Blakeslea trispora Photoreceptors: Identification and Functional Analysis

Studies have confirmed that light-regulated carotenogenesis is prevalent in filamentous fungi, especially in mucorales. However, few investigations have been done to understand photoinduced synthesis of carotenoids and related mechanisms in B. trispora, a well-known industrial microbial strains. In the present study, three photoreceptor genes in B. trispora were cloned, expressed, and characterized by bioinformatics and photoreception analyses, and then in vivo functional analyses of these genes were constructed in M. circinelloides. The results of this study will lead to a better understanding of photoreception and light-regulated carotenoid synthesis and other physiological responses in B. trispora.

ABSTRACT

Blakeslea trispora is an industrial fungal species used for large-scale production of carotenoids. However, B. trispora light-regulated physiological processes, such as carotenoid biosynthesis and phototropism, are not fully understood. In this study, we isolated and characterized three photoreceptor genes, btwc-1a, btwc-1b, and btwc-1c, in B. trispora. Bioinformatics analyses of these genes and their protein sequences revealed that the functional domains (PAS/LOV [Per-ARNT-Sim/light-oxygen-voltage] domain and zinc finger structure) of the proteins have significant homology to those of other fungal blue-light regulator proteins expressed by Mucor circinelloides and Neurospora crassa. The photoreceptor proteins were synthesized by heterologous expression in Escherichia coli. The chromogenic groups consisting of flavin adenine dinucleotide (FAD) and flavin mononucleotide (FMN) were detected to accompany BTWC-1 proteins by using high-performance liquid chromatography (HPLC) and fluorescence spectrometry, demonstrating that the proteins may be photosensitive. The absorbance changes of the purified BTWC-1 proteins seen under dark and light conditions indicated that they were light responsive and underwent a characteristic photocycle by light induction. Site-directed mutagenesis of the cysteine residual (Cys) in BTWC-1 did not affect the normal expression of the protein in E. coli but did lead to the loss of photocycle response, indicating that Cys represents a flavin-binding domain for photon detection. We then analyzed the functions of BTWC-1 proteins by complementing btwc-1a, btwc-1b, and btwc-1c into the counterpart knockout strains of M. circinelloides for each mcwc-1 gene. Transformation of the btwc-1a complement into mcwc-1a knockout strains restored the positive phototropism, while the addition of btwc-1c complement remedied the deficiency of carotene biosynthesis in the mcwc-1c knockout strains under conditions of illumination. These results indicate that btwc-1a and btwc-1c are involved in phototropism and light-inducible carotenogenesis. Thus, btwc-1 genes share a conserved flavin-binding domain and act as photoreceptors for control of different light transduction pathways in B. trispora.

IMPORTANCE Studies have confirmed that light-regulated carotenogenesis is prevalent in filamentous fungi, especially in mucorales. However, few investigations have been done to understand photoinduced synthesis of carotenoids and related mechanisms in B. trispora, a well-known industrial microbial strains. In the present study, three photoreceptor genes in B. trispora were cloned, expressed, and characterized by bioinformatics and photoreception analyses, and then in vivo functional analyses of these genes were constructed in M. circinelloides. The results of this study will lead to a better understanding of photoreception and light-regulated carotenoid synthesis and other physiological responses in B. trispora.

INTRODUCTION

Light is a source of energy and information for many living things and can regulate the growth and development of many strains and their physiological processes (1, 2). Studies on the light-regulated process in plants have taken considerable effort and have led to the discovery of photoreceptors and photoreceptor-mediated regulation of inducible gene expression (3, 4). In the past several decades, photosensitive pathways in microorganisms have been extensively investigated, and some photoreceptor proteins have been isolated and identified (5, 6). White Collar 1 (WC-1) is the first photoreceptor protein identified in the model filamentous fungi Neurospora crassa (7). Photoinduced synthesis of carotenoids disappeared in a wc-1-deficient strain, indicating that the wc-1 gene plays a key role in light-regulated carotenoid synthesis. The WC-1 protein contains three Per-ARNT-Sim (PAS) domains (8), a zinc finger structure located at the C terminus, and a nuclear localization sequence (9). The PAS domain near the N terminus is a structure with photoreceptor characteristics and is highly similar to those of other photoreceptor proteins containing the light-oxygen-voltage (LOV) domains. Many plant, bacterial, and fungal photoreceptors containing a LOV domain act as blue-light sensors and modulate the responses of living organisms to light (10, 11). The LOV domain usually contains 100 to 150 amino acids (aa) and belongs to the PAS family (12). Many of these LOV domains have the capacity to bind to other domains and form a functional macromolecule protein(s), such as the NPH1 protein of Arabidopsis thaliana (13). Others exist as only a single domain, such as the VIVID (VVD) protein of N. crassa (14, 15). The LOV domain interacts with the chromophore in light (16), but the mechanism of photosensitivity of the LOV domain needs to be clarified. The dimer white collar complex (WCC) is a signaling molecule that triggers current known light-regulated responses. WCC acts as a blue-light sensor and transduces light signals to downstream pathways by serving as a transcription factor to directly interact with the light response element (LRE) of the target gene promoter to regulate their expression (4).

A previous report has shown the existence of light-inducible carotenogenesis in Blakeslea trispora, where the crgA gene acts as a repressor in light-dependent signaling pathways (17). However, the molecular mechanism of light-inducible carotenogenesis in B. trispora, including photoreception and light-regulated expression of structural genes for carotenoids and other target genes, has not been studied in detail. In the present study, we showed by gene cloning and expression and bioinformatics analyses that B. trispora has three photoreceptors, BTWC-1A, BTWC-1B, and BTWC-1C. Functional in vitro and in vivo analyses were also constructed for BTWC-1 proteins to demonstrate the photoreceptor characteristics and their role in regulation of different light transduction pathways in B. trispora.

RESULTS

Cloning and bioinformatics analyses of BTWC-1 proteins.

btwc-1a (GenBank accession no. MN786804), btwc-1b (GenBank accession no. MN786805), and btwc-1c (GenBank accession no. MN786806) have almost the same coding sequence length (nearly 2,000 bp) but do not have the same intron number (Fig. 1A). Phylogenetic tree data reflect the phylogenetic relationships among three BTWC-1 proteins and indicate that they are evolutionarily different (Fig. 1B). These WC-1 proteins may have originated from the same ancestor, but BTWC-1 and MCWC-1 have differentiated into three proteins (indicated as proteins 1A, 1B, and 1C). WC-1 from B. trispora has the closest relationship with its Mucor circinelloides counterpart. Both BTWC-1A and BTWC-1B contain three PAS/LOV domains, while BTWC-1C contains only two. BTWC-1A and BTWC-1C both contain one zinc finger domain; however, no zinc finger domain is present in BTWC-1B. Domain architecture alignment data reveal that the PAS/LOV domains near the N terminus of BTWC-1 are highly similar to those of other photoreceptors (Fig. 2A) (5, 18), such as the LOV domain of NCWC-1 from N. crassa, the LOV domains of MCWC-1 from M. circinelloides, the LOV2 domain of PHY3 from Adiantum capillus-veneris, and the LOV domain of Phot1 from Arabidopsis thaliana (Fig. 2B). Sequence alignment indicates that all the LOV domains contain conserved amino acid residues and that they are also present in the well-characterized phototropin segment of the fern Adiantum capillus-veneris and Arabidopsis thaliana (Fig. 2B). The PAS/LOV domain is commonly found in photoreceptors of plants and acts as a blue-light sensor that regulates the light response (19, 20). We speculated that BTWC-1 is also a photoreceptor protein and acts as a light sensor to regulate light-inducible phenotypes, such as carotenoidogenesis in B. trispora (11). The secondary structures of each LOV domain shown in Fig. 2A indicate that these proteins can fold into similar spatial structures. In addition, the conserved Cys residues marked with a vertical solid triangle in the figure were observed in BTWC-1 proteins, indicating that it can bind FAD or FMN as shown in other photoreceptors previously (8, 21). Furthermore, a zinc finger domain is present at the C terminus in BTWC-1A and BTWC-1C, while this domain was not detected in BTWC-1B (Fig. 2C).

FIG 1.

(A) Schematic representation showing the exons and introns of btwc-1 genes. The shaded and blank parts indicate exons and introns, respectively. (B) A phylogenetic tree of BTWC-1 proteins and other counterparts of Mucorales was created by the use of MEGA 7.0 using sequence alignments and the neighbor-joining method. Branch lengths are proportional to the number of substitutions per site (bars). The numbers at the nodes represent bootstrap values (percent) for 100 replications. MADA, MADA protein from Phycomyces blakesleeanus (GenBank accession no. ABB77846.1); WCOA, White Collar One A from Phycomyces blakesleeanus (GenBank accession no. ABB77844); WCOB, White Collar One B from Phycomyces blakesleeanus (GenBank accession no. CAQ76857); FFWC-1, WC-1 from Fusarium fujikuroi (GenBank accession no. KLP20339.1); NCWC-1, GenBank accession no. CAA63964; MCWC-1A, GenBank accession no. AM040841.2; MCWC-1B, GenBank accession no. AM040842.1; MCWC-1C, GenBank accession no. AM040843.1.

FIG 2.

(A) The domain layout of BTWC-1 in B. trispora compared with those of NCWC-1 and MCWC-1. The length of each protein sequence in amino acids (a.a.) is indicated on the right side. PAS, Per-Arnt-Sim domain; PAC, the C terminus of PAS; ZnF GATA, GATA type zinc-finger DNA binding domain. Low-complexity regions are marked in pink, and unknown regions are marked in green. (B) Amino acid sequence alignment of LOV domains of BTWC-1, NCWC-1, MCWC-1A, MCWC-1B, MCWC-1C, PHY3 (Adiantum capillus-veneris; GenBank accession no. BAA36192), and Phot1 (Arabidopsis thaliana phototropin 1; GenBank accession no. AAC01753). The secondary structure is shown above the sequence alignment. Open boxes indicate α-helices, and open arrows indicate β-sheets. The vertical solid inverted triangle marks Cys residues of phototropin segments that interact with chromophore FAD or FMN. Pink highlighting represents a homology level equal to 100%, and blue highlighting represents a homology level between 75% and 100%. (C) Amino acid sequence alignment of ZnF GATA domains of BTWC-1A, BTWC-1C, NCWC-1, MCWC-1A, and MCWC-1C.

Preliminary analysis of the tertiary structure of BTWC-1 was performed by the use of a homologous modeling strategy. WC-1 always functions as a dimer in the form of either a WC-1/WC-1 homodimer (Fig. 3A) or a WC-1/WC-2 heterodimer. A typical LOV domain model consists of 3 to 4 α-helices and 5 β-sheets (22), which are also present in the LOV domains of BTWC-1 (Fig. 3B and C). The LOV domain (Fig. 3B; shown in purple in Fig. 3A) is located at the N terminus of BTWC-1. The LOV domains of BTWC-1 contain 3 α-helices and 5 β-sheets, and the cysteine residues of the 3 α′A-helices marked in yellow are at positions 131, 150, and 126, respectively (Fig. 3D to F).

FIG 3.

(A) Predicted tertiary structure model of WC-1-WC-1 self-dimerized (template PDB ID: 6HMJ). The LOV domain (purple part) is located at the N terminus of the WC-1 protein. The red portion is positively charged, and the blue portion is negatively charged. (B and C) The tertiary structure model of the LOV domain (B) and a display of its plane (C). Blue boxes indicate α-helices, and red arrows indicate β-sheets. (D to F) Predicted models of the LOV domains in (D) BTWC-1A, (E) BTWC-1B, and (F) BTWC-1C. They are similar in spatial structure and contain the same functional domains. The yellow part that the arrowhead points to is residual cysteine (Cys) and is used to bind the chromophore (FAD or FMN).

Heterogenous expression and photoresponse analyses of BTWC-1 in vitro.

We constructed a recombinant plasmid, pColdII-btwc-1, to express and obtain BTWC-1 (Fig. 4A). SDS-PAGE was performed to show the expression and purification of the recombinant proteins (Fig. 4B). The purified protein was accompanied by a yellow pigment substance, indicating chromophores binding to BTWC-1. In order to clarify the components of the chromophores, the ingredients of the protein and lipids were removed using chloroform to obtain the chromophores. Using such a chemical extraction process to separate the yellow pigment demonstrates that it was noncovalently bound to BTWC-1. In the supernatant of the extracted liquid, we detected two substances with the same retention times as the FAD and FMN standards (Fig. 5A), suggesting that the BTWC-1 proteins can bind to the chromogenic groups of FAD or FMN. Using UV-visible light and fluorescence full-wavelength scanning, we analyzed the purified BTWC-1 proteins and the released chromophores in the supernatant and measured the maximum absorbance at 380 and 450 nm (Fig. 5B), which are similar to the absorption spectra of other flavin-binding LOV domains (23, 24). Furthermore, the fluorescence spectra of BTWC-1 chromophores and FAD and FMN, with excitation peaks at 370 and 450 nm (Fig. 5C) and emission peak at 520 nm (Fig. 5D), have a remarkable similarity.

FIG 4.

(A) Map of the recombinant plasmid pColdII-btwc-1a. btwc-1a was attached to the pColdII after double-enzyme digestion at SacI/SalI sites. This operation is also suitable for btwc-1b and btwc-1c. (B) Expression in E. coli BL21(DE3) and purification of BTWC-1. Proteins were separated on a 10% SDS-PAGE gel (48). The pColdII protein (lane 1) and the recombinant proteins (lane 2, lane 4, and lane 6) were purified by affinity chromatography. The arrows (lane 3, lane 5, and lane 7) indicate the purified proteins. MW, molecular weight.

FIG 5.

BTWC-1 proteins are able to reversibly bind FAD or FMN. (A) HPLC analysis of FAD (peak 1), FMN (peak 2), and BTWC-1 chromophores. FAD and FMN exhibited absorption peaks at about 25 min (peak 1) and about 27 min (peak 2), respectively. BTWC-1 chromophores show two peaks with the same retention times as FAD and FMN. (B) Absorbance spectra of the purified BTWC-1 proteins and of FAD and FMN. (C and D) Fluorescence spectra of FAD, FMN, and purified BTWC-1 chromophores at excitation (C) and emission (D). In order to make the curve easier to observe on the graph, FAD, FMN, btwc-1, and btwc-1 chromophores were diluted into different concentrations. BTWC-1 had absorption peaks at wavelengths 380 nm and 450 nm and absorption peaks at fluorescence excitation wavelengths 370 nm and 450 nm and emission at wavelength 520 nm. (E) BTWC-1 proteins exhibit a reversible light-induced absorbance change. Absorbance spectra were recorded from the purified BTWC-1 proteins after they were maintained at 4°C in a dark environment overnight (spectrum 1) and after 30 s of light induction (spectrum 2, LED source of flashlight, 2.9 W/m²) followed by dark maintained for 10 min (spectrum 3), 2 h (spectrum 4), and 5 h (red spectrum 5).

To investigate whether BTWC-1 proteins would show photoreception, the absorbance spectra of the purified BTWC-1 proteins were recorded after overnight incubation in the dark (Fig. 5E, spectrum 1) or under light-pulse conditions (2.9 W/m² for 30 s, provided by light-emitting diodes [LED]) (Fig. 5E, spectrum 2). A remarkable change of absorbance spectra was shown by the disappearance of the absorption peak at 450 nm upon light pulsing. When the purified BTWC-1 proteins were placed in the dark again, recovery of absorbance peaks at 450 nm (Fig. 5E, red spectrum 5) was similar to the recovery indicated in previous reports (23). This indicates that the LOV domains of the BTWC-1 proteins undergo a reversible photochemical reaction following the transfer from dark to light (11, 25).

The role of cysteine residual in the LOV domain.

Our results thus far indicated that BTWC-1 can form a complex with FAD and FMN chromogenic groups and initiate a photo response and photocycle. Since cysteine residues in LOV domain of the WC-1 protein have been shown previously to be a site of binding to FAD or FMN in N. crassa (26), we investigated the role of cysteine residues in the LOV domain of BTWC-1 in initiating these photodependent pathways. Cysteine residues (Cys131 in BTWC-1A, Cys150 in BTWC-1B, and Cys126 in BTWC-1C) in the LOV domain were converted into alanine (Ala) (Fig. 6C) using site-directed mutagenesis. The mutated BTWC-1^CA proteins were successfully expressed in Escherichia coli BL21(DE3) as shown in the SDS-PAGE electrophoretogram (Fig. 6B). The absorbance spectra were then determined after incubation in dark and after a 30-s light pulse (LED white light of flashlight, 2.9 W/m²) (Fig. 6C). The absorbance curve of BTWC-1^CA in the range of 300 to 600 nm showed no change, suggesting that BTWC-1 lost the capability to bind and release chromophores. These results indicate that the cysteine residues represent a site for chromophore binding and can potentially mediate the photoreception process.

FIG 6.

(A) Sequencing results from BTWC-1 gene and BTWC-1^CA following site-directed mutagenesis. TGC (Cys) residues in the α′A-helix of the LOV domain were converted into GCC (alanine), respectively. (B) Expression in E. coli BL21(DE3) and purification of BTWC-1^CA. Proteins were separated on a 10% SDS-PAGE. The pColdII protein (lane 1) and recombinant proteins (lane 2, lane 4, and lane 6) were purified by affinity chromatography. The arrows (lane 3, lane 5, and lane 7) indicate the expressed and purified proteins (BTWC-1A^CA, BTWC-1B^CA, and BTWC-1C^CA, respectively). The arrows on the right indicate the BTWC-1 proteins. (C) Absorbance spectra of purified BTWC-1^CA proteins after incubation in the dark (spectrum 1) and after a 30-s light (LED white light of flashlight, 2.9 W/m²) induction (spectrum 2).

Functional analyses of BTWC-1 in vivo.

Previous reports have shown that it is possible to express the genes of B. trispora efficiently in M. circinelloides (27, 28), suggesting that they have a close genetic relationship. The comparison of domain architectures of WC-1 from B. trispora and M. circinelloides demonstrated that they are structurally conserved, which allowed us to conduct phenotypic analyses in M. circinelloides by complementing the null wc-1 mutants MU242 (mcwc-1a knockout strain), MU244 (mcwc-1b knockout strain), and MU247 (mcwc-1c knockout strain) (9). The recombinant plasmids pCambia1303-btwc-1a, pCambia1303-btwc-1b, and pCambia1303-btwc-1c were constructed (Fig. 7A) and were subsequently electrotransformed into Agrobacterium tumefaciens LBA4404 to infect the host Mucor circinelloides MU242, MU244, and MU247. By the use of resistance selection, transformants R-MU242 (complementing btwc-1a to MU242), R-MU244 (complementing btwc-1b to MU244), and R-MU247 (complementing btwc-1c to MU247) were obtained for phenotypic analyses.

FIG 7.

(A) Map of the recombinant pCambia1303-btwc-1a plasmid. btwc-1a was attached to pCambia1303 after double-enzyme digestion at SacI/SalI sites. This operation is also suitable for btwc-1c. btwc-1b contains a variety of restriction sites, so the homologous recombination method was used to construct pCambia1303-btwc-1b at SacI/SalI sites. (B) Electrophoresis analysis of existence of btwc-1a, btwc-1b, and btwc-1c in the genome of R-MU242, R-MU244, and R-MU247. btwc-1 genes were transformed into the knockout strains by Agrobacterium tumefaciens-mediated transformation (ATMT) and had bands at about 2,000 bp. Lanes 1 to 5 represent genes from R-MU242, lanes 6 to 11 represent genes from R-MU247, and lanes 12 to 17 represent genes from R-MU244. (C) Phototropism of the wild-type (R7B), knockout (MU242, MU244, and MU247), and complemented (R-MU242, R-MU244, and R-MU247) strains under dark (D) and white light (WL) conditions (14.6 W/m²). Open arrows indicated the light direction. (D) Carotene content of wild-type, knockout, and complemented strains under dark (D) and white light (WL) conditions. The values represent means ± standard errors (bars) of results from three independent experiments.

Gel electrophoresis analysis showed that integration of btwc-1a, btwc-1b, and btwc-1c into the genomic DNA of MU242, MU244, and MU247 had occurred since the specific primers used for cloning of btwc-1a, btwc-1b, and btwc-1c are active and the molecular weights shown in the electrophoretogram correspond to the sizes of btwc-1a, btwc-1b, and btwc-1c (Fig. 7B). To validate that the integrated fragments represented btwc-1a, btwc-1b, and btwc-1c, they were sequenced (by GENEWIZ Inc., Suzhou, China). Their sequences were consistent with btwc-1a, btwc-1b, and btwc-1c (data not shown), indicating that three btwc-1 genes were imported into the genomes of the corresponding wc-1 null strains, respectively. Phenotypic analyses were then carried out by comparing the phenotype of each transformant with that of the corresponding wc-1 null strain. The mycelium of MU242 grows in all directions regardless of the presence of or absence of irradiation, while R-MU242 showed phototropism, similarly to R7B (Fig. 7C). This indicates that complementing btwc-1a into MU242 recovered the phototropism that MU242 had lost due to the knockout of mcwc-1a. In addition, high-performance liquid chromatography (HPLC) analyses showed that R-MU247 accumulated more carotenoids than MU247 (Fig. 7D) when the strains were cultured under conditions of white light illumination (LED white light of illumination incubator, 14.6 W/m²). Therefore, BTWC-1 proteins were shown to be involved in the different light transduction pathways that control carotenogenesis or that affect the positive phototropism. However, we have not characterized the phenotype of the btwc-1b-complemented transformants, and the function of the btwc-1b gene remains unclear.

DISCUSSION

Light is an important signaling factor and can regulate the growth, development, and morphogenesis of bacteria and plants. Thus far, many investigations have attempted to reveal the molecular mechanisms of light-regulated physiological processes, including light perception, light signal transduction, and target gene and specific metabolic responses. Filamentous fungi, ascomycota, basidiomycota, and mucoromycota have been investigated with regard to their light-driven pigment formation and other light response phenotypes. Among them, the model N. crassa has become a well-understood strain, wherein the photoreceptor WC-1 was first discovered in filamentous fungi (7). These photoreceptors together determine the photoreception characteristics through LOV domains and the light-driven physiological response in N. crassa. M. circinelloides has also been used to investigate light response, due to the development of molecular tools to genetically manipulate this strain. A photoreception system containing three sets of photoreceptors homologous to the WC-1 protein of N. crassa controlling distinct light response has also been reported in M. circinelloides, indicating the complexity of light-regulated transduction pathways in mucorales (9). This is similar to what was observed previously in Phycomyces blakesleeanus, in which MADA protein was shown to mediate light-regulated transduction pathways (29, 30). Although light-inducible carotenogenesis exists in B. trispora, few investigations have been carried out to clarify the process and molecular mechanism (31, 32).

In the present study, one-step reverse transcription-PCR (RT-PCR) was conducted to identify the putative photoreceptors in B. trispora and three white collar-1 like genes (btwc-1a, btwc-1b, and btwc-1c) were successfully cloned from B. trispora (Fig. 1A). Although they resemble each other very closely, they present differences in gene lengths, protein sequences, molecular weights, and function. Thus, the three btwc-1 genes are discriminated as with the identifiers 1a, 1b, and 1c. The amino acid sequences of BTWC-1A, BTWC-1B, and BTWC-1C are 657 aa, 732 aa, and 670 aa in length, respectively. Those lengths are similar to those of other WC-1 proteins and are especially similar to those of each MCWC-1 protein from M. circinelloides (Fig. 2A). However, there are still many structural differences between them. In the ascomycetes N. crassa, only one WC-1 protein (NCWC-1) can be found and it contains three PAS/LOV domains and one zinc finger domain (Fig. 2A), whereas in mucorales (B. trispora and M. circinelloides), a more complex photoreception system that includes three photoreceptor proteins emerges. The photoreceptors contain different domain layouts, suggesting that they are functionally differentiated. By comparison, we found that the domain layouts of BTWC-1A and MCWC-1A are highly consistent with that of NCWC-1 but the domain layouts of WC-1B and WC-1C in B. trispora and M. circinelloides show different styles. This difference in domain layout may lead to a completely different physiological function, as has been demonstrated in M. circinelloides (9). Interestingly, the domain layout of each BTWC-1 protein is highly identical to that of MCWC-1, suggesting that they have similar functions. The N-terminal domains of BTWC-1 proteins show high sequence similarity with the LOV domains of the fungal photoreceptors WC-1 and the plant photoreceptors PHY3 and PHOT1. Furthermore, analyses of the tertiary structure of LOV domains in BTWC-1 revealed that these LOV domains possess three α-helices and five β-sheets (Fig. 3D to F). The findings described above suggest that the LOV domains of BTWC-1 proteins share a conserved function and that that function is involved in chromophore-binding processes. This structural conservation suggests that photoreceptors may have similar mechanisms in light sensing. Since most of the zinc finger domains are involved in the regulation of gene expression (6), the existence of zinc finger domains in BTWC-1A and BTWC-1C (Fig. 2A and C) indicates that they may directly exhibit certain traits or functions in B. trispora. However, BTWC-1B may need to bind to certain proteins in the form of the chaperone involved in the regulation of gene expression in B. trispora due to the lack of a zinc finger domain (21, 33).

Although the molecular weight of each of the three BTWC-1 proteins is different from that of the NCWC-1 protein (150 kDa) of N. crassa (18), they are very similar to those of the MCWC-1 proteins from M. circinelloides. Panels A to D of Fig. 5 indicate that the WC-1 proteins may couple with the chromophores (FAD and FMN), suggesting that BTWC-1 proteins can sense light. The absorbance spectra determined for the plant blue-light receptor under light and dark conditions show a regular reversible change, i.e., a photocycle (34). The photocycle phenomenon was also observed in BTWC-1 proteins (Fig. 5E), demonstrating they can function as the receptors to capture photoprotons. Cys in the LOV domain represents an important group for the sensing by photoreceptors of the presence of light in N. crassa (23). The photocycle was inhibited (Fig. 6C) when Cys (Cys131 for 1A, Cys150 for 1B, and Cys126 for 1C) residues were changed to alanine (Ala) in the LOV domains (Fig. 6A). This indicates that a thiol group is essential for the formation of cysteinyl adduct (23, 35–37).

Since the mucorales B. trispora and M. circinelloides are closely related, many of their genes are conserved and have similar functions. One example is the light induction of carotenogenesis seen in experiments where the crgA gene from B. trispora was complemented into the crgA null mutants of M. circinelloides, leading to the recovery of the wild-type phenotype of M. circinelloides (17). Here, we show that WC-1 of M. circinelloides (MCWC-1) and that of BTWC-1 are similar in structure; thus, the wc-1 null strains of M. circinelloides were selected as the hosts for the functional analyses of the btwc-1 genes. The functional characterization of the btwc-1 genes was performed by detection of light-regulated responses. The loss of positive phototropism and reduction of β-carotene levels in MU242 and MU247 are ascribable to the lack of mcwc-1a and mcwc-1c genes, respectively (9). For the transformants, mycelia of btwc-1a-complemented strain R-MU242 showed a phototropic response similar to that shown by the R7B wild-type strain (Fig. 7C), indicating that btwc-1a was involved in positive phototropism in response to light. The btwc-1c-complemented R-MU247 strains showed a light-inducible increase in β-carotene levels compared with MU247 (Fig. 7D), suggesting that the btwc-1c gene was involved in light induction of carotenogenesis. However, phenotypic analyses did not result in a conclusive answer about the trait or function of btwc-1b gene. A previous study showed that there are two light signaling pathways that control the synthesis of carotenoids (28). crgA and wc-1 (1b and 1c) play important roles independently or in combination. Expression of the crgA gene in M. circinelloides is activated by light, since its ability to inactivate MCWC-1B through its specific monoubiquitylation and diubiquitylation was demonstrated previously (38). In the present study, there were no changes in phenotypes (phototropism and carotenoid content) after btwc-1b was complemented into MU244, which does not mean that this gene plays no role in the regulation of carotenoid synthesis. In such a situation, the function of btwc-1b remains to be investigated by combinations of studies on crgA in B. trispora if we want to clarify the whole mechanism of light-inducible phenotype. Although the three BTWC-1 proteins share similar functional domains and tertiary structures, the signal transduction pathways that they control differ. Analysis of BTWC-1 is more conducive to investigation of the physiological processes of photoinduction (39), especially to clarify the light-inducible carotenogenesis pathway in B. trispora in future studies. It may be possible to achieve high levels of carotenoid synthesis in B. trispora by controlling and regulating the light-inducible carotenogenesis pathway (40).

MATERIALS AND METHODS

Medium, strains, and culture conditions.

Wort medium was prepared by saccharification of wheat flour. Wheat flour (200 g) was added to 800 ml water, and the mixture was then heated and stirred at 50°C for 1 h and then at 72°C for 1 h and 76°C for 1 h. The hydrolysate was then filtered with gauze, and the clarified wort was prepared to make wort medium (adding 2% agar to make solid wort medium). B. trispora NRRL 2896 (-) was stored in the laboratory and grown at 25°C in wort medium for 3 to 5 days. M. circinelloides R7B, MU242, MU244, and MU247 were provided by Victoriano Garre of Universidad de Murcia and grown at 25°C in YPG medium (0.3% yeast extract, 1% tryptone, and 2% glucose) for 3 to 5 days.

Gene cloning and bioinformatics analyses.

Total RNA of B. trispora NRRL 2896 (-) was obtained by using an RNAprep Pure Plant kit (Tiangen Biotech, Beijing, China). According to the sequences of mcwc-1 (mcwc-1a [GenBank accession no. AM040841.2], mcwc-1b [GenBank accession no. AM040842.1], and mcwc-1c [GenBank accession no. AM040843.1]) (9), BLAST procedures were performed to search for the counterpart (btwc-1a, btwc-1b, and btwc-1c) sequences in the whole genome of B. trispora NRRL 2456 stored in the JGI MycoCosm database. In order to obtain the coding sequences of these genes, primers btwc-1a-F (5′-ATGTCTCAGCAATATCACAGAAACG-3′), btwc-1a-R (5′-TTACACTGTGACAGACGAATTTTCTG-3′), btwc-1b-F (5′-ATGGATCCCTTTCAATCGTTC-3′), btwc-1b-R (5′-TTATGATTGTTGAGTGGCTAAAGTATTAG-3′), btwc-1c-F (5′-ATGAATAATTCCTCAGCGTCTTATTT-3′), and btwc-1c-R (5′-TTATTTTTTGCCTTGCGCAC-3′) were designed for one-step reverse transcription-PCR (RT-PCR) using a P612-HiScript II one-step RT-PCR kit (Dye Plus) (Vazyme Biotech Co., Nanjing, China). The functional domain of each protein was predicted by the use of the online procedure SMART, and the homologous modelings of BTWC-1A, BTWC-1B, and BTWC-1C were constructed by the use of SWISS-MODEL and PyMOL (as described in reference 41) with the structure of an RNA-binding LOV receptor (PDB identifier [ID]: 6HMJ) from Nakamurella multipartita dsm 44233 as the templates.

Expression, purification, and HPLC and spectroscopic analyses.

In order to improve the expression of photoreceptors in E. coli BL21(DE3), the coding sequences of btwc-1a, btwc-1b, and btwc-1c were codon optimized and sequence synthesis was completed by GENEWIZ (Suzhou, China). The entire open reading frames (ORFs) of btwc-1 were fused into SacI/SalI site of the expression vector pColdII using a ClonExpress MultiS one-step cloning kit (Vazyme, Nanjing, China) and were then transformed into E. coli BL21(DE3). pColdII-btwc-1 can be induced to express at low temperature (15°C) and contains a His tag for purification. Induced expression of BTWC-1 were carried out at 15°C under conditions of light irradiation (fluorescent lamp in constant-temperature shaker, 2.6 W/m²) for 10 h by addition of 0.4 mM isopropyl-β-d-thiogalactoside (IPTG) at an optical density at 600 nm (OD₆₀₀) of 0.5. Thereafter, 10 nM flavin adenine dinucleotide (FAD) and 10 nM flavin mononucleotide (FMN) were added and the engineering strains were cultured for another 10 h. The cells were then collected and broken by ultrasonication using the treatments and conditions listed in Table 1 (42, 43). The lysed cells were resuspended in phosphate-buffered saline (PBS; NaCl 8 g/liter, Na₂HPO₄ 1.42 g/liter, KCl 0.2 g/liter, KH₂PO₄ 0.27 g/liter), followed by centrifugation at 8,000 × g for 10 min to remove the cell wall fragments. The supernatant described above was incubated in ice under conditions of LED light irradiation (LED source of flashlight, 2.9 W/m²) and was subsequently purified by the use of an affinity chromatography column (HisTrap HP; GE, Shanghai, China) (1 ml) (44), and then the purified proteins were mixed with the same volume of chloroform (wrapped in tin foil) for 15 min in the dark to extract the chromophores.

TABLE 1.

Conditions of ultrasonic treatment

Ultrasonic treatment	Condition
Total time	10 min
Working time	1 s
Interval time	2 s
Temp	4°C
Sensor alarm temp	10°C
Power output	70%

The supernatant of extracted liquid was analyzed by the use of an HPLC apparatus equipped with a Zorbax SB-C₁₈ column (Agilent Technologies Inc.) (5-μm pore size, 4.6 by 250 mm) (45). A 40-min one-dimensional isocratic elution procedure was conducted using 15% (vol/vol) chromatography-grade acetonitrile. Meanwhile, purified proteins were scanned using visible light (wavelength, 300 to 600 nm) and a multifunction Synergy H4 microplate reader (Biotek, Shanghai, China). The supernatant of extracted liquid was then scanned by the use of fluorescence with the excitation wavelength at 300 to 500 nm and the emission wavelength at 450 to 600 nm. Fluorescence excitation and emission spectra were obtained with an F-7000 fluorescence spectrophotometer (Hitachi, Japan).

Site-directed mutagenesis.

Primers btwc-1a-CA-F (5′-ATCGTTGGCCGTAACGCCCGCTTTCTCCAA-3′), btwc-1a-CA-R (5′-GCGTTACGGCCAACGATTTCGGTGGCAA-’3), btwc-1b-CA-F (5′-GTGATCGGCCGTAACGCCCGCTTTCTGCAA-3′), btwc-1b-CA-R (5′-GCGTTACGGCCGATCACCTCATGTGGGG-3′); btwc-1c-CA-F (5′-ATCATCGGCAAAAACGCCCGTTTTCTGCAA-3′), and btwc-1c-CA-R (5′-GCGTTTTTGCCGATGATCTCTTCGTTCT-3′) were designed for site-directed mutagenesis of cysteine residues in the LOV domain using a Fast site-directed mutagenesis kit (KM101) (Tiangen Biotech, Beijing, China). The underlined characters indicate the mutated sequences. Expression was carried out at 15°C for 20 h in light as described above.

Agrobacterium tumefaciens-mediated transformation (ATMT) and complementation.

Plasmid pCambia1303 (Fig. 7A) was selected as the vector for complementing btwc-1 into wc-1 null strains of M. circinelloides. The pCambia1303 plasmid is a dual carrier with kanamycin and hygromycin resistance and contains a CaMV 35S promoter from Cauliflower mosaic virus, a β-glucuronidase (GUS) reporter gene, and a green fluorescent protein (GFP) gene. Agrobacterium tumefaciens can transfer exogenous genes in the transfer DNA (T-DNA) region to the genomes of infected hosts. First, recombinant plasmids pCambia1303-btwc-1a, pCambia1303-btwc-1b, and pCambia1303-btwc-1c were constructed and transformed into E. coli DH5α using primers pCambia1303-btwc-1a-F (5′-CGAGCTCATGTAGCTTGGCACAACCAATTAG-3′), pCambia1303-btwc-1a-R (5′-ACGCGTCGACCATAGTCACAAGCCATAGAGTT-3′), pCambia1303-btwc-1b-F (5′-CATGATTACGAATTCGAGCTCATGGATCCCTTTCAATCGTTCC-3′), pCambia1303-btwc-1b-R (5′-CTTGCATGCCTGCAGGTCGACTTATGATTGTTGAGTGGCTAAAGTATTAG-3′), pCambia1303-btwc-1c-F (5′-CGAGCTCATGCGCTATTTCGAGATCAACT-3′), and pCambia1303-btwc-1c-R (5′-ACGCGTCGACCGCTTACGAGTAAGCATAAGAA-3′). btwc-1a and btwc-1c were attached to the pCambia1303 plasmid at the SacI and SalI sites, respectively (Fig. 7A). btwc-1b is associated with a variety of restriction sites such as BamHI, HindIII, and XhoI, so the homologous recombination method was used to construct pCambia1303-btwc-1b. The recombinant plasmids were then electrotransformed into the Agrobacterium tumefaciens LBA4404 competent strain to construct pCambia1303-btwc-1 (AtLBA4404). A 200-μl volume of pCambia1303-btwc-1 (AtLBA4404) was mixed with a 200-μl spore suspension of M. circinelloides knockout strains and was induced by the addition of acetosyringone (AS) at a final concentration of 200 μg/ml and then cocultured at 28°C and 150 rpm for 2 to 3 days (46, 47). Subsequently, 200 μl of the cocultured solution was added to YPG selection medium containing 100 μg/ml cephalosporin and 500 μg/ml hygromycin B and then cultured at 28°C for 6 days. The mycelium of resistant strains was picked up and cultured in peptone-dextrose agar (PDA) liquid medium on a shaker for 4 to 5 days. The mycelia were collected, and the genomic DNA was extracted by liquid nitrogen milling using a plant genomic DNA kit (Tiangen, Beijing, China). PCR was performed using genomic DNA of the complemented strains as a template.

Functional analyses of complemented strains.

The light illumination experiments were conducted in an illumination incubator equipped with an LED white light source (Ningbo Prandt Instrument Co., Ltd., Ningbo, China) (14.6 W/m²). Carotene extracted from mycelia was analyzed by the use of an HPLC system equipped with a TC-C₁₈ column (Agilent Technologies Inc.) (5-μm pore size, 4.6 by 250 mm). A 40-min one-dimensional isocratic elution procedure was conducted using 20% methanol and 80% acetonitrile.

Data availability.

The sequences of genes btwc-1a, btwc-1b, and btwc-1c have been deposited in the GenBank database under accession numbers MN786804, MN786805, and MN786806, respectively.