Figure 1.
Schematic illustration of the method of protein C-terminal labeling.
The precursor protein is a recombinant fusion protein consisting of the target protein (TP) to be labeled, the 150-aa N-intein (IN) of the Ssp GyrB S11 split-intein for trans-splicing, and an affinity binder (AB) for easy purification of the precursor protein. The synthetic peptide consists of the 6-aa C-intein (IC) of the Ssp GyrB S11 split-intein for trans-splicing, followed by a serine residue (highlighted in red), a small linker sequence (blue box), and the labeling group (L) linked to the side chain of a suitable amino acid X. The N-intein and the C-intein are presented in presumed ribbon structures derived from the conserved intein crystal structure of the Ssp DnaB mini-intein intein [41], illustrating their association through structural complementation to reconstitute an active intein that catalyzes a trans-splicing reaction to produce the labeled protein and the excised inteins.
Figure 2.
Protein C-terminal labeling with a fluorophore.
A. Schematic illustration of the labeling reaction. IN and IC: components of the Ssp GyrB S11 split-intein. M: maltose binding protein as the target protein to be labeled. C: chitin-binding domain as the affinity binder for purification of the precursor protein. L: 5-carboxyfluorescein as the labeling group. In the synthetic peptide IC-L, IC is connected to L by the sequence SAGSGK, with L attached to the side chain of the K (lysine) residue. B. Analysis of the labeling results. The purified precursor protein was incubated at room temperature for 16 hours, with or without the peptide and the reducing agent TCEP, as indicated. The reaction products were resolved through SDS-PAGE and visualized either by Coomassie staining, by Western blotting using anti-C antibodies, or by fluorescence scan (excitation at 488 nm, filter for 520 nm). Positions of the precursor protein (MINC), the labeled protein (ML), and the excised N-intein (INC) are indicated. In lane 1, a minor protein band at the same position as ML is the endogenous E. coli maltose binding protein that is known to be co-purified in the amylose affinity chromatography. C. Time-course of the reaction between MINC precursor protein and IC-L peptide in the presence or absence of TCEP. Labeling efficiency was calculated from densitometry analysis on anti-C Western blots. Error bars represent standard deviations from triplicate experiments.
Figure 3.
Mass spectrometry identification of the fluorescently labeled maltose binding protein ML.
A. ESI-MS analysis of the labeled protein. The determined molecular weight of 43814 Da was in close agreement with the calculated mass (43830.7 Da) of the labeled protein ML. B. MS/MS analysis of the labeled protein after trypsin digestion. The deconvoluted spectrum of a 1378-Da peptide corresponded well to the calculated mass (1378.964 Da) of the C-terminal peptide of ML. The y-fragment ions of this peptide yielded to the amino acid sequence given below the spectrum, which unambiguously identified the trans-spliced peptide at the C-terminus of the labeled ML protein.
Figure 4.
Detection of possible N-cleavage as a side reaction.
A. Illustration of the detection strategy. The predicted N-cleavage product MC is compared with the trans-splicing (C-terminally labeled) product ML. Site-specific cleavage of MC and ML with protease Factor Xa releases the C-terminal peptide pepXa and pepXa-L, respectively. B. RP-HPLC analysis. The left chromatogram shows a synthetic pepXa used as a standard. The middle and right chromatograms show the sample peptide detected at 210 nm (middle) and 497 nm (right). The increasing baseline absorbance at 497 nm over time was most likely due to the absorbance of the increasing concentration of the mobile phase B solution. C. MS-coupled HPLC analysis. The pepXa-L was much more abundant than pepXa, and both peptides were identified by their corresponding molecular weights in mass spectrometry analysis (data not shown).
Figure 5.
On-column production of ML protein.
A. The co-purification labeling process was monitored through SDS-PAGE analysis with Coomassie blue staining or fluorescence scanning as indicated. Lanes 1 and 2: total E. coli proteins before and after IPTG-induced expression of the precursor protein MINC, respectively. Lane 3: soluble fraction of the cell lysate of lane 2. Lane 4: proteins of lane 3 bound to chitin beads. Lane 5: proteins released from the chitin beads of lane 4 after an overnight incubation with the labeling peptide IC-L in the presence of TCEP. B. Amylose resin was incubated with (panels 1 and 2) or without (panels 3 and 4) the purified ML protein. Panels 1 and 3 are differential interference contrast images, while panels 2 and 4 are fluorescence images. C. SDS-PAGE analysis of the ML protein fractions eluted from the amylose resin of panels 3 and 4 of B using a maltose-containing elution buffer.
Figure 6.
C-terminal biotinylation using the Ssp GyrB S11 split-intein.
A. The three precursor proteins (MINC, EINH, and GST-ACP-INC) were incubated with (+) or without (−) the peptide IC-B in the presence of 0.1 mM TCEP for 18 h at room temperature, and the reaction products were analyzed by Western blotting using antibodies against biotin. From the three precursor proteins, the three target proteins for biotinylation were a maltose binding protein (M, 43 kDa), an enhanced green fluorecent protein (E, 27 kDa), and a glutathione-S-transferase/acyl carrier fusion protein (GST-ACP, 39 kDa), respectively. B. Efficiency of C-terminal biotinylation of the three target proteins as determined by densitometry analysis on Western blots using anti-C antibodies (for MINC and GST-ACP-INC) and anti-H antibodies (for EINH). Error bars represent standard deviations from triplicate experiments.
Figure 7.
Immobilization of biotinylated enhanced green fluorescent protein (EB) to a solid surface.
A, The biotinylation-purification method. The precursor protein (EINC) in a cell lysate is first bound to chitin beads through its chitin binding domain (C). After washing away the unbound proteins, the beads are incubated with the peptide IC-B that consists of IC followed by the sequence SAGSK with biotin linked to the lysine (K) side chain. A trans-splicing reaction produces the biotinylated target protein (EB) that can bind to streptavidin-coated beads (S). In the experimental proof, streptavidin-coated beads were incubated either with the EINC precursor in a cell lysate (panels 1 and 2) or with the biotinylated and purified EB protein (panels 3 and 4). Panels 1 and 3 are differential interference contrast images, while panels 2 and 4 are fluorescence micrographs (45 ms exposure time). B, The biotinylation-fixation method. The IC-B peptide is first bound to streptavidin-coated beads and then incubated with the precursor protein EINC in a cell lysate to achieve trans-splicing, and the resulting biotinylated target protein (EB) is automatically fixed to the beads. In the experimental demonstration, streptavidin-coated beads with (panels 3 and 4) or without (panels 1 and 2) the pre-bound IC-B peptide were incubated with the EINC protein in an E. coli cell lysate for trans-splicing. After washing away unbound proteins, the beads were photographed as differential interference contrast images (panels 1 and 3) or as fluorescence images (panels 2 and 4, 600 ms exposure time).
Figure 8.
Labeling of transferrin receptor (TfR) protein on live cells.
A. Expression of the TfR-intein fusion protein. CHO-TRVb cells, which are Chinese hamster ovary cells deficient for endogenous TfR, were transfected with plasmid pcTR-HA expressing a human TfR fused to a C-terminal HA tag, plasmid pcTR-IN-HA expressing the TfR fused to IN and the HA tag (TfR-IN-HA fusion protein), or plasmid pcDNA3.1 as an empty vector, as indicated. On the left are confocal microscopy images of immunofluorescence using antibodies against the HA tag (α-HA), and the cells were counterstained with rhodamine-phalloidin for actin. On the right are Western blots using α-HA or α-TfR antibodies, with the monomeric and dimeric forms of the receptor proteins indicated. B, Function of the TfR-intein fusion protein. On the left, CHO-TRVb cells transfected with the indicated plasmids were incubated with rhodamine-labeled transferrin (Rh-Tf) to show function of the TfR-intein fusion protein in cellular uptake of transferrin. The cells were counterstained using antibodies against tubulin. On the right, cells transfected with the pcTR-IN-HA plasmid were incubated with Rh-Tf and also visualized by immunofluorescence using α-HA antibodies, which showed co-localization of the TfR-IN-HA fusion protein with Rh-Tf. The cells were counterstained with TO-PRO-3 for the nucleus. C. Observation of labeling of TfR. CHO-TRVb cells transfected with the indicated plasmids were incubated with increasing amounts (as indicated) of the fluorescent IC-L peptide (described in Figure 2). Only cells expressing the TfR-intein fusion protein from the pcTR-IN-HA plasmid showed discrete green fluorescent spots, indicating successful labeling of TfR. The cells were counterstained with rhodamine-phalloidin for actin in the confocal microscopy.