A J-like protein influences fatty acid composition of chloroplast lipids in Arabidopsis

Abstract

A comprehensive understanding of the lipid and fatty acid metabolic machinery is needed for optimizing production of oils and fatty acids for fuel, industrial feedstocks and nutritional improvement in plants. T-DNA mutants in the poorly annotated Arabidopsis thaliana gene At1g08640 were identified as containing moderately high levels (50-100%) of 16∶1Δ7 and 18∶1Δ9 leaf fatty acids and subtle decreases (5-30%) of 16∶3 and 18∶3 (http://www.plastid.msu.edu/). TLC separation of fatty acids in the leaf polar lipids revealed that the chloroplastic galactolipids monogalactosyldiacylglycerol (MGDG) and digalactosyldiacylglycerol (DGDG) were the main lipid types affected by this mutation. Analysis of the inferred amino acid sequence of At1g08640 predicted the presence of a transit peptide, three transmembrane domains and an N-terminal J-like domain, and the gene was named CJD1 for Chloroplast J-like Domain 1. GFP reporter experiments and in vitro chloroplast import assays demonstrated CJD1 is a chloroplast membrane protein. Screening of an Arabidopsis cDNA library by yeast-2-hybrid (Y2H) using the J-like domain of CJD1 as bait identified a plastidial inner envelope protein (Accumulation and Replication of Chloroplasts 6, ARC6) as the primary interacting partner in the Y2H assay. ARC6 plays a central role in chloroplast division and binds CJD1 via its own J-like domain along with an adjacent conserved region whose function is not fully known. These results provide a starting point for future investigations of how mutations in CJD1 affect lipid composition.

Figure 1. T-DNA mutants in CJD1 possess altered fatty acid profiles.

A, Salk_032130C (cjd1-1) contains a T-DNA insertion in the first exon of CJD1 (At1g08640) while Salk_039694 (cjd1-2) harbors a T-DNA insertion in intron 6. T-DNA insertions are illustrated as triangles; exons, introns and untranslated regions are depicted by empty rectangles, solid lines and black rectangles, respectively. B, Semi-quantitative RT-PCR analysis shows that leaves of cjd1-1 and cjd1-2 do not accumulate detectable CJD1 transcript under the conditions tested. Wild-type plants (WT) and the elongation initiation factor 1 alpha (EFlα, GenBank accession no. X16432) were used as controls. C, and D, FAME profiles from GC-FID expressed in mol % for WT, cjd1-1 and cjd1-2. The error bars represent the standard deviation of three biological replicates and statistically significant differences relative to wild type (Student's t test P<0.01) are indicated with asterisks.

Figure 2. Analysis of Arabidopsis CJD1 inferred amino acid sequence.

Clustal W (1.83) alignment of CJD1 with selected homologues. Identical residues are depicted by black boxes while similar residues are shaded with grey boxes. The bracket delineates the predicted transit peptide, grey bars indicate predicted transmembrane domains (TM1, 2 and 3) and the double arrow defines the J-like domain. Abbreviations and GenBank Protein ID: CJD1, 18390922, Populus, Populus trichocarpa, 222862208; Vitis, Vitis vinifera, 225453038; Zea, Zea mays, 194705880; Oryza, Oryza sativa (japonica cultivar), 78708817, Ostreo, Ostreococcus lucimarinus, 145347386; Chlamy, Chlamydomonas reinhardtii, 159491044; Synecho, Synechocystis sp. (PCC6803), 16329734; Prochlo, Prochlorococcus marinus (NATL2A), 72001786.

Figure 3. CJD1 protein resides in chloroplast membranes.

A, Confocal images of Arabidopsis leaves expressing CJD1-GFP indicate that the fusion protein is targeted to chloroplasts. B, Immunoblotting of fractionated chloroplasts (membrane, P; soluble, S) probed with anti-GFP, anti-HSP70 and anti-HSP93. WT, untransformed wild-type plants; CJD1-GFP, transgenic lines expressing CJD1-GFP. C, Chloroplast import experiments with radiolabeled recombinant CJD1, CJD1-GFP, ARC6 and rubisco small subunit (SS). Chloroplasts were isolated following treatment with (+Tr) or without Trypsin (−Tr) and fractionated into membrane (p) and soluble fractions (s). TP, translation product; MW, molecular weight; m, mature protein; pr, precursor protein. D, Predicted CJD1 topology based on import assay results, published proteomics studies – and the location of putative transmembrane domains. OEM, chloroplast outer membrane; IEM, chloroplast inner membrane; IMS, chloroplast intermembrane space; Stroma, chloroplast stroma; JL, J-like domain.

Figure 4. Assay of CJD1 J-like domain as a possible co-chaperone.

A, Modular organization and classification of the different types of J proteins (I, II, III and J-like) proposed by . J, J domain; G, Glycine rich domain; Zn, Zn-finger domain; CTD, C-terminal domain. B, Results of Y2H experiments with the two Arabidopsis chloroplastic HSP70 proteins and CJD1 J-like domain or atDjA24 HSP40 co-chaperone J-domain. C, The J-like domain of CJD1 does not rescue the temperature sensitivity of an E. coli dnaJ/cbpA double knockout mutant. The empty vector and CJD1_60–164 transformed mutants were viable at 39°C, but inviable at 42°C, while the cells transformed with the full E. coli DnaJ protein and only the J domain were viable at both temperatures. Cells were spotted on LB media supplemented with 0.5% w/v arabinose and 20 µg/ml ampicillin.

Figure 5. Schematic representation of selected yeast 2-hybrid clones and results.

TP, transit peptide; Roman numerals, transmembrane domains; CR, conserved region; PDV2, PDV2 binding domain. A, CJD1 protein. The amino terminal soluble portion (amino acids 60–164) was used as the bait. This peptide includes the J-like domain (amino acids 74–135). B, ARC6 clones and Y2H results. Top drawing: modular organization of ARC6. Second drawing: representative clone identified by library screening. This clone has the full J-like domain and the shortest CR domain of the six clones recovered (see Table 2 for details of other clones). Constructs defined by bracket: results from directed Y2H screening of truncated ARC6 proteins with positive and negative results indicated with ‘+’ and ‘−’, respectively.