Cloning and characterization of scon-3+, a new member of the Neurospora crassa sulfur regulatory system

Abstract

The sulfur regulatory system of Neurospora crassa consists of a group of sulfur-regulated structural genes (e.g., arylsulfatase) that are under coordinate control of the CYS3 positive regulator and sulfur controller (SCON) negative regulators. Here we report on the cloning of scon-3(+), which encodes a polypeptide of 171 amino acids and is a Skp1 family homolog. Repeat-induced point mutation of scon-3(+) resulted in a phenotype of constitutive expression of arylsulfatase, a phenotype consistent with other sulfur controller mutants. Northern analysis indicated that, unlike other members of the sulfur regulatory system, expression of scon-3(+) is not under the direct control of the CYS3 transcriptional activator. In particular, scon-3(+) mRNA was detectable under sulfur repressing or derepressing conditions in a Deltacys-3 mutant. In yeast, Skp1p and an F-box protein binding partner are core constituents of a class of E3 ubiquitin ligases known as SCF complexes. The N. crassa negative regulator SCON2 contains an F-box motif essential for the operation of the sulfur regulatory system and suggests a role for an SCF complex in the N. crassa sulfur regulatory system. A crucial set of experiments, by using a yeast two-hybrid approach with confirming coimmunoprecipitation assays, demonstrated that SCON3 interacts with SCON2 in a manner dependent upon the F-box motif of SCON2. The protein-protein interaction detected between SCON2 and SCON3 represents the initial demonstration in a filamentous fungus of functional interaction between putative core components of a SCF complex.

FIG. 1.

Nucleotide and predicted amino acid sequence of the scon-3⁺ gene. The sequence is shown from 1,547 nucleotides upstream of the translation start codon to 339 nucleotides downstream of the stop codon (indicated by an asterisk). The nucleotides are numbered relative to the initiator ATG codon. The predicted polypeptide sequence is given below the nucleotide sequence in single-letter code. The 5′ splice sites, the 3′ splice sites, and the internal lariat sequences within the introns are underlined. The sequence corresponding to the N. crassa transcriptional initiation site consensus is boxed. A vertical arrow indicates the major transcriptional initiation site at −72. A potential CYS3 binding site at −1100 within the scon-3⁺ promoter region is bracketed.

FIG. 2.

Determination of the transcription start site for scon-3⁺. (A) S1 nuclease analysis. The size of the S1 nuclease-protected fragment was determined by comparison with a DNA sequencing reaction generated with the same primer and template used to produce the S1 probe. An arrow indicates the protected fragment corresponding to a start site at nucleotide −72 (Fig. 1). (B) Primer extension. Primer extension products were sized by comparison against a dideoxy-sequencing reaction generated with the same primer and with pSCON3 as a template. An arrow indicates the major primer extension product corresponding to a transcriptional initiation site at −72.

FIG. 3.

SCON3 sequence homologies. The deduced amino acid sequences of N. crassa (Nc) SCON3 (AF402682), M. canis (Mc) SCONC (AF40848), A. nidulans (An) SCONC (AAB18274), and the Skp1 proteins from H. sapiens (Hs) (AAH25673), A. thaliana (At) (NP_565123), D. melanogaster (Dm) (NP_477340), S. cereviseae (Sc) (AAC49492), and S. pombe (Sp) (NP_595455) were optimally aligned by using CLUSTALW. The resulting alignment was shaded by using the BOXSHADE program. Sequences, other than SCON3, were obtained from BLASTP searches. Identical residues are shown as white on black, whereas similar residues are shown as black on gray. Brackets indicate PEST and P-loop sequences. Invariant residues are represented in the consensus line as capital letters, whereas conserved residues are represented as lowercase letters.

FIG. 3.

SCON3 sequence homologies. The deduced amino acid sequences of N. crassa (Nc) SCON3 (AF402682), M. canis (Mc) SCONC (AF40848), A. nidulans (An) SCONC (AAB18274), and the Skp1 proteins from H. sapiens (Hs) (AAH25673), A. thaliana (At) (NP_565123), D. melanogaster (Dm) (NP_477340), S. cereviseae (Sc) (AAC49492), and S. pombe (Sp) (NP_595455) were optimally aligned by using CLUSTALW. The resulting alignment was shaded by using the BOXSHADE program. Sequences, other than SCON3, were obtained from BLASTP searches. Identical residues are shown as white on black, whereas similar residues are shown as black on gray. Brackets indicate PEST and P-loop sequences. Invariant residues are represented in the consensus line as capital letters, whereas conserved residues are represented as lowercase letters.

FIG. 4.

Northern blot analysis of scon-3⁺ expression. (Left panel) Northern analysis of poly(A)⁺ RNA isolated from wild-type N. crassa grown under derepressing (lane 1) and repressing conditions (lane 2). Northern blots were prepared and probed with ³²P-labeled scon-2⁺, scon-3⁺, and am⁺ DNA. scon-2⁺ is included to demonstrate the expression pattern typically seen for genes within the N. crassa sulfur control system. The am⁺ gene served as a control to ensure comparability between samples. (Right panel) Northern analysis of poly(A)⁺ RNA isolated from the Δcys-3 strain of N. crassa grown for 24 h under derepressing (lane 3) and repressing (lane 4) conditions. Northern blots were prepared as described above and probed with ³²P-labeled scon-2⁺, scon-3⁺, and am⁺ DNA. Note that the scon-2⁺ transcript was not observed in Δcys-3. The am⁺ gene again served as a control to ensure comparability between samples.

FIG. 5.

In vivo association of SCON2 and SCON3. The Gal4 AD, an AD fusion with full-length SCON2 (residues 1 to 650), an AD fusion with the N-terminal domain of SCON2 (residues 1 to 250), or an AD fusion with the C-terminal domain of SCON2 (residues 265 to 650) were tested for their ability to interact with a Gal4 BD fusion with SCON3 in the yeast two-hybrid assay. Each patch represents an independent transformation of the yeast YRG-2 host strain expressing the indicated proteins. The various pAD-scon2 constructs or the pAD control were cotransformed with pBD-scon3 into the YRG-2 yeast host strain. Interaction between fusion proteins was assayed by their ability to induce expression of β-galactosidase on SD-Leu-Trp media augmented with X-Gal and having sucrose as the sole carbon source. Note that YRG-2 cotransformed with pBD-scon3 and either pAD-scon2 or pAD-scon2Δwd40 tested positive for expression of lacZ, indicating the presence of protein-protein interaction. In contrast, yeast cotransformed with pBD-scon3 and pAD-scon2Δfbox tested negative for induction of lacZ expression. Controls, including pAD with pBD (not shown), were all negative for interactions in the assays.

FIG. 6.

Coimmunoprecipitation of SCON2 and SCON3. The proteins were transcribed and translated in vitro from PCR products. Prior to transcription and translation, c-Myc epitope and T7 promoter sequences were added upstream of the scon-3 coding region via PCR, whereas an HA epitope tag and T7 promoter region were added upstream of the scon-2 constructs via PCR. [³⁵S]methionine was included in the translation mixture to generate the radiolabeled products: SCON3-Myc, SCON2-HA, SCON2ΔWD40-HA, and SCON2ΔFbox-HA. After translation, coimmunoprecipitation was carried out as described in Materials and Methods. After elution from the protein A-beads, 10 μl of the immunoprecipitate was loaded onto a sodium dodecyl sulfate-15% polyacrylamide gel. Lane 1, SCON3 plus c-Myc antibody; lane 2, SCON3 plus HA antibody; lane 3, SCON2 plus c-Myc antibody; lane 4, SCON2 plus HA antibody; lane 5, SCON2ΔWD40 plus c-Myc antibody; lane 6, SCON2ΔWD40 plus HA antibody; lane 7, SCON2ΔFbox plus c-Myc antibody; lane 8, SCONΔFbox plus HA antibody; lane 9, SCON3 plus SCON2 plus c-Myc antibody (note that both proteins were precipitated); lane 10, SCON3 plus SCON2 plus HA antibody (note that both proteins were precipitated); lane 11, SCON3 plus SCON2ΔWD40 plus c-Myc antibody (note that both proteins were precipitated); lane 12, SCON3 plus SCON2ΔWD40 plus HA antibody, note that both proteins were precipitated; lane 13, SCON3 plus SCON2ΔFbox plus c-Myc antibody (note that only the SCON3 band is present); lane 14, SCON3 plus SCON2ΔFbox plus HA antibody (note that only the SCON2ΔF-box band is present). Lanes 9 to 12 demonstrate coimmunoprecipitation and protein-protein interactions between SCON3 and SCON2, whereas lanes 13 and 14 demonstrate the necessity of an F-box in SCON2 for the interaction with SCON3 to occur.