of Science
  8. 8.↵
    1. Bosu DR,
    2. Kipreos ET
    . 2008. Cullin-RING ubiquitin ligases: global regulation and activation cycles. Cell Div 3:7. doi:10.1186/1747-1028-3-7.
    OpenUrlCrossRefPubMed
  9. 9.↵
    1. Hori T,
    2. Osaka F,
    3. Chiba T,
    4. Miyamoto C,
    5. Okabayashi K,
    6. Shimbara N,
    7. Kato S,
    8. Tanaka K
    . 1999. Covalent modification of all members of human cullin family proteins by NEDD8. Oncogene 18:6829–6834. doi:10.1038/sj.onc.1203093.
    OpenUrlCrossRefPubMedWeb of Science
  10. 10.↵
    1. Pan ZQ,
    2. Kentsis A,
    3. Dias DC,
    4. Yamoah K,
    5. Wu K
    . 2004. Nedd8 on cullin: building an expressway to protein destruction. Oncogene 23:1985–1997. doi:10.1038/sj.onc.1207414.
    OpenUrlCrossRefPubMedWeb of Science
  11. 11.↵
    1. Hochstrasser M
    . 2000. Biochemistry. All in the ubiquitin family. Science 289:563–564. doi:10.1126/science.289.5479.563.
    OpenUrlFREE Full Text
  12. 12.↵
    1. Geyer R,
    2. Wee S,
    3. Anderson S,
    4. Yates J, III,
    5. Wolf DA
    . 2003. BTB/POZ domain proteins are putative substrate adaptors for cullin 3 ubiquitin ligases. Mol Cell 12:783–790. doi:10.1016/S1097-2765(03)00341-1.
    OpenUrlCrossRefPubMedWeb of Science
  13. 13.↵
    1. Ou CY,
    2. Lin YF,
    3. Chen YJ,
    4. Chien CT
    . 2002. Distinct protein degradation mechanisms mediated by Cul1 and Cul3 controlling Ci stability in Drosophila eye development. Genes Dev 16:2403–2414. doi:10.1101/gad.1011402.
    OpenUrlAbstract/FREE Full Text
  14. 14.↵
    1. Pintard L,
    2. Willis JH,
    3. Willems A,
    4. Johnson JL,
    5. Srayko M,
    6. Kurz T,
    7. Glaser S,
    8. Mains PE,
    9. Tyers M,
    10. Bowerman B,
    11. Peter M
    . 2003. The BTB protein MEL-26 is a substrate-specific adaptor of the CUL-3 ubiquitin-ligase. Nature 425:311–316. doi:10.1038/nature01959.
    OpenUrlCrossRefPubMedWeb of Science
  15. 15.↵
    1. Duda DM,
    2. Borg LA,
    3. Scott DC,
    4. Hunt HW,
    5. Hammel M,
    6. Schulman BA
    . 2008. Structural insights into NEDD8 activation of cullin-RING ligases: conformational control of conjugation. Cell 134:995–1006. doi:10.1016/j.cell.2008.07.022.
    OpenUrlCrossRefPubMedWeb of Science
  16. 16.↵
    1. Zheng N,
    2. Schulman BA,
    3. Song L,
    4. Miller JJ,
    5. Jeffrey PD,
    6. Wang P,
    7. Chu C,
    8. Koepp DM,
    9. Elledge SJ,
    10. Pagano M,
    11. Conaway RC,
    12. Conaway JW,
    13. Harper JW,
    14. Pavletich NP
    . 2002. Structure of the Cul1-Rbx1-Skp1-F boxSkp2 SCF ubiquitin ligase complex. Nature 416:703–709. doi:10.1038/416703a.
    OpenUrlCrossRefPubMedWeb of Science
  17. 17.↵
    1. Saha A,
    2. Deshaies RJ
    . 2008. Multimodal activation of the ubiquitin ligase SCF by Nedd8 conjugation. Mol Cell 32:21–31. doi:10.1016/j.molcel.2008.08.021.
    OpenUrlCrossRefPubMedWeb of Science
  18. 18.↵
    1. Schwechheimer C
    . 2001. Interactions of the COP9 signalosome with the E3 ubiquitin ligase SCFTIR1 in mediating auxin response. Science 292:1379–1382. doi:10.1126/science.1059776.
    OpenUrlAbstract/FREE Full Text
  19. 19.↵
    1. Lyapina S,
    2. Cope G,
    3. Shevchenko A,
    4. Serino G,
    5. Tsuge T,
    6. Zhou C,
    7. Wolf DA,
    8. Wei N,
    9. Shevchenko A,
    10. Deshaies RJ
    . 2001. Promotion of NEDD-CUL1 conjugate cleavage by COP9 signalosome. Science 292:1382–1385. doi:10.1126/science.1059780.
    OpenUrlAbstract/FREE Full Text
  20. 20.↵
    1. Wei N,
    2. Chamovitz DA,
    3. Deng XW
    . 1994. Arabidopsis COP9 is a component of a novel signaling complex mediating light control of development. Cell 78:117–124. doi:10.1016/0092-8674(94)90578-9.
    OpenUrlCrossRefPubMedWeb of Science
  21. 21.↵
    1. Lingaraju GM,
    2. Bunker RD,
    3. Cavadini S,
    4. Hess D,
    5. Hassiepen U,
    6. Renatus M,
    7. Fischer ES,
    8. Thomä NH
    . 2014. Crystal structure of the human COP9 signalosome. Nature 512:161–165. doi:10.1038/nature13566.
    OpenUrlCrossRefPubMedWeb of Science
  22. 22.↵
    1. Wolf DA,
    2. Zhou C,
    3. Wee S
    . 2003. The COP9 signalosome: an assembly and maintenance platform for cullin ubiquitin ligases? Nat Cell Biol 5:1029–1033. doi:10.1038/ncb1203-1029.
    OpenUrlCrossRefPubMedWeb of Science
  23. 23.↵
    1. Bosu DR,
    2. Feng H,
    3. Min K,
    4. Kim Y,
    5. Wallenfang MR,
    6. Kipreos ET
    . 2010. C. elegans CAND-1 regulates cullin neddylation, cell proliferation and morphogenesis in specific tissues. Dev Biol 346:113–126. doi:10.1016/j.ydbio.2010.07.020.
    OpenUrlCrossRefPubMed
  24. 24.↵
    1. Cope GA,
    2. Deshaies RJ
    . 2003. COP9 signalosome: a multifunctional regulator of SCF and other cullin-based ubiquitin ligases. Cell 114:663–671. doi:10.1016/S0092-8674(03)00722-0.
    OpenUrlCrossRefPubMedWeb of Science
  25. 25.↵
    1. Hotton SK,
    2. Callis J
    . 2008. Regulation of cullin RING ligases. Annu Rev Plant Biol 59:467–489. doi:10.1146/annurev.arplant.58.032806.104011.
    OpenUrlCrossRefPubMedWeb of Science
  26. 26.↵
    1. Zheng J,
    2. Yang X,
    3. Harrell JM,
    4. Ryzhikov S,
    5. Shim E,
    6. Lykke-Andersen K,
    7. Wei N,
    8. Sun H,
    9. Kobayashi R,
    10. Zhang H
    . 2002. CAND1 binds to unneddylated CUL1 and regulates the formation of SCF ubiquitin E3 ligase complex. Mol Cell 10:1519–1526. doi:10.1016/S1097-2765(02)00784-0.
    OpenUrlCrossRefPubMedWeb of Science
  27. 27.↵
    1. Chew EH,
    2. Hagen T
    . 2007. Substrate-mediated regulation of cullin neddylation. J Biol Chem 282:17032–17040. doi:10.1074/jbc.M701153200.
    OpenUrlAbstract/FREE Full Text
  28. 28.↵
    1. Liu J,
    2. Furukawa M,
    3. Matsumoto T,
    4. Xiong Y
    . 2002. NEDD8 modification of CUL1 dissociates p120 (CAND1), an inhibitor of CUL1-SKP1 binding and SCF ligases. Mol Cell 10:1511–1518. doi:10.1016/S1097-2765(02)00783-9.
    OpenUrlCrossRefPubMedWeb of Science
  29. 29.↵
    1. Zhang W,
    2. Ito H,
    3. Quint M,
    4. Huang H,
    5. Noel LD,
    6. Gray WM
    . 2008. Genetic analysis of CAND1-CUL1 interactions in Arabidopsis supports a role for CAND1-mediated cycling of the SCFTIR1 complex. Proc Natl Acad Sci U S A 105:8470–8475. doi:10.1073/pnas.0804144105.
    OpenUrlAbstract/FREE Full Text
  30. 30.↵
    1. Enchev RI,
    2. Scott DC,
    3. Da Fonseca PCA,
    4. Schreiber A,
    5. Monda JK,
    6. Schulman BA,
    7. Peter M,
    8. Morris EP
    . 2012. Structural basis for a reciprocal regulation between SCF and CSN. Cell Rep 2:616–627. doi:10.1016/j.celrep.2012.08.019.
    OpenUrlCrossRefPubMed
  31. 31.↵
    1. Emberley ED,
    2. Mosadeghi R,
    3. Deshaies RJ
    . 2012. Deconjugation of Nedd8 from Cul1 is directly regulated by Skp1-F-box and substrate, and the COP9 signalosome inhibits deneddylated SCF by a noncatalytic mechanism. J Biol Chem 287:29679–29689. doi:10.1074/jbc.M112.352484.
    OpenUrlAbstract/FREE Full Text
  32. 32.↵
    1. Bennett EJ,
    2. Rush J,
    3. Gygi SP,
    4. Harper JW
    . 2010. Dynamics of cullin-RING ubiquitin ligase network revealed by systematic quantitative proteomics. Cell 143:951–965. doi:10.1016/j.cell.2010.11.017.
    OpenUrlCrossRefPubMedWeb of Science
  33. 33.↵
    1. Wang J,
    2. Hu Q,
    3. Chen H,
    4. Zhou Z,
    5. Li W,
    6. Wang Y,
    7. Li S,
    8. He Q
    . 2010. Role of individual subunits of the Neurospora crassa CSN complex in regulation of deneddylation and stability of cullin proteins. PLoS Genet 6:e1001232. doi:10.1371/journal.pgen.1001232.
    OpenUrlCrossRefPubMed
  34. 34.↵
    1. Busch S,
    2. Eckert SE,
    3. Krappmann S,
    4. Braus GH
    . 2003. The COP9 signalosome is an essential regulator of development in the filamentous fungus Aspergillus nidulans. Mol Microbiol 49:717–730. doi:10.1046/j.1365-2958.2003.03612.x.
    OpenUrlCrossRefPubMedWeb of Science
  35. 35.↵
    1. Wei N,
    2. Deng XW
    . 2003. The COP9 signalosome. Annu Rev Cell Dev Biol 19:261–286. doi:10.1146/annurev.cellbio.19.111301.112449.
    OpenUrlCrossRefPubMedWeb of Science
  36. 36.↵
    1. He Q,
    2. Cheng P,
    3. He Q,
    4. Liu Y
    . 2005. The COP9 signalosome regulates the Neurospora circadian clock by controlling the stability of the SCFFWD-1 complex. Gene Dev 19:1518–1153. doi:10.1101/gad.1322205.
    OpenUrlAbstract/FREE Full Text
  37. 37.↵
    1. Gusmaroli G,
    2. Figueroa P,
    3. Serino G,
    4. Deng XW
    . 2007. Role of the MPN subunits in COP9 signalosome assembly and activity, and their regulatory interaction with Arabidopsis Cullin3-based E3 ligases. Plant Cell 19:564–581. doi:10.1105/tpc.106.047571.
    OpenUrlAbstract/FREE Full Text
  38. 38.↵
    1. Brockway H,
    2. Balukoff N,
    3. Dean M,
    4. Alleva B,
    5. Smolikove S
    . 2014. The CSN/COP9 signalosome regulates synaptonemal complex assembly during meiotic prophase I of Caenorhabditis elegans. PLoS Genet 10:e1004757. doi:10.1371/journal.pgen.1004757.
    OpenUrlCrossRefPubMed
  39. 39.↵
    1. Pierce NW,
    2. Lee JE,
    3. Liu X,
    4. Sweredoski MJ,
    5. Graham RL,
    6. Larimore EA,
    7. Rome M,
    8. Zheng N,
    9. Clurman BE,
    10. Hess S,
    11. Shan SO,
    12. Deshaies RJ
    . 2013. Cand1 promotes assembly of new SCF complexes through dynamic exchange of F box proteins. Cell 153:206–215. doi:10.1016/j.cell.2013.02.024.
    OpenUrlCrossRefPubMed
  40. 40.↵
    1. He Q,
    2. Cha J,
    3. He Q,
    4. Lee HC,
    5. Yang Y,
    6. Liu Y
    . 2006. CKI and CKII mediate the FREQUENCY-dependent phosphorylation of the WHITE COLLAR complex to close the Neurospora circadian negative feedback loop. Genes Dev 20:2552–2565. doi:10.1101/gad.1463506.
    OpenUrlAbstract/FREE Full Text
  41. 41.↵
    1. Zhao Y,
    2. Shen Y,
    3. Yang S,
    4. Wang J,
    5. Hu Q,
    6. Wang Y,
    7. He Q
    . 2010. Ubiquitin ligase components Cullin4 and DDB1 are essential for DNA methylation in Neurospora crassa. J Biol Chem 285:4355–4365. doi:10.1074/jbc.M109.034710.
    OpenUrlAbstract/FREE Full Text
  42. 42.↵
    1. Aronson BD,
    2. Johnson KA,
    3. Loros JJ,
    4. Dunlap JC
    . 1994. Negative feedback defining a circadian clock: autoregulation of the clock gene frequency. Science 263:1578–1584. doi:10.1126/science.8128244.
    OpenUrlAbstract/FREE Full Text
  43. 43.↵
    1. Cheng P,
    2. Yang Y,
    3. Liu Y
    . 2001. Interlocked feedback loops contribute to the robustness of the Neurospora circadian clock. Proc Natl Acad Sci U S A 98:7408–7413. doi:10.1073/pnas.121170298.
    OpenUrlAbstract/FREE Full Text
  44. 44.↵
    1. Zheng L
    . 2004. An efficient one-step site-directed and site-saturation mutagenesis protocol. Nucleic Acids Res 32:e115. doi:10.1093/nar/gnh110.
    OpenUrlCrossRefPubMed
  45. 45.↵
    1. Xu H,
    2. Wang J,
    3. Hu Q,
    4. Quan Y,
    5. Chen H,
    6. Cao Y,
    7. Li C,
    8. Wang Y,
    9. He Q
    . 2010. DCAF26, an adaptor protein of Cul4-based E3, is essential for DNA methylation in Neurospora crassa. PLoS Genet 6:e1001132. doi:10.1371/journal.pgen.1001132.
    OpenUrlCrossRefPubMed
View Abstract
Back to top
Download PDF
Citation Tools
Print

Alerts
Email

KEYWORDS

COP9 signalosome (CSN)
neddylation
cullin1 (Cul1)
unneddylated Cul1 (Cul1K722R)
SCF E3 ligases

Related Articles

Cited By...