CaM/BAG5/Hsc70 signaling complex dynamically regulates leaf senescence

Calcium signaling plays an essential role in plant cell physiology, and chaperone-mediated protein folding directly regulates plant programmed cell death. The Arabidopsis thaliana protein AtBAG5 (Bcl-2-associated athanogene 5) is unique in that it contains both a BAG domain capable of binding Hsc70 (Heat shock cognate protein 70) and a characteristic IQ motif that is specific for Ca2+-free CaM (Calmodulin) binding and hence acts as a hub linking calcium signaling and the chaperone system. Here, we determined crystal structures of AtBAG5 alone and in complex with Ca2+-free CaM. Structural and biochemical studies revealed that Ca2+-free CaM and Hsc70 bind AtBAG5 independently, whereas Ca2+-saturated CaM and Hsc70 bind AtBAG5 with negative cooperativity. Further in vivo studies confirmed that AtBAG5 localizes to mitochondria and that its overexpression leads to leaf senescence symptoms including decreased chlorophyll retention and massive ROS production in dark-induced plants. Mutants interfering the CaM/AtBAG5/Hsc70 complex formation leads to different phenotype of leaf senescence. Collectively, we propose that the CaM/AtBAG5/Hsc70 signaling complex plays an important role in regulating plant senescence.


Gene Expression Analysis
For semi-quantitative RT-PCR, total RNA was extracted using TRIzol reagent (Invitrogen) and treated with RNase-free DNase I (Fermentas) according to the manufacturer's instructions. Complementary DNA (cDNA) was produced using 1 g of total RNA and TransScript First-Strand cDNA Synthesis SuperMix from TransGen Biotech (Code#AE301-02) according to the manufacturer's instructions.
Semi-quantitative RT-PCR was performed using 2 l of cDNA samples. Actin2 was used as an internal control to quantify the relative transcript levels of the target genes.
For real-time RT-PCR detection, RNA extraction and transcription were performed as previously described. cDNA samples were diluted tenfold, and 2 l was used as the template. The reaction was performed using 2 l of SYBR Green Perfect mix (TaKaRa, Dalian, China) and a CFX96 Real-Time PCR Detection System (Bio-Rad). The Tip41 transcript was used as an internal control to quantify the relative transcript levels of the target genes. Each of the three biological replicates was examined in triplicate.
To generate the overexpression transgenic plants, the full-length cDNA of BAG5 was cloned into the pBA002 binary vector in the sense orientation downstream of the CaMV 35S promoter.
To generate the overexpression of BAG5 point mutation transgenic plants, the full-length cDNA of AtBAG5 with specific point mutants in BAG domain (SS) and IQ motif (IQR) was obtained, respectively. Then the fragments were cloned into the 2 pBA002 binary vector in the sense orientation downstream of the CaMV 35S promoter.
All constructs were confirmed via DNA sequencing. The recombinant plasmid was introduced into wild-type Arabidopsis using the floral-dip method 1 . Seeds were harvested and sown in soil, and Basta screening was performed. Homozygous T3 plants with single-insert were used for further analysis.

Agrobacterial Infiltration
One-month-old plants of Nicotiana benthamiana were used for agrobacterial infiltration. Agrobacterium tumefaciens cells (C58C1) containing the ProAtBAG5:BAG5-EGFP construct and the mitochondrial marker (CD3-991), respectively were cultured to OD600 = 0.6, then the bacteria were pelleted by centrifugation at 5,000 rpm, resuspended in infiltration buffer (10 mM MgCl2, 10 mM MES, pH 5.9, and 150 μM acetosyringone) to OD600 = 1, and incubated at room temperature for 3 h before infiltration. The agroinfiltrated plants were allowed to grown in growth chambers for 48 h before examined by confocal laser-scanning microscopy.

Protein Expression and Purification
Gene fragment corresponding to AtBAG5-long (residues 49-153) was PCR Human CaM was expressed in bacteria and purified as described previously 2,3 .
The Se-Met derivative protein of CaM was produced following the same protocol as that of the wild-type protein, with the exception that methionine auxotroph E.coli B834 (DE3) cells and minimal medium were used.
The ATPase domain of human Hsc70 (residues 5-381) was amplified from a human liver cDNA library and cloned into the pET-M vector (Novagen). The Hsc70 protein was expressed and purified as described previously 4 .

Crystallization and Data Collection
All crystals were grown using sitting-drop vapor-diffusion. AtBAG5-long was All data were collected on the BL19U1 beam line at Shanghai Synchrotron 5 Radiation Facility (SSRF) and processed using the HKL2000 software 5 .
Single-wavelength anomalous data were collected for Se-Met-substituted crystals at the peak wavelength for Se.

Structure Determination and Refinement
The initial phase of AtBAG5-long structure was obtained by the program PHASER 6 using molecular replacement. Our previously solved AtBAG4 structure 4 was used as the searching model. Model was built using the program of COOT 7 and refined both by CNS 8 and PHENIX 9 programs. The noncrystallographic restraints were applied through the whole refinement. Iterative cycles of positional refinement were carried out until free R factor was converged.
The initial phases of AtBAG5-long/apo-CaM complex were obtained by molecular replacement using structures of AtBAG5-long, C-and N-terminal lobe of apo-CaM as searching models. The selenomethionines in CaM was used as a guide to locate the exact position of CaM molecules. The model was built manually using COOT program and refined using CNS program. The detailed data collection and refinement statistics are summarized in Table SI