Epitope mapping of monoclonal antibodies for the Deinococcus radiodurans bacteriophytochome

Abstract

Bacteriophytochromes (BphP) are phytochrome-like light sensing proteins in bacteria, which use biliverdin as a chromophore. In order to study the biochemical properties of the DrBphP protein, five (2B8, 2C11, 3B2, 3D2, and 3H7) anti-DrBphP monoclonal antibodies were produced through the immunization of mice with purified full-length DrBphP and DrBphN (1–321 amino acid) proteins, and epitope mapping was then carried out. Among the five antibodies, 2B8 and 2C11 preferentially recognized the N-terminal region of BphP whereas 3B2, 3D2, and 3H7 showed preference for the C-terminal region. We performed further epitope mapping using recombinant truncated BphP proteins to narrow down their target sequences. The results demonstrated that each of the five monoclonal antibodies recognized different regions on the DrBphP protein. Additionally, epitopes of 2B8 and 3H7 antibodies were discovered to be shorter than 10 amino acids (2B8: RDPLPFFPP, 3H7: PGEIEEA). These two antibodies with such specific recognition epitopes could be especially valuable for developing new peptide tags for protein detection and purification.

Introduction

Phytochromes are light-sensing plant proteins that inter-convert to either a red light absorbing Pr structure or far-red light absorbing Pfr structure depending on the specific wavelength (660 nm and 730 nm) of light to which they are exposed.^1,2 Plant phytochromes generally consist of P2/PAS, P3/GAF, and P4/PHY domains in the N-terminal photosensory region and PAS1, PAS2, and protein kinase domains in the C-terminal regulatory regions.³ For a long period of time, plant phytochromes were thought to be the only existing red light absorbing photoreceptor proteins. However, since the discovery that the RcaE protein in cyanobacteria Fremyella diplosiphon is homologous to the chromophore binding domain (CBD) of PhyE in Arabidopsis, many phytochrome-like photoreceptors have been reported in proteobacteria, actinobacteria, fungi, and slime molds.^4,5 Even some non-photosynthetic eubacteria, such as Deinococcus radiodurans, Psuedomonas putia, and Pseudomonas aerugnosa, contain light sensing chromophores that express phytochrome-like photoreceptors designated as bacteriophytochromes (BphP).^6,7 The DrBphP covalently associated with the biliverdin α (BV) chromophore has linear tetrapyrroles to generate reversible R/FR isoform structures like plant phytochromes.⁷ The chromophore attached to holo-DrBphP showed an identical photochromic property to those of higher-plant phytochromes, except for a little red shifted absorption maxima (698 nm and 750 nm). This difference is thought to be due to the attachment of BV chromophore, instead phytochromes attach to phytochromobilin, which is a derivative of BV.^7,8 BphP has domain structures that are similar to plant phytochromes, in that P2/PAS (Per: period circadian protein; Arnt: arylhydrocarbon receptor nuclear translocator protein; Sim: single-minded protein), P3/GAF (cGMP-specific phosphodiesterases, adenylyl cyclases, and FhlA), and P4/PHY (phytochrome-associated) domains are located in the N-terminal region followed by HKD (histidine kinase domain) in the C-terminal region.⁵ Since phytochrome domain structures and spectral properties are similar to those of BphPs, it is possible that BphPs are the progenitors of phytochrome-type photoreceptors.⁷ Biochemical similarities between plant phytochromes and BphPs prompted the structural study of BphPs as a model protein for better understanding the structures of plant phytochromes.

We have investigated the biochemical properties of BphP and reported that the C-terminal PHY domain is necessary for spectral integrity and protein stabilization.⁹ Recently, crystal structures of BV-bound PAS and GAF domains of DrBphP in Pr form and Pseudomonas aeruginosa (PaBphP) were revealed by 2.5 Å resolution.^10,11 Yang et al. reported the essential role of the PHY domain in the efficient phototransformation of a chromophore. They also identified residues that are involved in the stabilization of the 15Ea (Pfr) and 15Za (Pr) configuration of the BV chromophore, and that a Ser-261 is important for a stable Pr state in PaBphP.¹²

Since we do not know the real structures of plant phytochromes, structural and biochemical studies of BphPs could be an alternative for better understanding phytochrome structures and their structural (Pr/Pfr) changes. For this reason, we generated DrBphP specific monoclonal antibodies to use for structural and biochemical studies of BphP. Here, we report the specific epitopes of five novel monoclonal antibodies for BphP and their biochemical properties. These new antibodies will expand the possibilities for biochemical research of BphP as well as plant phytochromes.

Results

To analyze the biochemical and surface properties of DrBphP, epitopes of five anti-DrBphP monoclonal antibodies were determined using various engineered recombinant DrBphP proteins. CBDs of DrBphP structures were identified in the Pr state.^10,13 However, structural information about PHY and the histidine kinase domain is still unclear. To analyze the surface topography of the DrBphP C-terminal domain as well as the previously reported N-terminal domain, epitope mapping of the DrBphP protein was carried out using newly generated DrBphP-specific monoclonal antibodies.

Five DrBphP-specific monoclonal antibodies (2B8, 2C11, 3B2, 3D2, and 3H7) were generated by using highly purified full-length DrBphP and DrBphN (N-terminal domain: 1–321 amino acid region) proteins.¹⁴ An initial ELISA test showed that the 2B8 and 2C11 reacted with both DrBphP and DrBphN, whereas 3B2, 3D2, and 3H7 reacted only with a full length DrBphP protein. To confirm more specific regions of the DrBphP protein for five generated monoclonal antibodies, Western blot analysis was carried out using purified DrBphP and DrBphN proteins. The 2B8 and 2C11 bound to both BphP and BphN proteins, whereas 3B2, 3D2, and 3H7 bound only to DrBphP protein, which was consistent with the ELISA test results (Fig. 1). These results suggest that 2B8 and 2C11 recognize the N-terminal region (1–321 aa, PAS and GAF domain) whereas 3B2, 3D2, and 3H7 recognize the C-terminal region (322–755 aa, PHY and HK domain). Since the five generated monoclonal antibodies reacted with denatured BphP or BphN proteins, epitopes for these antibodies were considered to be sequential epitopes, which require only specific linear and continuous amino acids, as opposed to conformational epitopes, which only function in specific tertiary structures or specific peptide orientations.^15–18 ELISA results using denatured and native antigen proteins confirmed that epitopes identified were sequential (Supporting Information Fig. 3). All five antibodies recognized both denatured and native antigens with similar reactivity.

Figure 1.

Western blot analysis of selected monoclonal antibodies using purified BphP and BphN proteins. BphP and BphN apoproteins were purified by Ni⁺-NTA affinity chromatography and subjected to Western blotting. Detection was carried out using purified monoclonal antibodies of DrBphP (2B8, 2C11, 3B2, 3D2, and 3H7). BphP: 1–755 aa, full-length; BphN: 1–321 aa containing PAS and GAF domain.

Figure 3.

List of constructed recombinant partial DrBphP peptides fused with GST for epitope identification and Western blot analysis. Short fragment peptides of DrBphP were inserted into the pGEX4T1 vector that includes GST protein in the cloning site. Listed all of target DrBphP peptides were connected in the C-terminal end of GST protein (A). Western blot analyses were carried out to find out correct epitope sequences for five monoclonal antibodies. Western blot analysis using (B) 3B2, (C) 3D2, (D) 2C11, (E) 3H7, and (F) 2B8 antibodies.

We hypothesized that the five monoclonal antibodies would recognize different regions on the DrBphP protein (Fig. 1). Because no DrBphP-specific monoclonal antibody had ever been reported, our work of antibody generation as well as epitope mapping is valuable to future DrBphP research. For detailed epitope mapping, we constructed various truncated recombinant DrBphP proteins by deleting several amino acids from either the N-terminus or the C-terminus. We expressed a series of deletion recombinant DrBphP proteins from both of N- and C-terminal ends by cloning in pET28 vectors [Fig. 2(A)]. All of the expressed proteins were analyzed by Western blot with each of the five monoclonal antibodies resulting in the 3H7 bound to all of C-terminal deleted DrBphP proteins whereas 3B2 recognized 1–727 and 1–755, and 3D2 recognized only the full-length (1–755) DrBphP protein [Fig. 2(B)]. These results suggest that the recognition site of the 3H7 is in between amino acids 320 and 580, the 3B2 in between amino acids 698 and 727, and the 3D2 in between amino acids 728 and 755 [Fig. 2(B)]. The epitope of 2B8 was located somewhere in between amino acids 1 and 20, while the epitope of 2C11 is located in between amino acids 40 and 321 (or within the general vicinity of those regions) [Fig. 2(C)].

Figure 2.

Western blot analysis of recombinant truncated DrBphP proteins. All of these recombinant truncated DrBphP proteins were fused with a His-tag at the N-terminal end and expressed by a pET28a vector in E. coli. Proteins were then subjected to SDS-PAGE followed by Western blot using purified monoclonal antibodies. A: Progressive amino acid deletions of DrBphP, either from the C-terminal or N-terminal end, were generated using the pET28a(+) or pET21a vector including 1–755 aa (Full-length). B: C-terminal truncated recombinant DrBphP proteins were analyzed by Western blot using 3B2, 3D2, 3H7 antibodies. C: N-terminal truncated recombinant DrBphP proteins were analyzed by Western blot using 2B8, 2C11 antibodies.

To find out an exact epitope sequence of DrBphP for five monoclonal antibodies, we generated recombinant glutathione S-transferase (GST) proteins that contained short, partial peptide sequences selected from the previous mapping result of DrBphP [Fig. 3(A)]. These recombinant DrBphP peptides, containing GST proteins, were constructed by using pGEX vectors because they contain the GST gene. Although DrBphP full-length fusion proteins with GST exhibited reduced solubility in a previous report, short peptides of DrBphP fused with GST proteins were well expressed and soluble.¹⁹ 3B2 antibody recognized peptide sequences 691–727, 691–737, and 691–747 [Fig. 3(B)]. 3D2 recognized both peptides 727–755 and 730–755 [Fig. 3(C)]. These results are consistent with previous Western blot analysis that the 3B2 antibody requires amino acids fragment of 691–727 for detection and 3D2 requires 730–755 of DrBphP. Since Figure 2C indicated that the N-terminal domain-specific 2C11 antibody recognized 40–755 without recognition of 80–755, we initially produced a GST fusion with amino acids 20–180 and 40–180 of DrBphP. 2C11 recognition sequence was finally narrowed down that amino acids fragment 38–128 corresponding to the PAS domain of DrBphP is the recognition sequence for the 2C11 antibody [Figs. 2(B) and 3(D)].

Analysis of the 3H7 antibody with GST fusions with amino acids 320–420, 400–500, and 480–580 of DrBphP resulted in the detection of amino acids 400–500 and 480–580, but not 320–420 [Fig. 3(A,E)]. This result suggests the binding site of 3H7 is around 480–500. An additional experiments with various mutant antigens surprisingly showed that a six-amino acid fragment 485–490 could be an epitope of 3H7 [Fig. 3(E)]. The six amino acid epitope (⁴⁸⁵PGEIEE⁴⁹⁰) was critical for detection by Western blot of the 3H7 antibody. However, the 3H7 antibody maximally reacted with a seven-amino acid peptide epitope (⁴⁸⁵PGEIEEA⁴⁹¹) suggesting that those seven amino acids are a valuable epitope for 3H7 antibody [Fig. 3(E)]. 2B8 antibody could not detect any BphP fragment in which 20 or more amino acids had been deleted from the N-terminal end [Figs. 2(C) and 3(A)]. Additional analysis indicated that the 2B8 antibody recognized only peptide 1–12 of DrBphP [Fig. 3(F)]. Even shorter fragments including amino acids 3–12, 3–11, 3–10, and 4–12 were analyzed and resulted in an amino acid fragment 3–11 as an essential minimal epitope required for the recognition of 2B8 antibody [Fig. 3(F)]. Thus, the nine amino acid stretch ³RDPLPFFPP¹¹ of DrBphP protein is the minimal-length epitope for 2B8 antibody.

Discussion

Our epitope mapping elucidated that the five DrBphP specific monoclonal antibodies recognize various surface regions of DrBphP protein (Fig. 4). 2B8 recognized amino acids 3–11 of the N-terminal extension fragment of DrBphP, the region very close to the chromophore (BV) binding site, Cys-24.¹⁰ Even though the crystal structure shows that Cys-24 is located in the chromophore binding pocket, which is comprised of a P2/PAS domain and a GAF domain containing a conserved trefoil knot structure, our results showed that amino acids 3–11 (N-terminal fragment) are highly exposed to the surface. The 2C11 recognized a broad region of the PAS domain (amino acids 38–128), which is a critical constituent of the “knot” structure that generates the chromophore-binding pocket with a GAF domain,¹³ suggesting that the PAS domain is more exposed to the surface in the N-terminal region. 3H7 recognized a very short, 6-amino acid peptide (residues 485–490) located in the end region of the PHY domain, the domain necessary for Pfr stabilization and spectral integrity of DrBphP.⁹ Photoconversion of Pr to Pfr, triggered by a light-induced isomerization of the chromophore, changes N-terminal structure as well as the region of the C-terminal that contains the PHY and HK domains. Residue 532, located on the hinge in between PHY domain and HK domain, has been reported to be more exposed in the Pfr state.¹⁹ Residue 466, on the other hand, appears to be more exposed in a Pr state because this residue can be more easily cleaved by trypsin digestion in the Pr state.⁹ These results indicate that the 3H7 epitope is located between residues 466 and 532, the region in between PHY and HK domains, which is highly exposed in both Pr and Pfr states. To analyze whether this epitope region has photoconversion-specific surface properties, we examined the immunoreactivity of both Pr and Pfr states of DrBphP proteins via immunoprecipitation experiment. Results showed no significant difference between the Pr and Pfr states suggesting that the epitopes are not located on the critical surface on the conformational changes by lights (Supporting Information Fig. 1). Additionally, antibodies did not interfere with the spectral (Pr/Pfr) changes of DrBphP strongly indicating that epitopes are exposed in both Pr and Pfr conformations (Supporting Information Fig. 2).

Figure 4.

Schematic representation of epitope location of the five monoclonal antibodies. PSM: photosensory module, OPM: output module, CBD: chromophore binding domain.

3B2 and 3D2 antibodies recognized residues 691–727 and 730–755 located on the histidine kinase domain that is classified as an output module in BphPs.¹⁹ These results suggest that the HK domain is more exposed than other domains of the DrBphP protein. The HK domain has historically been less investigated than N-terminal domain probably due to difficulties with the generation of crystals. However, the HK domain is important for proper dimerization as well as for spectral/structural perfection.^8,19 For proper function, the HK domain interacts with and phosphorylates a response regulator (RR) in order to transfer light signals through two-component systems (TCSs).^7,20,21

In this study, we produced five DrBphP-specific monoclonal antibodies and characterized the specific binding epitopes. 2B8 and 2C11 recognized residues 3–11 of the N-terminal extension and residues 38–128 of the PAS domain. 3B2, 3D2, and 3H7 recognized the C-terminal region of the hinge and HK domains. Whether or not identified epitopes require specific structural conformations for their immune response against five monoclonal antibodies, denatured and native antigen proteins were reacted with each antibody. Antibodies reacted similarly to the denatured and native antigen proteins suggesting that the epitopes are probably sequential rather than conformational (Supporting Information Fig. 3). Since 3-D current structural information of the C-terminal region including the PHY and HK domains is insufficient, our epitope mapping results will provide valuable information especially about the C-terminal region. In addition, the DrBphP-specific monoclonal antibodies we developed are valuable tools for future biochemical research of BphP proteins. Additionally, extremely short peptide epitopes for 2B8 and 3H7 antibodies (2B8: RDPLPFFPP, 3H7: PGEIEEA) could be developed for use as a new tag for protein detection and purification.

Materials and Methods

Production of five monoclonal antibodies of DrBphP

Monoclonal antibodies for the DrBphP protein were produced by the commercial company, ATGEN (KOREA). We supplied ATGEN with two highly purified recombinant proteins DrBphP (full-length) and DrBphN (1–321 aa), and they utilized mice to produce 5 specific antibodies, out of 24 candidate antibodies, in response to antigen injection.

Six to eight weeks old male BALB/c mice were used for antibody production. Mice were first immunized with highly purified full length DrBphP protein mixed with an equal amount of complete Freund's adjuvant by intra-peritoneal injection and followed by boosters after 4 weeks with incomplete Freund's adjuvant. Antibody levels in serum isolated from the immunized mice were monitored by ELISA. The spleen cells of immunized mice were fused with mouse myeloma cell using PEG 1500. The fused cells were selected in HAT medium and screened by ELISA and Western blot using hybridoma supernatants. Monoclonal hybridoma was selected by the method of limiting dilution. Eight weeks old female BALB/c mice were inoculated with 1 × 10⁷ hybridoma cell producing monoclonal antibodies through an intra-peritoneal injection. After 7 days, ascites fluids were harvested. Antibodies were purified by Protein A/G agarose followed by buffer change with 0.02% sodium azide in phosphate buffer saline (pH 7.4) and concentration.²²

Preparation of truncated bacteriophytochrome constructs

Full-length D. radiodurans bacteriophytochrome DNA was used as a starting material for the construction of all deletion mutants of DrBphP. Five N-terminal [Δ(N-19), Δ(N-39), Δ(N-79), Δ(N-159), Δ(N-239)] and four C-terminal deletion mutants [Δ(581-C), Δ(661-C), Δ(698-C), Δ(728-C)] were generated by using polymerase chain reaction (PCR). Deletion mutant DNA was cloned into pET-28(a)+ vectors. Other short fragments [1–8 aa, 1–12 aa, 1–17 aa, 1–27 aa, 3–10 aa, 3–11 aa, 3–12 aa, 4–12 aa, 9–17 aa, 9–27 aa, 20–180 aa, 38–128 aa, 40–180 aa, 320–420 aa, 400–500 aa, 480–580 aa, 420–480 aa, 480–500 aa, 481–495 aa, 485–500 aa, 485–495 aa, 485–494 aa, 485–493 aa, 485–492 aa, 485–491 aa, 485–490 aa, 485–489 aa, 486–495 aa, 691–727 aa, 691–737 aa, 691–747 aa, 727–755 aa, and 730–755 aa] DNA were also generated by using PCR and cloned into the pGEX4T1 vector which contains the GST gene. Table 1 shows all primers used for DNA constructs (Supplementary Table. 1).

Expression and purification of recombinant DrBphP, DrBphN, and recombinant truncated BphP proteins

Truncated BphP DNA constructs were cloned into the pET-21(a), pET-28(a)+, and pGEX4T1 vectors and expressed in Escherichia coli, BL-21 cells. The cells were grown at 25°C and protein expression was induced under 1 mM Isopropyl-beta-D-thiogalactopyranoside (IPTG). After 6-h protein induction, the cells were collected by centrifugation. In the case of constructions in pET vector, cells were resuspended in binding buffer (100 mM Tris-Cl, 200 mM NaCl, and 10 mM immidazole, pH 8.0). Resuspended cells were lysed by sonication on ice for a total of 10 mL lysate. The cell lysate was centrifuged at 10,000 rpm for 10 min at 4°C. Clarified total protein solution was then subjected to an Ni⁺-NTA affinity column purification process (Qiagen, Valencia, CA) and target proteins were eluted with a buffer containing 100 mM Tris-HCl (pH 8.0), 200 mM NaCl, and 200 mM imidazole. The purified recombinant target BphP proteins were stored at −70°C for later use.

Sodium dodecyl sulfate - polyacrylamide gel electrophoresis (SDS-PAGE)

All samples were mixed with SDS sample buffer (125 mM Tris-HCl, pH 6.8; 4% (wt/vol) SDS; 0.005% (wt/vol) bromophenol blue; 20% (vol/vol) glycerol; 5% (vol/vol) β-mercaptoethanol) and SDS-PAGE was performed with 10% Tris-glycine gels in Tris running buffer (25 mM Tris pH 8.3, 0.1% (vol/vol) SDS, 250 mM glycine) using a Hofer™ Dual Gel Caster (GE Healthcare, Piscataway, MA) at 20 mA for 90 min. Proteins were stained with Coomassie brilliant blue R-250 (Sigma-Aldrich, St.Louis, MO).

Western blot

Proteins were separated on a 10% SDS-PAGE gel and transferred to polyvinylidene difluoride membrane (Invitrogen, Carlsbad, CA). Anti-His (H-3, Santa Cruz Biotechnology, Dellas, TX), anti-GST (GST Ab-1, Neo Markers, Fremont, CA), and five DrBphP-specific monoclonal antibodies (2B8, 2C11, 3H7, 3B2, and 3D2) were then used for detection of His6, GST, and expressed recombinant DrBphP proteins, including the short DrBphP peptides. Horseradish peroxidase-conjugated anti-mouse IgG was used as a secondary antibody, and Western signals were detected by chemiluminescence using an ECL reaction (GE Healthcare).

Supporting Information

Additional Supporting Information may be found in the online version of this article.

Supplementary Information