MFE1, a member of the peroxisomal hydroxyacyl coenzyme A dehydrogenase family, affects fatty acid metabolism necessary for morphogenesis in Dictyostelium spp

Abstract

Beta-oxidation of long-chain fatty acids and branched-chain fatty acids is carried out in mammalian peroxisomes by a multifunctional enzyme (MFE) or D-bifunctional protein, with separate domains for hydroxyacyl coenzyme A (CoA) dehydrogenase, enoyl-CoA hydratase, and steroid carrier protein SCP2. We have found that Dictyostelium has a gene, mfeA, encoding MFE1 with homology to the hydroxyacyl-CoA dehydrogenase and SCP2 domains. A separate gene, mfeB, encodes MFE2 with homology to the enoyl-CoA hydratase domain. When grown on a diet of bacteria, Dictyostelium cells in which mfeA is disrupted accumulate excess cyclopropane fatty acids and are unable to develop beyond early aggregation. Axenically grown mutant cells, however, developed into normal fruiting bodies composed of spores and stalk cells. Comparative analysis of whole-cell lipid compositions revealed that bacterially grown mutant cells accumulated cyclopropane fatty acids that remained throughout the developmental stages. Such a persistent accumulation was not detected in wild-type cells or axenically grown mutant cells. Bacterial phosphatidylethanolamine that contains abundant cyclopropane fatty acids inhibited the development of even axenically grown mutant cells, while dipalmitoyl phosphatidylethanolamine did not. These results suggest that MFE1 protects the cells from the increase of the harmful xenobiotic fatty acids incorporated from their diets and optimizes cellular lipid composition for proper development. Hence, we propose that this enzyme plays an irreplaceable role in the survival strategy of Dictyostelium cells to form spores for their efficient dispersal in nature.

FIG. 1.

(A) Structures of Rat MFE2 and Dictyostelium MFE1. Rat MFE2 consists of three domains, as do other mammalian MFE2/DBPs, while Dictyostelium MFE1 consists of two domains, also found in the C. elegans counterpart. The ECH domain exists as a distinct protein MFE2. (B and C) Comparison in amino acid sequence of the HCD domain (B) and SCP2 domain (C) among Dictyostelium MFE1 (accession number: BAA94961) and worm, rat, and human counterparts (accession numbers T16638, NP077368, and P51659, respectively) by using the MegAlign program of DNASTAR.

FIG. 2.

Generation of the mfeA-null mutant. (A) A schematic view of the disruption construct and the parental gene. The disruption construct was made by the insertion of a BSR cassette into mfeA at the NdeI site. The NdeI and HindIII fragment of mfeA cDNA was used as the probe for both Southern and Northern blotting. The shaded boxes indicate the regions undetected by the probe. (B) Southern blotting was performed. Genomic DNA of wild-type and mfeA-null cells and mfeA cDNA and the disruption construct were analyzed after digestion with NsiI and HindIII. A 0.3-kb shift in band size occurred in mfeA-null cells. (C) Northern blotting was performed. Total RNA samples were extracted from wild-type (wt), mfeA-null (mfeA−), and mfeA-null cells expressing mfeA under the actin 15 promoter (A15::mfeA in mfeA−) at the time points indicated during the starvation. In the wild type, a single 1.45-kb transcript was constitutively expressed during development. In mfeA-null cells, this transcript was missing. In mfeA-null cells expressing mfeA under the actin 15 promoter, a slightly larger transcript (1.85 kb) was expressed depending on the actin 15 promoter activity. Ten micrograms of genomic DNA and total RNA was analyzed for each sample.

FIG. 3.

The development of mfeA-null cells cultivated in HL5 or on E. coli B/r. (A) Plaques formed from mfeA-null (a and c) and wild-type cells (b and d) on an E. coli lawn. Panels c and d give higher magnification of the images in the rectangles in panels a and b, respectively. Scale bars, 1 mm. All plaques formed from mfeA-null cells did not show any sign of multicellular development as seen in panels a and c. (B) Effects of culture conditions on development of the mutant cells on nonnutrient agar. The axenically grown mutant cells formed slugs (a) and fruiting bodies (b) with normal stalk cells (c) and spores (d). mfeA-null cells that had been cultivated on bacteria for 3 h often formed aberrant multicellular structures (e) that eventually developed into fruiting body-like structures with a stalk but filled with abnormal cells (f) and a globular mass atop the stalk, which consisted of a mixture of ellipsoidal cells or unhealthy amoebae (g). mfeA-null cells that had been cultivated for 24 h on bacterial lawn did not develop into multicellular structures on nonnutrient agar plates (h). Scale bars, 0.2 mm and 20 μm.

FIG. 4.

Subcellular localization of MFE1 was visualized in mfeA-null cells expressing GFP-MFE1 after axenic (a) or bacterial (c) cultivation. (b and d) Differential interference contrast images of panels a and c, respectively. Scale bar, 10 μm.

FIG. 5.

Deletion analysis of each domain of MFE1. The construct to express the gene coding for MFE1 or truncated proteins (lanes a to d) was introduced into mfeA-null cells. Phenotypes of the resulting transformants are shown (a′ to d′). Scale bar, 1 mm.

FIG. 6.

Lipid analyses. (A) Lipid droplets in the bacterially grown mfeA-null (a) and wild-type (b) cells were visualized with Nile red. Inset in panel a shows higher magnification of a single point with a white arrow. Scale bar: 20 μm. (B) TLC of total lipids in wild-type (W) and mfeA-null (M) cells after various periods of bacterial cultivation. Lipid moieties were visualized as described in Materials and Methods. The standard lipids (STD), cholesterol oleate (1), methyl oleate (2), triolein (3), oleic acid (4), and cholesterol (5) were tested. (C) Two-dimensional TLC of total lipids extracted from Dictyostelium cells cultivated for 24 h on an E. coli lawn. Left panel, wild-type cells; right panel, mfeA-null cells. Blue arrows, SEs; red arrows, TGs.

FIG. 7.

Effect of bacterial PE on the morphogenesis of mfeA-null cells. Morphogenesis of mfeA-null cells was monitored on a filter after cultivation in axenic medium HL5 supplemented with bacterial PE (a and b) or dipalmitoyl PE (c and d). It is clearly shown that bacterial PE that contains abundant CFAs suppressed the mutant development. The pictures were taken at 16 (a and c) and 22 (b and d) h of starvation. Scale bar, 0.5 mm.