Thylakoid-bound FNR can be sub-divided into fractions with differential binding affinities. Washing with salt readily releases a fraction of the enzyme, whereas another, stronger bound sub-pool is not affected by such a treatment (Matthijs et al., 1986). Comparison of the dissociation of FNR from isolated thylakoids of WT and tic62 mutant plants using high ionic- strength buffer was employed to investigate which of those fractions contains the Tic62- bound FNR, possibly allowing to extract information about the binding mode. For quantification of the amount of solubilized enzyme, a fraction of the supernatant after high- salt wash was used in Cyt c-reduction assays (Figure 29 C). In parallel, supernatant and pellet were immunoblotted and probed with antibodies against FNR (Figure 29 A/B).
As the presented data demonstrate, the Cyt c-activity of the high-salt washes prepared from tic62 thylakoids was ~ 70% of WT level, consistent with the amount of FNR in the supernatant as determined by immunoblotting (compare Figure 29 B and C). Provided that
0 4 8 12 16 20 Cyt C -red uctio n [µM/min x mg protein -1 ] WT tic62 0 4 8 12 16 20
C
FNR S 10 6.6 3.3 10 5 µg WT tic62 FNR P 20 10 5 20 10 5 µg WT tic62A
0 20 40 60 80 100 120 140 NaCl-Sup PB
S P 0 20 40 60 80 100 120 FNR amount [% ] 140 WT tic62Figure 29: The interaction of Tic62 and FNR is high-salt insensitive. (A) FNR is more readily washed from tic62 thylakoids than from WT using high salt concentrations. Isolated thylakoids of WT and tic62 plants were washed with high ionic-strength buffer (0.5 M NaCl in Hepes/MgCl2 buffer) in the dark, and membrane and soluble fractions were separated by centrifugation (5 min, 5,000 x g). Shown is a representative immunoblot of supernatant (S) and pellet (P) fractions obtained from WT and tic62 thylakoids, using a dilution series of protein and Chl concentrations and probed with
FNR antibody. (B) Additionally, the quantification of the FNR amount in WT (white; 100%) and tic62 (grey) supernatant and pellet signals from the immunoblots is shown. The dotted line represents the amount of FNR detected in native tic62 thylakoids. Standard error bars are included; the experiment was performed in triplicate. (C) The FNR activity in the supernatant was determined by Fd-dependent Cyt c-reduction, monitored with a spectrophotometer at 550 nm. The experiment was performed in triplicate, standard error bars are included.
FNR was equally well solubilized from the thylakoids in all samples, the resulting activity in the supernatant should match the difference in total quantity present in the membrane. This was estimated to ~ 50% in both mutant lines compared to the WT (see Figure 17 A), indicating that FNR was more readily washed from the mutant thylakoids than from WT samples, leading to an overrepresentation in the supernatant fraction. Accordingly, less FNR than expected remained in the pellet fraction (~ 30% of WT). Analysis of the second tic62 knockout line (tic62-2) confirmed the results presented above, displaying an even stronger effect (only ~ 20% residual FNR in the pellet; data not shown). Thus, FNR is more readily washed from the thylakoid membranes in presence of high ionic-strength when Tic62 is not present, indicating a strong binding between the two proteins in the WT.
In the following, it was investigated whether Tic62 preferentially binds to one leaf FNR isoform over the other. For this purpose, heterologously expressed and purified Tic62 Ct, containing the FNR-binding repeats, was bound via its (His)6-tag to Ni2+-beads. These
were used as an affinity matrix for LFNR1 and LFNR2 from Arabidopsis tic62 stroma, being devoid of endogenous Tic62 (Figure 30). Elution was carried out by increasing first the ionic- strength of the solution and subsequent addition of 4 M or 8 M urea, respectively, to denature the proteins still bound to the matrix. Finally, the column was stripped with imidazole. Analysis of the resulting fractions by immunoblotting revealed that both FNR isoforms eluted equally well from the Tic62-Ct matrix (Figure 30, upper lane), indicating a similar affinity for Tic62. In addition, only little FNR could be eluted with salt, in line with the findings from the dissociation experiments from thylakoids. For a major fraction of bound FNR, denaturation with urea was necessary to release the protein from Tic62. As controls, stroma was also incubated with equal amounts of immobilized fructose-1,6-bisphosphatase (FBPase) or with Ni2+-sepharose only. No (unspecific) FNR binding was detected in either sample.
Figure 30: Tic62 binds both LFNR1 and LFNR2 equally well. LFNR1/ LFNR2 binding assay on Tic62Ct-His affinity matrix. Overexpressed and purified Tic62 Ct and FBPase were bound via a (His)6-tag to Ni2+-beads and used as an affinity matrix for LFNR1 and LFNR2 from tic62Arabidopsis stroma. An empty column without the addition of His-tagged protein was used as additional negative control. After incubation for 1 h at 4°C the matrix was washed (W, last wash), and bound proteins were eluted by addition of 750 mM NaCl (E1), 1 M NaCl (E2), 4 M urea (E3), 8 M urea (E4), 200 mM imidazole (E5) and 400 mM imidazole (E6). The resulting samples including 1/40 of load (L) and 2/5 of flow-through (FT) were subjected to Urea/SDS-PAGE and immunoblotting with FNR antibody.
FNR
L FT W E1 E2 E3 E4 E5 E6 Tic62Ct-His
FBPase-His control
2 Heterologous expression and initial characterization of Tic20
As described above (Introduction, chapter 2.1), Tic20 was proposed to act as an alternative preprotein import channel in the IE next to, or in combination with, Tic110 (Inaba et al., 2003). However, all studies so far were based on database predictions, immunological detection of the protein after cross-linking reactions, or the phenotypical characterization of plants with drastically reduced amounts of Tic20 (Ma et al., 1996; Kouranov and Schnell, 1997; Kouranov et al., 1998; Chen et al., 2002; Reumann et al., 2005; Teng et al., 2006; van Dooren et al., 2008). No direct biochemical evidence has been brought forward either demonstrating that Tic20 indeed has channel activity in vitro (or in vivo) or verifying any of the in silico predictions regarding e.g. topology or structure of the protein.
Production of pure protein therefore is a prerequisite for the investigation of biochemical and (electro-) physiological functions in more detail but no successful protocol has been published yet. Therefore, several new heterologous expression and purification strategies for Tic20 were tested, two of which are presented in the following: (I) expression of a codon-optimized Tic20 from Pisum sativum in a self-made Escherichia coli (E. coli) cell- free coupled transcription-translation system (S12-lysate) and (II) expression of Tic20 from Arabidopsis thaliana in standard E. coli cultures using a cold-induced promoter system.