Zinc Finger Protein LipR Represses Docosahexaenoic Acid and Lipid Biosynthesis in Schizochytrium sp.

ABSTRACT

The heterotrophic marine microalgae Schizochytrium sp. is an important industrial producer of docosahexaenoic acid (DHA). Increased production of DHA and lipids in Schizochytrium sp. has been achieved by standard fermentation optimization and metabolic engineering methods; however, regulatory mechanisms for DHA and lipid biosynthesis remain unknown. In this study, the C₂H₂ zinc finger protein LipR was identified in Schizochytrium sp. ATCC 20888 by transcriptional analysis. Deletion of the lipR gene significantly (P < 0.001) increased production of total lipids and DHA by 33% and 48%, respectively. LipR repressed DHA and lipid production by directly inhibiting transcription of polyunsaturated fatty acid (PUFA) and fatty acid synthase (FAS) genes (pfa1, pfa2, pfa3, and fas). Specific binding of LipR to 9-bp recognition sequence 5′-(C/A)(A/G)CCATCTT-3′ in upstream regions of target genes was demonstrated by electrophoretic mobility shift assays (EMSAs) and DNase I footprinting assays. Expression of several key genes (acc, acl, ampD, fabD, mae, zwf, and dga1) related to levels of precursors and NADPH, and to triacylglycerol storage rate, were also directly repressed by LipR. Our findings, taken together, indicate that the evolutionarily unique regulator LipR is an essential repressor of DHA and saturated fatty acid biosynthesis in Schizochytrium sp.

IMPORTANCE Regulatory mechanisms for DHA and saturated fatty acid biosynthesis in the heterotrophic marine microalgae Schizochytrium sp. are unclear. We demonstrate here that deletion of the gene (lipR) encoding the C₂H₂ zinc finger protein LipR promotes DHA and saturated fatty acid production in this genus. LipR acts as a key repressor of such production by binding to 9-bp consensus sequence 5′-(C/A)(A/G)CCATCTT-3′ in the upstream regions of polyunsaturated fatty acid and fatty acid synthase genes (pfa1, pfa2, pfa3, and fas), and genes related to levels of precursors and NADPH (acc, acl, ampD, fabD, mae, and zwf), and to triacylglycerol storage rate (dga1). This is the first demonstration that a regulator inhibits synthesis of DHA and lipids in Schizochytrium sp. by directly controlling transcription of PUFA synthase and fas genes. Manipulation of the lipR gene provides a potential strategy for enhancing accumulation of polyunsaturated fatty acids and lipids in thraustochytrids.

INTRODUCTION

The heterotrophic marine microalgae Schizochytrium sp. has been studied extensively because of its ability to accumulate large quantities of triacylglycerols (TAGs) rich in docosahexaenoic acid (DHA, C_22:6ω-3) (1). DHA is the principal ω-3 polyunsaturated fatty acid (PUFA) in vertebrate brain membranes, and it plays essential roles in brain development and function (2). It also has clinical application in diseases such as hypertension, arteriosclerosis, arthritis, and depression because of its cardioprotective and anti-inflammatory properties (3).

In the Schizochytrium spp., DHA comprises 25 to 45% of total fatty acids (TFAs) (4), and the other major TFA components are hexadecanoic acid (C_16:0), tetradecanoic acid (C_14:0), and docosapentaenoic acid (DPA, C_22:5ω-6) (5, 6). Saturated fatty acids (C_16:0, C_14:0) are synthesized by a single large type I fatty acid synthase, formed by fusion of a fungal FAS α-subunit at the N terminus and a β-subunit at C terminus (7). Most eukaryotes synthesize 20- and 22-carbon PUFAs from saturated fatty acids via an elongation-desaturation pathway (8); in contrast, the Schizochytrium sp. utilizes a polyketide synthase-like PUFA synthase to synthesize DHA and DPA de novo (7, 9). Schizochytrium PUFA synthase is encoded by three open reading frames (ORFs), pfa1, pfa2, and pfa3 (also known as orf A, B, and C), each having multiple catalytic domains. The PUFA synthase complex apparently synthesizes DHA and DPA by catalyzing iterative processing of fatty acyl chain from malonyl-CoA, with trans-cis isomerization and enoyl reduction in certain cycles (7, 9, 10). However, precise reactions involved in the de novo DHA synthesis remain unclear. Schizochytrium sp. has a partial desaturase-elongase system, which presumably functions in scavenging intermediate fatty acids from the de novo DHA biosynthesis system or environmental PUFAs (10). There have been numerous attempts to promote DHA production (11–14), but the regulatory mechanisms for DHA and saturated fatty acid biosynthesis in Schizochytrium spp. remain unknown.

Zinc finger proteins are a major transcriptional regulator family in eukaryotes and are widely distributed in fungi, yeast, plants, and animals. They are classified into many types based on the structure of zinc-binding amino acids, e.g., Cys₂/His₂ (C₂H₂), C₃H, C₃HC₄, and C₂HC₅. C₂H₂ zinc finger (known as “classic” zinc finger) is the most common type (15). The C₂H₂ zinc finger type contains a consensus sequence (F/Y)-X-C-X_2-5-C-X₃-(F/Y)-X₅-Ψ-X₂-H-X_3-4-H, in which X represents any amino acid, and Ψ represents a hydrophobic residue (16). Some C₂H₂ zinc finger proteins play key regulatory roles in growth, development, secondary metabolism, and stress responses of yeast and filamentous fungi (17–19). C₂H₂ regulator Mig1/Mig2/Mig3 and its homologous protein CreA/CRE1 are key regulators of glucose repression in these organisms (20–22). In the presence of glucose, Mig1 mediates glucose repression by inhibiting expression of uptake and utilization genes of secondary carbon source (20, 23). In the basidiomycete Ustilago maydis, the C₂H₂ zinc finger Rua1 activates expression of biosurfactant ustilagic acid biosynthetic genes (24). In the ascomycote filamentous fungus Aspergillus nidulans, the C₂H₂ transcription factor FlbC regulates conidiospore development (25), and the C₂H₂ regulator MtfA regulates secondary metabolism and morphogenesis (26). However, functions of most C₂H₂ transcriptional regulators in fungi are unknown.

Schizochytrium spp. are an increasingly important source of DHA. Our limited knowledge of regulatory mechanisms of DHA and saturated fatty acid biosynthesis is a strategic barrier to construction of DHA high-yield strains through genetic manipulation. The present study is focused on the regulatory role of the C₂H₂ zinc finger protein LipR (lipogenesis regulator) in DHA biosynthesis, lipogenesis, and central carbon metabolism in Schizochytrium sp. ATCC 20888. Our findings indicate that LipR directly represses PUFA synthase and fas genes, the TAG accumulation gene, and key genes related to malonyl-CoA and NADPH production and thereby inhibits DHA and lipid production. We constructed a DHA high-producer strain (yield increased 48%) from ATCC 20888 by deletion of the lipR gene.

RESULTS

Deletion of the lipR gene promotes lipid accumulation in Schizochytrium sp. ATCC 20888.

C₂H₂ zinc finger regulators are common transcription factors in eukaryotes and play essential regulatory roles in a variety of physiological processes. Negative regulators of lipid accumulation are typically downregulated under low-nitrogen conditions, resulting in an increase in cellular lipid content (27, 28). We identified possible regulators of lipogenesis by measuring transcription levels of the C₂H₂ zinc finger regulator genes in ATCC 20888 after 48 or 96 h growth in liquid medium under high- (C/N ratio, 12.6) or low- (C/N, 94.3) nitrogen conditions through quantitative real-time RT-PCR (qRT-PCR). Among the tested genes, the transcription level of a putative C₂H₂ zinc finger gene (termed lipR; GenBank accession number OL687154) was significantly downregulated in C/N 94.3 relative to C/N 12.6 medium, with respective C/N 94.3 and C/N 12.6 fold changes (log₂) of −3.18 and −2.25 at 48 and 96 h, respectively (Fig. S1 in the supplemental material), suggesting that LipR may be a lipogenesis repressor. The predicted LipR molecule (466 amino acids) contained four C₂H₂ zinc fingers (amino acid sequences 88 to 110, 120 to 139, 147 to 169, and 177 to 199) at the N terminus and one PHA03247 superfamily domain (amino acids 242 to 443) at the C terminus (Table S1; Fig. S2A). The lipR gene contains no intron.

To evaluate possible regulation by LipR of lipogenesis in Schizochytrium sp., we deleted most of the lipR ORF (+221 from ATG to stop codon) in wild-type (WT) strain ATCC 20888 by homologous recombination. The selective marker ble gene was subsequently removed from the gene replacement mutant (ΔlipR::ble) using the Cre-loxP system (Fig. S2B). The resulting gene deletion mutant was termed ΔlipR (Fig. S2). In comparison with ATCC 20888, which has ∼14 lipid bodies per cell, ΔlipR had fewer (∼9) and larger lipid bodies (Fig. 1). Lipid body-staining intensity was much stronger for ΔlipR than for WT. ΔlipR and WT had similar glucose consumption rate and dry cell weight (DCW) (Table 1; Fig. 2A and B). At 120 h culture, the lipid yield of ΔlipR (16.5 g/L) was significantly higher (P < 0.001) than that of WT (12.4 g/L); it increased by 33%. Lipid content of ΔlipR reached 69.3% (wt/wt), much higher than that of WT (52.1%) (Table 1; Fig. 2C and D). Introduction of the empty vector (pPICZαA) did not affect growth and lipid accumulation of ΔlipR; while the lipR gene under its own promoter was reintroduced into ΔlipR, lipid production of the complementation strain ΔlipR/pPICZαA-lipR decreased to WT levels. These findings indicate that LipR inhibits lipid accumulation in ATCC 20888.

FIG 1.

Effect of lipR gene deletion on lipid body formation in Schizochytrium sp. (A) Imaging analyses of WT and ΔlipR. Cells were cultured 48 h and stained for neutral lipids with Nile Red dye. Scale bar, 10 μm. (B) Numbers of lipid bodies per cell. ***, P < 0.001 (Student's t test).

TABLE 1.

Fermentation characteristics of Schizochytrium strains^a

Strain	DCW (g/L)	Lipid yield (g/L)	Lipid content (%)	DHA yield (g/L)	% TFA
WT	23.8 ± 0.4a	12.4 ± 1.0b	52.1b	4.0 ± 0.1b	36.6a
ΔlipR	23.8 ± 0.6a	16.5 ± 1.4a	69.3a	5.9 ± 0.1a	40.0a
ΔlipR/pPICZαA-lipR	23.2 ± 1.2a	12.3 ± 0.2b	53.0b	4.0 ± 0.4b	37.6a
ΔlipR/pPICZαA	24.0 ± 0.3a	16.6 ± 1.6a	69.2a	5.8 ± 0.2a

^aData were analyzed by one-way ANOVA and Duncan’s multiple range test (SPSS V. 23.0). Differing lowercase letters indicate significant difference (P < 0.05) between values.

FIG 2.

Effects of lipR deletion on cell growth, lipid accumulation, and DHA production. (A to C) Time courses of fermentation profiles in WT, ΔlipR, and ΔlipR/pPICZαA-lipR cultured in fermentation medium over 120 h. The fermentation broth was collected every 24 h to determine concentration of glucose, dry cell weight (DCW; g/L), and total lipid content (% DCW). (D and E) DCW, lipid content, and DHA yield (g/L) in WT, ΔlipR, ΔlipR/pPICZαA-lipR, and empty vector control ΔlipR/pPICZαA. Values shown are mean ± SD from three replicate flasks grown in fermentation medium over 120 h. (F) Fatty acid composition (TFA, %) in WT, ΔlipR, and ΔlipR/pPICZαA-lipR cultured for 120 h. *, P < 0.05; ***, P < 0.001; NS, not significant.

DHA production is enhanced by lipR deletion.

Major fatty acid components of Schizochytrium sp. are DHA, C_16:0, DPA, and C_14:0 (5). lipR deletion resulted in increased lipid accumulation. To determine whether biosynthesis of unsaturated and saturated fatty acids was also enhanced by the lipR deletion, we measured DHA yields and fatty acid compositions at 120 h culture for WT, ΔlipR, ΔlipR/pPICZαA, and ΔlipR/pPICZαA-lipR. DHA yield was much higher (P < 0.001) for ΔlipR (5.9 g/L) or ΔlipR/pPICZαA (5.8 g/L) than for WT (4.0 g/L) or ΔlipR/pPICZαA-lipR (4.0 g/L) (Table 1; Fig. 2E). In comparison with WT, DHA yield was 48% higher for ΔlipR, greater than the increase of lipid yield (33%). The percentage of DHA in TFAs was slightly higher for ΔlipR (40.0%) than for WT (36.6%) or ΔlipR/pPICZαA-lipR (37.6%) (Fig. 2F), but the differences were not significant. These findings indicate that the lipR deletion promoted lipid production and DHA biosynthesis. Fermentation characteristics of ΔlipR/pPICZαA-lipR were similar to those of WT, suggesting that increased lipid and DHA production in ΔlipR were due to the lipR deletion.

LipR represses transcription of PUFA synthase genes and the FAS gene.

The lipR gene encodes a C₂H₂ family zinc finger protein, and the above findings demonstrate that the lipR deletion in the Schizochytrium sp. promotes lipid and DHA biosynthesis. To determine whether LipR affects lipid and DHA biosynthesis at the transcription level, we analyzed expression of PUFA synthase genes (pfa1, pfa2, and pfa3) and the FAS gene (fas) by qRT-PCR at 2 and 4 days culture. Expression of all four genes was much higher (>2.5-fold, P < 0.001) at both time points for ΔlipR than for WT, particularly at 4 days when DHA and lipids are rapidly synthesized (Fig. 3), consistent with the increased DHA and lipid production observed for ΔlipR. The lipR gene also showed significantly higher expression (P < 0.001) in ΔlipR at both time points (Fig. 3). These findings suggest that the lipR gene is negatively autoregulated and that expression of the fas and pfa genes is under negative control of LipR.

FIG 3.

qRT-PCR analysis of transcription levels of the PUFA synthase genes and fas gene in WT and ΔlipR. RNAs were isolated from cells grown in fermentation medium for 2 and 4 days. ***, P < 0.001. The actin gene was used as internal control.

LipR directly regulates expression of PUFA synthase and FAS genes.

LipR negatively regulates transcription of fas, lipR, and pfa genes as demonstrated above. To determine whether this regulation is direct, we evaluated interactions of LipR with the DIG-labeled upstream regions of these genes by electrophoretic mobility shift assays (EMSAs). LipR protein with an N-terminal His₆ tag was overexpressed in E. coli Rosetta (DE3) and purified for EMSAs (Fig. S3). His₆-LipR bound to the fas and pfa upstream regions, and retarded bands were abolished by addition of 150-fold excess of unlabeled specific probe but not nonspecific competitor probe (Fig. 4), indicating that the interactions of LipR with these upstream regions were specific. LipR also bound specifically to the lipR upstream region (Fig. 4), thus exerting feedback control of its own expression. These findings, in combination with the results of transcriptional analysis as above, demonstrate that LipR directly represses expression of pfa genes, fas, and lipR by binding to their upstream regions.

FIG 4.

EMSAs of His₆-LipR binding to upstream regions of the PUFA synthase genes and fas gene. Each lane contained 0.2 nM labeled probe and indicated quantities of His₆-LipR (50, 100, 150, or 200 nM). For specific (lane S) or nonspecific (lane N) competition assays, a 150-fold excess of unlabeled competitor DNA was used. Arrow, free probe.

Identification of LipR-binding site.

DNase I footprinting assays using His₆-LipR and a 660-bp 5-carboxyfluorescein (FAM)-labeled probe containing the pfa1 upstream region were performed to elucidate the regulatory mechanism of LipR on its target genes. A 31-bp region of the pfa1 probe was protected by 0.8 to 1.6 μM His₆-LipR (Fig. 5A). C₂H₂ zinc fingers typically recognize a continuous 3-bp DNA sequence (29). LipR contains four C₂H₂ zinc fingers, but only fingers 2, 3, and 4 have putative nucleic acid-binding sites, suggesting that LipR recognizes a 9-bp DNA sequence. The LipR-binding site was identified by mutating successive 9-bp sequences of the LipR-protected sequence in the pfa1 upstream region (probe a1). LipR bound to probes a1, a1-1m, a1-2m, and a1-4m, but not to probe a1-3m (Fig. 5B), suggesting that the third 9-bp sequence is the LipR-binding site. To confirm this possibility, the third 9-bp sequence was flanked by nonbinding sequence of probe a1-3m to produce probe a1-3, and specific binding of LipR to probe a1-3 was shown by EMSA (Fig. 5B). The LipR-binding site was thus identified as 9-bp sequence 5′-CACCATCTT-3′, located at −641 to −633 nucleotides (nt) from start codon ATG in the pfa1 upstream region (Fig. 5A). Similarly, DNase I footprinting assays of the FAM-labeled pfa2 upstream region with His₆-LipR revealed a 9-bp sequence (5′-AGCCATCTT-3′) next to the 12-bp protected region (Fig. S4), and binding of LipR to this sequence was confirmed by EMSA (Fig. 5B). Thus, the LipR-binding sites are 5′-CACCATCTT-3′ and 5′-AGCCATCTT-3′. ATCTT or GCCAT alone (which are present in probe a1-3m) were not sufficient for binding to LipR (Fig. 5B). Characteristic LipR-binding sites were also found in the upstream regions of fas, lipR, and pfa3 (Table 2 and Table S2).

FIG 5.

Determination of LipR-binding site. (A) DNase I footprinting assay of LipR on the pfa1 upstream region. Protection fluorogram was obtained with increasing amounts of His₆-LipR (control, reaction without His₆-LipR) and 0.8 nM FAM-labeled probe. Nucleotide sequence of the pfa1 upstream region is shown at bottom. Shaded box, region protected by LipR; arrow, LipR-binding site. (B) EMSAs of probe a1, probe a2, and mutated probes to identify LipR-binding site. We introduced 9-bp mutations into the LipR-protected region of probe a1 to generate mutated probes. Blue font, altered nucleotides.

TABLE 2.

LipR-binding sites in upstream regions of putative LipR target genes

Gene	Function	Putative LipR-binding site(s)a	EMSA resultb
acc	Acetyl-CoA carboxylase	TCCCTTCTT (−316, −324), CACCTACCT (−780, −788), CTCCATCCA (−805, −813), CGAGATCTT (−845, −853)	+
acl	Citrate lyase subunit beta-like protein	CATCATCAT (−559, −551), TAGCATTTT (−571, −579)	+
zwf	Glucose-6-phosphate 1-dehydrogenase	CTCCTTCTC (−136, −144), CTCCTGCTT (−570, −562), GCCCTTCTT (−585, −577), CTCCTTCAT (−609, −601)	+
dga1	Phospholipid, diacylglycerol acyltransferase	CACGATCTT (−639, −647), TATCATTTT (−838, −846), ATCCATGTT (−892, −884)	+
mae	NADP-dependent malic enzyme	CCCCTTCTT (−74, −82), CATCATCAT (−537, −546) (−540, −549) (−543, −552) (−547, −555) (−954, −962) (−957, −965) (−960, −968) (−963, −971)	+
ampD	AMP deaminase	CAACAGCTT (−202, −210), CACCAACTA (−719, −711), CACGATCCT (−901, −893), GGCCATTTT (−742, −734)	+
fabD	Polyketide synthase	CAGTATCTT (−404, −412), CCGCATCTC (−532, −524)	+
gpd1	Glycerol-3-phosphate dehydrogenase	CATCATCAT (−976, −984), CGCCATCGA (−964, −972)	−
fabF	3-Oxoacyl-acyl-carrier-protein synthase	TGTCATCTT (−268, −276) CAAATTCTT (−343, −351)	−
lipR	ZFP42	CATCATTTT (−94, −102), CTCCTTGTT (−557, −565), CGCGTTCTT (−808, −816)	+

^aNumbers indicate distance from ATG.

^b+, binding; −, no binding.

Identification of putative LipR targets.

In the ascomycote yeast Yarrowia lipolytica, the C₂H₂ zinc finger protein Mig1 indirectly inhibits lipogenesis by repressing expression of fatty acid biosynthesis-related genes acl1, me1 (mae1), gdp1, gut1, and gut2 (30). To test the possibility that LipR affects lipogenesis by regulating expression of genes other than fas and pfa genes, we scanned the 1,000-bp upstream regions of lipid metabolism-related genes in ATCC 20888 using 9-bp sequences 5′-(C/A)(A/G)CCATCTT-3′. Putative LipR-binding sites were detected in the upstream regions of acc, acl, ampD, dga1, fabD, fabF, gpd1, mae, and zwf (Table 2). Corresponding enzyme proteins acetyl-CoA carboxylase (ACC), ATP-citrate lyase (ACL), AMP deaminase (AMPD), glycerol-3-phosphate dehydrogenase (GPD1), malic enzyme (MAE), and glucose-6-phosphate dehydrogenase (ZWF) catalyze production of fatty acid precursors and NADPH for lipid biosynthesis. Malonyl-CoA/acyl carrier protein malonyltransferase (FABD) and 3-oxoacyl-ACP synthase (FABF) form malonyl-ACP and 3-oxoacyl-ACP intermediates involved in fatty acid synthesis. Diacylglycerol acyltransferase (DGA1) catalyzes production of TAG from diacylglycerol. The regulatory effect of LipR on the putative target genes was confirmed by qRT-PCR analysis of gene expression using the RNAs described in Fig. 3. Transcription of these genes (except fabF and gpd1) at 2 and 4 days was significantly higher (P < 0.001) in ΔlipR than in WT (Fig. 6A). EMSAs revealed that LipR bound specifically to the upstream regions of acc, acl, ampD, dga1, fabD, mae, and zwf, but not to those of fabF and gpd1 (Fig. 6B). These findings demonstrate that LipR directly represses expression of acc, acl, ampD, dga1, fabD, mae, and zwf in Schizochytrium sp.

FIG 6.

Confirmation of putative LipR target genes. (A) qRT-PCR analysis of transcription levels of putative LipR target genes in WT and ΔlipR. (B) EMSAs of His₆-LipR binding to upstream regions of putative LipR target genes. PCR and EMSA conditions as in Fig. 3 and 4. ***, P < 0.001; **, P < 0.01; NS, not significant.

DISCUSSION

We identified the C₂H₂ zinc finger transcription factor LipR in Schizochytrium sp. ATCC 20888. Deletion of the lipR gene enhanced lipid body formation and DHA and lipid accumulation. We propose a model of LipR-mediated DHA and lipid biosynthesis in this genus based on our findings (Fig. 7). According to this model, LipR binds to the 9-nt sequence 5′-(C/A)(A/G)CCATCTT-3′ in the upstream regions of pfa genes and fas gene and directly regulates DHA and fatty acid biosynthesis by repressing expression of the PUFA and fatty acid synthase genes. LipR also represses expression of acc, acl, ampD, dga1, fabD, mae, and zwf by binding to their upstream regions. The enzymes AMP deaminase (AMPD), ATP-citrate lyase (ACL), acetyl-CoA carboxylase (ACC), malonyl-CoA/acyl carrier protein malonyltransferase (FABD), malic enzyme (MAE), and glucose-6-phosphate dehydrogenase (ZWF) catalyze production of fatty acid precursors and NADPH for lipid synthesis, and lipid accumulation in oleaginous yeasts is upregulated by overexpression of these enzymes (31–33). Diacylglycerol acyltransferase (DGA1) catalyzes the final step of TAG synthesis by incorporating a third acyl-CoA onto the diacylglycerol backbone and transporting it into the lipid body (34). Lipogenesis in Y. lipolytica was greatly enhanced by DGA1 overexpression (31). Thus, LipR controls DHA and lipid biosynthesis by regulating expression of pfa and fas genes, supplies of precursors and NADPH, and TAG storage rate. Deletion of the lipR gene resulted in increased expression of precursor synthesis genes, pfa and fas genes, and TAG synthesis gene through derepression, leading to enhanced levels of precursors and NADPH for lipogenesis and accelerated biosynthesis and storage of DHA and saturated fatty acids, with consequent overproduction of DHA and lipids.

FIG 7.

Proposed model of LipR-mediated lipid and DHA biosynthesis in Schizochytrium sp. Red bars, direct repression of the genes by LipR.

A BLAST search revealed that LipR has high homology to a hypothetical zinc finger protein (identity 100%) of Schizochytrium sp. CCTCC M209059 and a hypothetical ZFP42 (identity, 97%; positives, 98%; cover, 97%) of Hondaea fermentalgiana, another thraustochytrid species which is phylogenetically closely related to Aurantiochytrium limacinum, although morphologically and genomically distinct (35). The zinc finger domain (ZFD) sequence of H. fermentalgiana ZFP42 is identical to that of LipR. The regulatory role of ZFP42_Hf is unknown; however, the fact that LipR and ZFP42_Hf are highly conserved proteins suggests that they have a conserved regulatory function in thraustochytrids. BLAST search also revealed that the N-terminal ZFD of LipR is homologous (identity, 52%; positives, 64%) to the C-terminal ZFD of transcription factor Yin Yang 1 (YY1) of Homo sapiens (Fig. S5A in the supplemental material). YY1 is conserved in vertebrates and invertebrates; its DNA-binding domain has been maintained for ∼600 million years without any amino acid changes through functional selection (36, 37). YY1 functions as a repressor, activator, or transcription initiator and is involved in control of cell proliferation, development, and DNA damage responses (37). LipR, similar to YY1, has four C₂H₂ zinc fingers, of which only fingers 2, 3, and 4 have putative nucleic acid-binding sites (Fig. S5B). YY1 recognizes a 9-nt consensus sequence 5′-CGCCATNTT-3′, and ∼10% of human genes have YY1-binding sites near their promoter regions (38). ZFDs of LipR and YY1 share only 52% identity; however, the residues for zinc-binding sites and putative DNA-binding sites are highly conserved. In comparison of LipR with YY1, only one amino acid difference in putative DNA-binding sites is observed for fingers 2 and 4, and two amino acid differences for finger 3 (Fig. S5B). ZFD of LipR shares more amino acid differences in putative DNA-binding sites to those of Homo sapiens YY2 (identity, 50%) and REX1 (also termed ZFP42; identity, 50%), two retroposed copies of YY1 (Fig. S5C) (36). The highly conserved residues in DNA-binding domains suggest that LipR and YY1 share similar DNA-binding sites. We identified binding sites for LipR as 5′-CACCATCTT-3′ and 5′-AGCCATCTT-3′; these are similar to the YY1-binding consensus sequence 5′-CGCCATNTT-3′, which is also recognized by LipR (Fig. 5B). According to a phylogenetic tree of LipR orthologs, this YY1-like ZFD is present in animals, oomycetes, and thraustochytrids, but not in yeast or fungi (Fig. S6). The non-ZFD portion of LipR has no homologous proteins except for H. fermentalgiana ZFP42 and Schizochytrium sp. CCTCC M209059 zinc finger protein, indicating that LipR is an evolutionarily unique zinc finger protein in thraustochytrids. Characteristic LipR-binding sites are present in the upstream regions of fas and pfa genes of genome-sequenced thraustochytrid strains (Table S2), confirming the conserved regulatory role of LipR in DHA and saturated fatty acid biosynthesis in this family.

Several regulators are involved in regulation of lipid accumulation in oleaginous microorganisms. Studies of Y. lipolytica revealed that the metabolic sensor SNF1 pathway plays a key role in the transition from growth phase to oleaginous phase (39). Deletion of Ylsnf1 significantly upregulated expression of Ylacl1 and Ylacs2, increased the cytoplasmic acetyl-CoA pool for fatty acid synthesis, and promoted lipid accumulation. The zinc finger protein Mig1 controls lipid accumulation by regulating genes related to precursor production and fatty acid degradation (30). mig1 deletion enhanced lipid accumulation, as shown by overexpression of me1 (mae1), acl1, acl2, kgd1, id, gdp1, gut1, and gut2 and reduced expression of mfe1, while acc1, fas1, and fas2 expression were not affected. Inactivation of mhy1, another zinc finger protein-encoding gene, significantly increased the lipid content of Y. lipolytica from 30.2% (wt/wt) to 43.1% by redirecting carbon flux from amino acid biosynthetic pathways to lipid biosynthetic pathways (40). Here, we found that LipR directly regulates genes not only related to fatty acid precursor production but also genes responsible for PUFA and saturated fatty acid synthesis and TAG storage.

In conclusion, the present findings indicate that evolutionarily unique C₂H₂ zinc finger protein LipR in Schizochytrium sp. directly regulates transcription of genes involved in fatty acid precursor production, PUFA synthesis, saturated fatty acid synthesis, and TAG storage. LipR is the first transcription regulator shown to directly control DHA biosynthesis by regulating expression of PUFA synthase genes. LipR homologs and their target genes are highly conserved in thraustochytrids. Manipulation of the lipR gene therefore provides a potential strategy for enhancing accumulation of unsaturated fatty acids and lipids in thraustochytrids.

MATERIALS AND METHODS

Strains and media.

The Schizochytrium sp. and E. coli strains used in this study are listed in Table 3. Schizochytrium sp. was grown on solid GPY medium (2% glucose, 1% peptone, 0.5% yeast extract, and 2% sea crystal) at 28°C. Schizochytrium transformants were selected on GPY supplemented with 40 μg/mL zeocin. For DHA production, Schizochytrium sp. was grown for 24 h in 50 mL of seed medium (3% glucose, 1% peptone, 0.5% yeast extract, and 2% sea crystal) in 250-mL flasks on a rotary shaker (230 rpm). Seed culture was inoculated (5% [vol/vol]) into 50 mL of fermentation medium [10% glucose, 1% sodium glutamate, 0.5% yeast extract, 0.39% NaCl, 0.026% KCl, 0.05% (NH₄)₂SO₄, 0.1% KH₂PO₄, 0.143% MgSO₄, 0.004% CaCl₂] in 250-mL flasks and incubated on a rotary shaker (250 rpm) for 5 days. For transcriptional analysis of C₂H₂ zinc finger regulator genes, Schizochytrium sp. ATCC 20888 was grown in 50 mL of seed medium in 250-mL flasks at 230 rpm overnight. The cells were harvested and washed twice with sterile water and inoculated into 50 mL GYN12.6 [30 g/L glucose, 5 g/L yeast nitrogen base without amino acids and (NH₄)₂SO₄, 4.5 g/L (NH₄)₂SO₄, and 20 g/L sea crystal] or GYN94.3 [30 g/L glucose, 5 g/L yeast nitrogen base without amino acids and (NH₄)₂SO₄, 0.6 g/L (NH₄)₂SO₄, and 20 g/L sea crystal] in a 250-mL flask and incubated on a rotary shaker (230 rpm) for 4 days. E. coli transformants were grown in LB medium.

TABLE 3.

Strains and plasmids used in this study

Strain or plasmid	Description	Source
Schizochytrium strain or mutant
ATCC 20888	Wild-type strain	American Type Culture Collection
ΔlipR::ble	lipR deletion mutant in which lipR is replaced by ble	This study
ΔlipR	lipR deletion mutant	This study
ΔlipR/pPICZαA-lipR	Complemented strain of ΔlipR	This study
ΔlipR/pPICZαA	ΔlipR containing empty vector pPICZαA	This study
E. coli strains
JM109	General cloning host for plasmid manipulation	Laboratory stock
Rosetta (DE3)	Host for protein overexpression	Laboratory stock
Plasmids
pPICZαA	Yeast expression vector	Invitrogen
pJN44-cre	Cre expression vector	Washington State University
pET28a(+)	Vector for His6-tagged protein overexpression in E. coli	Novagen
pPICZαA-lipR	lipR-complemented vector based on pPICZαA	This study
pET-lipR	lipR-overexpressing vector based on pET-28a(+)	This study

Transformation of Schizochytrium by electroporation.

Schizochytrium cells (∼10⁶) were grown in 50 mL of seed medium in a 250-mL flask for 24 h, harvested, washed with 30 mL of ice-cold sterile water and 30 mL of 1 M sorbitol, and suspended in 3 mL of 1 M sorbitol. We placed 5 μg of linearized plasmid DNA, and competent cells were placed in a cuvette (0.1-cm gap) for electroporation (1.5 kV, 4.5 s, twice). Cells were then incubated 4 h at 28°C, plated on GPY medium with 40 μg/mL zeocin, and grown at 28°C for transformant selection.

Construction of lipR gene deletion and complementation strain.

The lipR deletion mutant (ΔlipR) was constructed by homologous recombination (Fig. S2B in the supplemental material). The ble gene was amplified from yeast expression vector pPICZαA (Invitrogen; Carlsbad, CA, USA) using primer pair ble-loxp-Fw/ble-loxp-Rev (Table 4). Upstream (940-bp) and downstream (1,059-bp) fragments flanking the lipR gene were amplified from ATCC 20888 genomic DNA by PCR using primer pairs lipR-up-Fw/lipR-up-Rev and lipR-dw-Fw/lipR-dw-Rev, respectively (Table 4). The 5′-flanking and 3′-flanking targeting fragments were ligated to the ble gene by PCR using primer pairs lipR-up-Fw/ble-loxp-Rev and ble-loxp-Fw/lipR-dw-Rev. The resulting fusion products were transformed jointly to ATCC 20888, and transformants (ΔlipR::ble) were selected on GPY medium with 40 μg/mL zeocin. pJN44-cre (containing the cre gene) was introduced into the ΔlipR::ble mutant, and the selection marker ble was removed by Cre recombinase at loxP sites to generate ΔlipR. For complementation of ΔlipR, an 1,808-bp DNA fragment bearing the lipR ORF and its upstream region (407 bp) was amplified from ATCC 20888 genomic DNA and cloned into pPICZαA to generate pPICZαA-lipR. The resulting plasmid and pPICZαA were introduced into ΔlipR to produce the complementation strain (ΔlipR/pPICZαA-lipR) and vector control strain ΔlipR/pPICZαA.

TABLE 4.

Primers used in this study

Purpose	Primer	DNA sequence (5′–3′)a	Length (bp)
Construction of ΔlipR mutant
	lipR-up-Fw	AGGAAGAACGCGACGAAA	940
	lipR-up-Rev	CATACATTATACGAAGTTATGCGGGCCCCACTTATGTT
	lipR-dw-Fw	ATAACTTCGTATAATGTATGAACCCCACCGAAGACTTT	1,059
	lipR-dw-Rev	TAAATTTATGCAGAGAAT
	ble-loxp-Fw	ATAACTTCGTATAATGTATGCTATACGAAGTTATTTTTCTCTTTCAGTGACC	1,173
	ble-loxp-Rev	ATAACTTCGTATAATGTATGCTATACGAAGTTATTGAGAAAGCGCCACGCTT
	V1	CCGCGCGAGCCCGGCGAC	2,210
	V2	GACGGGAGCCACACCCTA
	V3	GGCGAAAAGGACTTCGTC	350
	V4	GCGGTGGCTAGCACGGTG
Complementation of ΔlipR mutant
	lipR-C-Fw	GGGGTACCCAGAAACAGAAACTCCTG (KpnI)	1,808
	lipR-C-Rev	ATAAGAATGCGGCCGCTTAAGGGCTGCTCGACATTTT (NotI)
Construction of His-tagged LipR
	His-lipR-Fw	CGGGATCCATGATGCAGCAGCAGCAG (BamHI)	1,401
	His-lipR-Rev	CCCAAGCTTTTAAGGGCTGCTCGACAT (HindIII)
Footprinting assay
	pfa1p-FAM-Fw	CGCCAGGGTTTTCCCAGTCACGACCGGACCAAGCGTCCGGGG	660
	pfa1p-Rev	TGCTCCTTTTTTTTTACG
	pfa2p-FAM-Fw	CGCCAGGGTTTTCCCAGTCACGACGGCTCATTCAATCGACCC	375
	pfa2p-Rev	TTCCGAGCGGCCATTTTG
EMSA
	accp-Fw	GGTCATCGATCCCAAGCG	563
	accp-Rev	CCTGGGTCACACCTCTTG
	aclp-Fw	ACAGATTGAGCGCCAGCC	592
	aclp-Rev	AAATCAATCCCTGCCTGC
	ampdp-Fw	AAGTCACGAAAAAGTTTG	574
	ampdp-Rev	GCGCGCTGCAACGACGTT
	dga1p-Fw	ACTTCGTCCGATCTCGCG	568
	dga1p-Rev	AGCAAAGCAAAGCAGAGC
	fabDp-Fw	GTGTCGTGTCTGCCTTCT	511
	fabDp-Rev	CGGAGCCCTTTATCACAG
	fabFp-Fw	TTGGCTCAGACGCGACAT	526
	fabFp-Rev	GGCCGCTCCTGTCGTTAC
	fasp-Fw	GCCGGCGGCGTCTGCACT	548
	fasp-Rev	GGAGGGGGGGGGGGCGCG
	gpd1p-Fw	CCTCGTTTCGACTTGCTC	497
	gpd1p-Rev	TGGCCAGCAGTTAGTCAT
	lipRp-Fw	GGCCACGCCACGCCAGGC	502
	lipRp-Rev	AGAGAGGGGACCCGTTTG
	maep-Fw	GACAACAAGCCCAAGATG	579
	maep-Rev	GGGGGAGGAATCAGGGAG
	pksap-Fw	CGGACCAAGCGTCCGGGG	660
	pksap-Rev	TGCTCCTTTTTTTTTACG
	pksbp-Fw	GGCTCATTCAATCGACCC	375
	pksbp-Rev	TTCCGAGCGGCCATTTTG
	pkscp-Fw	CGACAAGGCCGGCGGCGT	524
	pkscp-Rev	TCTCCCTCTCGTGGCTGC
	zwfp-Fw	CCGGGAAGGTGTTCTTGA	535
	zwfp-Rev	CTCTCTGTTCTGCTCTGC
	gpop-Fw	AAACGTTAGTCTTTGTCG	570
	gpop-Rev	GTTCGCCATGGCTGACAA
qRT-PCR
	actin-QP-Fw	GCGACATCAAGGAGAAGC	130
	actin-QP-Rev	GAAGGACGGCTGGAAGAG
	acc-QP-Fw	TCAAGAGCAACATTGCCG	109
	acc-QP-Rev	CTCATGTAGTCAGCGCCA
	acl-QP-Fw	CAGGGAGCGCAACTTGCT	150
	acl-QP-Rev	GAGGGCTCGCTGCCGCGA
	ampd-QP-Fw	CAAACAGAGCGTGAGTCC	115
	ampd-QP-Rev	AAGAAGTCCTTGGTGTCCC
	dga1-QP-Fw	AGACCAGCGTGAACACCC	106
	dga1-QP-Rev	TCGCCAGACGACGAGTAG
	fabDp-QP-Fw	AAGCCAATGACATCCTCG	81
	fabDp-QP-Rev	TGTAAAGGGTCTCGCTGAG
	fabFp-QP-Fw	GCGATGGCTTTGTAATGG	92
	fabFp-QP-Rev	TTCAGCAAGAATGGGAGC
	fas-QP-Fw	GCGACCAACAACACGGAC	99
	fas-QP-Rev	CTGGGACTCCACAAATCC
	gpd1-QP-Fw	TGGAGATTGGACGCTTTATC	101
	gpd1-QP-Rev	TCCGTAGCAAGTCGTGATG
	lipR-QP-Fw	CATTGCCAGCACCACGAG	90
	lipR-QP-Rev	AGGACCTTGCCGCAGAAC
	mae-QP-Fw	CAGTGTCTTCCCGTCATC	116
	mae-QP-Rev	ACAAGGTGGTCGTAGGTC
	pksa-QP-Fw	CGCCTCGGATTCACTTCG	141
	pksa-QP-Rev	GCCCTGAGCAATGTCCAC
	pksb-QP-Fw	GACATTCACCGCATTTGG	134
	pksb-QP-Rev	AACTCTTGCTGCTGGCTC
	pksc-QP-Fw	ACGATAACGACCACACCC	179
	pksc-QP-Rev	GTCAGACACAGACACGGC
	zwf-QP-Fw	CCGCATTCAGCCCAACAG	105
	zwf-QP-Rev	GGGTCGTCTCCGAGCAGG

^aUnderlining is used to indicate restriction enzyme sites.

Glucose concentration and DCW.

Glucose concentration was determined by 3,5-dinitrosalicylic acid (DNS) method (41). DCW (biomass) was determined by freeze-drying cell pellets from 40 mL broth for 36 h to constant weight.

Lipid extraction and fatty acid composition analysis.

These steps were performed as described previously (42). Freeze-dried pellets were used for lipid extraction. Approximately 0.3 g pellet was mixed with 6 mL of 4 M HCl for 30 min and incubated at 100°C for 8 min. Sixteen milliliters methanol/chloroform (1:1 [vol/vol]) was added to the mixture, mixed vigorously, and centrifuged. The lower phase was transferred to a glass tube and evaporated under a gentle nitrogen stream. Fatty acids in lipids were methylated and analyzed by gas chromatography (Wufeng; model GC522) equipped with Agilent J&W DB23 capillary column (30 m by 0.25 mm inner diameter). Nitrogen was used as the carrier gas. The injector was at 250°C. The column temperature was raised from 150 to 200°C at 4°C per min, kept at 200°C for 1 min, further raised to 250°C at 5°C per min, and kept at 250°C for 4 min.

Microscopic observation of lipid bodies in Schizochytrium cells.

A 1-mL sample from culture at 48 h was collected by centrifugation, washed, and resuspended in phosphate-buffered saline (PBS; 38.7 mM Na₂HPO₄·12H₂O, 11.3 mM NaH₂PO₄·2H₂O, and 150 mM NaCl). Cells were stained with Nile Red dye (0.5 mg/L in dimethyl sulfoxide [DMSO]) in the dark for 5 min at room temperature and observed by confocal laser scanning microscopy (excitation at 488 nm) (model TCS SP8; Leica Microsystems) with oil immersion objective.

RNA preparation and qRT-PCR analysis.

Schizochytrium cells cultured in fermentation medium for indicated times were collected, frozen in liquid nitrogen, and ground to a fine powder. Total RNA was prepared using TRIzol reagent (Tiangen Biotech; Beijing, China) as per the manufacturer’s recommendations. cDNA was synthesized by Moloney murine leukemia virus (M-MLV) (RNase H⁻; TaKaRa) with oligo(dT)18 from 4 μg total RNA. qRT-PCR analysis was performed using SYBR Green qPCR master mix by an Applied Biosystems QuantStudio 6 Flex with the primers listed in Table 4. PCR included a 10-min preincubation at 95°C, followed by 40 cycles of denaturation at 95°C for 10 s and annealing and extension at 60°C for 30 s. For each sample, three technical replicates were performed, with one negative control. Relative expression level was determined by comparative threshold cycle (C_T) method, with actin as internal control.

Expression and purification of N-terminal His₆-tagged LipR.

The coding region of lipR (1,401 bp) was amplified from ATCC 20888 cDNA using primer pair His-lipR-Fw/His-lipR-Rev and cloned into pET28a(+) to produce expression vector pET-lipR, which was then expressed in E. coli strain Rosetta (DE3). Cells were grown at 37°C to an optical density at 600 nm (OD₆₀₀) of 0.5, induced by 0.1 mM IPTG (isopropyl-β-d-thiogalactopyranoside), and incubated for 3 h at 37°C. Cell pellets were resuspended in lysis buffer (20 mM Tris, 500 mM NaCl, 5 mM imidazole, and 5% glycerol, pH 7.9) and lysed by sonication on ice. His₆-LipR was purified from supernatant using Ni²⁺-nitrilotriacetic acid (NTA) resin (Qiagen, Hilden, Germany) as per the manufacturer’s protocol.

EMSA.

EMSAs were performed using a digoxigenin (DIG) gel shift kit (Roche) according to the manufacturer’s instructions. DNA probes carrying upstream regions of target genes were prepared by PCR with corresponding primers (Table 4) and 3′ end-labeled with DIG-11-ddUTP. We incubated 0.2 nM labeled DNA probes with various quantities of His₆-LipR (50, 100, 150, or 200 nM) in a binding reaction mixture for 30 min at 25°C. Protein-bound and unbound probes were separated by electrophoresis on 5.0% native polyacrylamide gels. Binding specificity was confirmed by adding 150-fold excess of unlabeled specific or nonspecific probe to the reaction mixture. A 570-bp upstream region of glutathione peroxidase (gpo), which does not interact with LipR, was used as nonspecific probe.

DNase I footprinting assay.

DNase I footprinting assays were performed by nonradiochemical method. For determination of LipR-binding site, FAM-labeled DNA fragments containing the pfa1 (660 bp) and pfa2 (375 bp) upstream regions were prepared by PCR with FAM-labeled primers (Table 4) and purified from agarose gel. We incubated 0.8 nM FAM-labeled probe and various quantities of His₆-LipR (0.8 or 1.6 μM) a in 25 μL binding reaction mixture for 30 min at 25°C, and it was subjected to DNase I digestion. Digested DNA was purified and subjected to capillary electrophoresis, and data were analyzed using GeneMarker v.2.2.0 software program.

Phylogenetic tree construction for LipR orthologs.

Homologous sequences of 32 species were taken to construct the phylogenetic tree by MEGA using the neighbor-joining (NJ) method. Bootstrap analyses of 1,000 replicates were carried out.

Data availability.

The GenBank accession numbers for fas, pfa1, pfa2, and pfa3 are EF015632, AF378327, AF378328, and AF378329 (7, 9). The GenBank accession numbers for lipR, acc, acl, ampD, dga1, fabD, fabF, gpd1, mae, and zwf are OL687154 and OM240807 to OM240815.