Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/springer-journals/development-of-fluorescence-labeled-antibody-for-immune-checkpoint-IWQY6QkyMO
- Publisher
- Springer Journals
- Copyright
- Copyright © The Author(s) 2023
- eISSN
- 2191-0855
- DOI
- 10.1186/s13568-023-01509-y
- Publisher site
- See Article on Publisher Site
Abstract
Introduction with a green fluorescent protein (GFP) that is capable of Lactic acid bacteria (LAB) are among the most exten- recognizing programmed death ligand 1 (PD-L1), and sively studied human symbiotic gut bacteria and fer- that can be used as a tool for the treatment, alleviation, menters of livestock products and have been classified by and diagnosis of cancer. PD-L1, known as an immune the US Food and Drug Administration (FDA) as generally checkpoint molecule, is an important target in cancer being safe for use in the food sciences (Hill et al. 2017). therapy (Dermani et al. 2019). Therefore, anti-PD-L1 Although considerable research has been conducted on antibodies, which are immune checkpoint inhibitors improving quality of life by utilizing the functionality of (ICIs) that can inhibit the interaction between PD-L1 LAB, their application as biopharmaceuticals has been and its receptor, programmed death 1 (PD-1), can be anticipated in recent years. In particular, the use of genet- applied clinically as powerful antitumor agents (Jiang ically modified LAB (gmLAB) has gained attention in et al. 2019b). Antibody drugs are expensive to manufac- the prevention and alleviation of mucosal disorders (Shi- ture due to their sophisticated structure, but scFv, which gemori and Shimosato 2017). gmLAB can be used to pro- contains an antigen recognition site for an antibody con- duce a variety of recombinant proteins by transforming nected by a flexible peptide linker, can be produced in a cells with gene expression vectors. gmLAB are ingested prokaryotic expression system and thus can be procured and remain viable in the intestinal tract, where they can at a low cost (Arbabi-Ghahroudi et al. 2005; Samaranay- produce recombinant proteins in situ; as a result, they ake et al. 2009). In addition, the presence or absence of have been studied for use as an intestinal drug delivery PD-L1-positive cells in the tumor microenvironment system for recombinant proteins (Cano-Garrido et al. can have a significant impact on the selection of an 2015). In addition, a recent study showed that intranasal appropriate treatment strategy (Jiang et al. 2019a), mak- administration of gmLAB can also deliver proteins to the ing the simple detection of PD-L1-positive cells a criti- respiratory mucosa, thereby contributing to disease relief cal issue. Therefore, we considered that constructing a (Yumoto et al. 2020). In this context, gmLAB capable of gmLAB strain capable of producing GFP-fused anti-PD- producing beneficial or therapeutic proteins are classified L1 scFv (GFP-PDL1scFv) could be an attractive alterna- as next-generation probiotics (NGPs) or microbial thera- tive antibody drug and diagnostic/treatment tool. This peutics and are expected to be exploited further in the study aimed to design an scFv based on the anti-PD-L1 future (Jimenez et al. 2019; O’Toole et al. 2017). antibody, construct a gmLAB strain that produces GFP- In this study, we constructed a strain of gmLAB that PDL1scFv, and verify the PD-L1 protein recognition abil- produces a single chain variable fragment (scFv) fused ity of GFP-PDL1scFv. Namai et al. AMB Express (2023) 13:4 Page 3 of 11 Materials and methods was cultured in M17 broth (BD Difco , Becton, Dick- Bacterial strains, growth conditions, and plasmid inson and Co., MD, USA) containing 0.5% glucose Lactococcus (L.) lactis subsp. cremoris NZ9000 (GM17) at 30 °C without shaking. Constructed gmLAB (NZ9000) was purchased from MoBiTec GmbH (Got- were cultured using GM17 supplemented with 10 µg/ tingen, Germany). NZ9000 is a derivative of L. lactis mL chloramphenicol (GM17cm). Escherichia (E.) coli subsp. cremoris MG1363 in which the pepN gene has MC1061 was purchased from MoBiTec GmbH and been replaced with the constituent genes of the NICE cultured using Luria-Bertani (LB) broth (Invitrogen system, i.e., nisR and nisK (Ruyter et al. 1996). NZ9000 Fig. 1 Scheme of plasmid construction. a The lactococcal gene expression vector, pNZ8148#2:CY T, derived from commercially available pNZ8148. b The green fluorescent protein (GFP) gene was integrated into pNZ8148#2:CY To construct the lactococcal GFP expression vector, pNZ8148#2:CY T-GFP. c Predicted 3D model of PDL1scFv designed using SWISS-MODEL. The locations of the VH chain, VL chain, and peptide linker (EGKSSGSGSESKS) are shown. VH chain; variable region of the heavy chain, VL chain; variable region of the light chain. d The PDL1scFv gene was integrated into pNZ8148#2:CY T-GFP by restriction digestion using BamHI and HindIII to construct GFP-conjugated PDL1scFv expression vector, pNZ8148#2:CYT-GFP-PDL1scFv. P : nisin-inducible promoter, His-tag six-histidine tag, MCS multiple cloning site, T terminator, repC and repA origins nisA of replication, Cat chloramphenicol acetyltransferase, GFP green fluorescent protein Namai et al. AMB Express (2023) 13:4 Page 4 of 11 Corp., CA, USA) containing 25 µg/mL chlorampheni- (TBS: 50 mM Tris-HCl, 140 mM sodium chloride, pH col (LBcm) at 37 °C with intense shaking. 8.0) or phosphate-buffered saline (PBS; 137 mM sodium The NICE-system plasmids, pNZ8148#2:CYT and chloride, 2.7 mM potassium chloride, 10 mM disodium pNZ8148#2:CYT-GFP, were constructed based on the hydrogen phosphate, 1.76 mM potassium dihydrogen commercially available pNZ8148 (MoBiTec GmbH), as phosphate, pH 7.4). Then, the cell pellets were crushed described in Shigemori et al., and were used as the gene using a bead beater (µ-12; TITEC, Saitama, Japan), and expression vectors (Fig. 1a, b) (Shigemori et al. 2012, the soluble fraction was obtained by centrifugation (4 °C, 2017b). 20,000×g, 15 min). An equal volume of 2× sample buffer (Wako, Osaka, Japan) was added to the soluble fraction Design of the scFv and construction of the gene expression and boiled at 95 °C for 5 min to prepare the sample for vector western blotting (WB) (Ishida et al. 2020; Namai et al. PDL1scFv was designed by connecting the amino 2018a). acid sequences of the variable region of atezolizumab using the flexible peptide linker EGKSSGSGSESKS. Confocal laser scanning microscopy The three-dimensional (3D) structure of the designed Freshly prepared TBS-washed bacterial pellets were PDL1scFv was predicted using SWISS-MODEL, suspended in 400 µL of TBS, and 10 µL was placed on an automated protein homology-modeling server a microscope slide and observed under a confocal laser (Schwede et al. 2003). The designed amino acid scanning microscope (FluoView FV1000, Olympus, sequences were then converted to nucleotide sequences Tokyo, Japan) using an oil immersion objective lens by Eurofins Genomics (Tokyo, Japan), based on L. lac - (×60). tis subsp. cremoris MG1363 codon usage. In addition, restriction enzyme recognition sites, BamHI and Hin- Immunoreactivity assay of GFP‑PDL1scFv dIII, were inserted on each side of the scFv sequence. Recombinant gene expression was induced, and the cell The resulting gene was subcloned into pEX-K4J2 by pellet was crushed to obtain a soluble fraction containing Eurofins Genomics (Tokyo, Japan). General molecular recombinant protein, as described above. The total pro - cloning techniques were performed using modifica - tein concentration of the soluble fraction was measured tions of previously described methods (Namai et al. using a BCA Protein Assay Kit (Thermo Fisher Scien - 2018a). Briefly, the gene segment was excised using tific, MA, USA) according to the manufacturer’s instruc - BamHI and HindIII and cloned into the multi-cloning tions, and the solution was adjusted to 5 mg/mL. The site of pNZ8148#2:CYT-GFP. The resulting GFP-con - immunoreactivity of the recombinant GFP-PDL1scFv jugated PDL1scFv (GFP-PDL1scFv) expression vec- (rGFP-PDL1scFv) was examined using an enzyme-linked tor (designated pNZ8148#2:CYT-GFP-PDL1scFv) was immunosorbent assay (ELISA), as described previously sequenced by Eurofins Genomics (Tokyo, Japan) to (Namai et al. 2020a, b; Shigemori et al. 2017a). confirm the absence of mutations and/or deletions. Purification of rGFP‑PDL1scFv Recombinant gene expression was induced by add- Construction of gmLAB for GFP‑PDL1scFv gene expression ing nisin (final: 1.25 ng/mL) to 1 L culture, as described pNZ8148#2:CYT-GFP-PDL1scFv (DDBJ accession num- above. The cell pellet was then collected by centrifuga - ber: LC739557) was introduced into NZ9000 to construct tion (4 °C, 8,000×g, 5 min), washed using MilliQ water, a gmLAB strain (designated as NZ-GFP-PDL1scFv) by and frozen at −80 °C. The frozen pellets were pulver - electroporation (Namai et al. 2018a). Simultaneously, ized into a fine powder in liquid nitrogen using a Cryo pNZ8148#2:CYT and pNZ8148#2:CYT-GFP were also Press disruptor (Microtec Co., Chiba, Japan), followed introduced into NZ9000 to generate the vector con- by adding 10 mL binding buffer (20 mM imidazole, 20 trol gmLAB (designated as NZ-VC and NZ-GFP). The mM Na PO , 0.5 M NaCl, pH 7.4). The soluble fraction 3 4 constructed gmLAB were then cultured to induce gene was collected (4 °C, 12,000×g, 15 min), and DNA was expression (Namai et al. 2018b). Briefly, the pre-incu - removed using a Nucleic Acid Removal Kit (ProFoldin, bated gmLAB were inoculated into GM17cm (final con - Hudson, MA, USA) as per the manufacturer’s instruc- centration 5%), and nisin, a gene expression inducer, tions. The resulting fraction was filtered using a DISMIC- was added when the optical density at 600 nm (OD ) 25AS filter (pore size 0.45 μm, Toyo Roshi, Tokyo, Japan). reached 0.4 (1-1.5 h). Cells were harvested at 3 h after The filtrate was loaded onto a HisTrap HP column (1 mL, the addition of nisin by centrifugation (4 °C, 8,000×g, GE Healthcare) equilibrated with binding buffer, and 5 min) and washed with ice-cold Tris-buffered saline the column was washed with five column volumes (CV) Namai et al. AMB Express (2023) 13:4 Page 5 of 11 Statistical analysis of binding buffer. The column-absorbed proteins were GraphPad Prism software (version 8, GraphPad, San then eluted with a linear gradient of 0–500 mM imida- Diego, CA, USA) was employed for statistical analysis, zole over 40 CVs at 1 mL/min using a fast protein liquid and significance was accepted at p < 0.05. For flow cytom - chromatography system (AKTA pure 25, GE Healthcare). etry analysis, the data were analyzed using unpaired The collected fractions (soluble fraction; cell, wash, flow- t-tests. through, and eluate; F1-8) were analyzed by WB with CBB staining following SDS-PAGE, as described above. Results The eluted fractions were then dialyzed against PBS, and Construction of GFP‑conjugated PD‑L1scFv expression the His-tagged protein concentration in the dialyzed vector sample was measured using a His-Tag ELISA Detection PDL1scFv amino acid sequences were designed based Kit (GenScript, Piscataway, NJ, USA). on the sequences of the anti-PD-L1 antibody, atezoli- zumab (KEGG Drug: D10773). A flexible peptide linker, Culture conditions for Raw264.7 cells EGKSSGSGSESKS, connected variable regions of the The mouse macrophage cell line, Raw264.7 cells (ATCC, heavy and light chains. To confirm the 3D structure, Manassas, VA, USA), was maintained in complete Dul- the structure of PDL1scFv was predicted using SWISS- becco’s Modified Eagle’s medium (DMEM; with 10% fetal MODEL. The results suggested that the VH chain and bovine serum [GE Healthcare], penicillin [100 U/mL; VL chain, which have independent structures, were Nacalai Tesque, Kyoto, Japan], and streptomycin [100 µg/ connected by the peptide linker (Fig. 1c). The result - mL; Nacalai Tesque]) at 37 °C in 5% CO and passaged ing amino acid sequences were converted to DNA once every three days. sequences based on L. lactis subsp. cremoris MG1363 codon usage by Eurofins Genomics (Tokyo, Japan). The Flow cytometry of Raw264.7 cells restriction enzyme recognition sites, BamHI and Hin- Raw264.7 cells were seeded on a 24-well plate at 2.0 × 10 dIII, were added to both sides of the DNA sequences cells/well and cultured at 37 °C for 2 h. After remov- and then subcloned into pEX-K4J1. PDL1scFv sequences ing the supernatant, a complete DMEM containing 0 or and pNZ8148#2:CYT-GFP (Fig. 1b) derived from 10 µg/mL of lipopolysaccharide (LPS; InvivoGen, San pN8148#2:CYT (Fig. 1a) were excised by restriction Diego, CA, USA) was added. After 4 h of incubation at digestion to construct the GFP-conjugated PDL1scFv 37 °C, cells were collected by centrifugation (4 °C, 500×g, expression vector, pNZ8148#2:CYT-GFP-PDL1scFv 5 min) and washed with PBS containing 1% fetal bovine (Fig. 1d). serum. Cells were then stained using 1/100 dilution of PE anti-mouse CD274 (B7-H1, PD-L1) Antibody (BioLe- Construction of GFP‑PDL1scFv‑producing gmLAB gend, San Diego, CA, USA) or purified rGFP-PDL1scFv pNZ8148#2:CYT-GFP-PDL1scFv was introduced into (500 ng/mL) at RT for 1 h. After washing, cells were ana- NZ9000 by electroporation to construct the gmLAB lyzed using a Cell Sorter SH800 (SONY, Tokyo, Japan), strain, NZ-GFP-PDL1scFv. pNZ8148#2:CYT and and cell populations were identified using FlowJo soft - pNZ8148#2:CYT-GFP were also introduced into NZ9000 ware (v10.5.3; BD Biosciences, San Jose, NJ, USA). to construct the vector control gmLAB strains, NZ-VC, and NZ-GFP, respectively. Each gmLAB strain was cul- Fluorescence observation of Raw264.7 cells tured with/without nisin at 30 °C for 3 h, and the cell Raw264.7 cells were seeded on a 24-well plate at 2.0 × 10 extracts were subjected to WB. As a result, when NZ- cells/well and cultured at 37 °C for 2 h. After removing GFP-PDL1scFv and NZ-GFP were incubated with the supernatant, a complete DMEM containing 0 or 1 µg/ nisin, bands corresponding to either rGFP-PDL1scFv mL of LPS was added. After 24 h incubation at 37 °C, the (59.0 kDa) or recombinant GFP (31.0 kDa) were detected, supernatant was removed, and cells were fixed using 200 respectively (Fig. 2a). In contrast, no bands were detected µL of 10% formalin neutral buffer solution for 10 min. in the NZ-VC, NZ-GFP-PDL1scFv, and NZ-GFP sam- Then, cells were washed twice with PBS containing 0.05% ples without nisin stimulation (Fig. 2a). Next, each Tween 20 (Nacalai) (PBS-T) and stained using 200 µL of gmLAB strain was cultured with or without nisin. The 1/100 dilution of PE anti-mouse CD274 (B7-H1, PD-L1) washed cells were observed under a confocal laser scan- antibody or purified rGFP-PDL1scFv (500 ng/mL) for ning microscope. GFP fluorescence was observed in the 1 h. Cells were washed twice using PBS-T and mounted NZ-GFP and NZ-GFP-PDL1scFv samples incubated using DAPI-Fluoromount-G (Southern Biotech, Birming- with nisin. However, no fluorescence was detected in the ham, AL, USA). The resulting slides were observed under NZ-VC, non-stimulated NZ-GFP, and non-stimulated a BZ-X800 microscope (Keyence, Osaka, Japan). NZ-GFP-PDL1scFv samples (Fig. 2b). Namai et al. AMB Express (2023) 13:4 Page 6 of 11 Fig. 2 Gene expression and fluorescence analysis. NZ-VC, NZ-GFP, and NZ-GFP-PDL1scFv were cultured with/without nisin, and gene expression and fluorescence were analyzed. a The cell extracts were subjected to WB using an anti-His tag antibody to detect the expression of rGFP-PDL1scFv by the gmLAB. The black arrow indicates the band corresponding to rGFP-PDL1scFv (59.0 kDa). The gray arrow indicates the band corresponding to rGFP (31.0 kDa). M: molecular mass marker (kDa). b Fluorescence analyses of gmLAB. Each gmLAB was observed by confocal laser scanning microscopy using an oil immersion objective lens (×60) under visible and fluorescent light, and merged images were generated. −/+: absence/ presence of nisin stimulation Immunoreactivity assay of NZ‑GFP‑PDL1scFv or when NZ-VC cell extracts were added to the PD-L1 The antigen recognition ability of rGFP-PDL1scFv was immobilized well (Fig. 3). These results suggested that investigated by ELISA. The target protein, PD-L1, was rGFP-PDL1scFv produced by NZ-GFP-PDL1scFv exhib- immobilized on a multi-well plate, and a cell extract its immunoreactivity against the PD-L1 protein. of the nisin-stimulated NZ-GFP-PDL1scFv containing rGFP-PDL1scFv was added. The bound rGFP-PDL1scFv Purification of rGFP‑PDL1scFv from gmLAB was detected using an anti-His-tag antibody. As a result, rGFP-PDL1scFv was purified from the NZ-GFP-PDL1s - the absorbance at 450 nm increased with an increase in cFv cell extract. A chromatogram of the absorbance at the concentration of total protein in the NZ-GFP-PDL1s-280 nm (A ) is shown in Fig. 4a. Based on the absorb- cFv cell extract when added to PD-L1-immobilized wells ance values, the purified solution was separated into five (Fig. 3). In contrast, the absorbance remained at baseline fractions (F1–F5), and sample purity was assessed by levels when the NZ-GFP-PDL1scFv cell extracts were WB (Fig. 4b) or by SDS-PAGE followed by staining with added to wells that did not contain immobilized PD-L1 CBB (Fig. 4c). Bands corresponding to rGFP-PDL1scFv Namai et al. AMB Express (2023) 13:4 Page 7 of 11 Fig. 3 Immunoreactivity assay of NZ-GFP-PDL1scFv against PD-L1. A binding assay using an anti-His tag antibody was performed to verify the immunoreactivity of rGFP-PDL1scFv. Serially diluted cellular extracts of each gmLAB were added to wells immobilized (orange and black line) or not immobilized (gray line) with PD-L1 protein. The bound scFv was detected using an anti-His tag antibody, and the absorbance at 450 nm was measured. The data are representative of two independent experiments. Orange line: PD-L1+, NZ-GFP-PDL1scFv, Gray line: PD-L1-, NZ-GFP-PDL1scFv, Black Fig. 4 Purification of rGFP-PDL1scFv. a Chromatogram of protein line: PD-L1+, NZ-VC elution. Protein adsorbed on a His-Trap column was eluted by adding imidazole. The eluted protein was collected as five fractions (F1–F5) based on A . b The presence of rGFP-PDL1scFv in each fraction was analyzed by WB using an anti-His-tag antibody. c Results (59.0 kDa) and highly purified rGFP-PDL1scFv were of SDS-PAGE (CBB staining) analysis. The black arrow indicates the confirmed, particularly in fraction F2. Therefore, in sub - band corresponding to the size of rGFP-PDL1scFv (59.0 kDa) sequent experiments, F2 was used as purified rGFP- PDL1scFv. The ELISA results indicated that F2 contained 2.19 µg/mL (600 µL) of rGFP-PDL1scFv. Discussion PD-L1 is a well-studied immune checkpoint molecule. Cell surface antigen recognition ability of rGFP‑PDL1scFv Under normal conditions, PD-L1 is primarily expressed Next, we verified whether the rGFP-PDL1scFv pro - by antigen-presenting cells, such as monocytes, mac- duced by the gmLAB recognized PD-L1 expressed on rophages, and dendritic cells. It contributes to host the cell surface of Raw264.7 cells. Since Raw264.7 cells immune homeostasis by binding to PD-1 on T cells, express PD-L1 on the cell surface when stimulated with thereby suppressing excessive immunity (Francisco et al. LPS (Xiao et al. 2020), the cells were stimulated with 2010). In contrast, tumor cells are also known to express LPS (0, 1 or 10 µg/mL) and stained with commercially PD-L1, which enables them to escape antitumor immu- available PE anti-PD-L1 Ab. As a result, an increase in nity (Ai et al. 2020). The tumor microenvironment accu - the number of PE-positive cells was confirmed by flow mulates immune cells that highly express PD-L1 and cytometry, and the findings were compared with those PD-1, and these cell populations have been shown to obtained with no LPS stimulation (Fig. 5a). PE fluores - support tumor survival (Jiang et al. 2019a). Inhibition of cence was also observed by fluorescence microscopy PD-1/PD-L1 signaling is a target in developing antitumor using Raw 264.7 cells stimulated with LPS (Fig. 5d, f ). In strategies. An anti-PD-L1 antibody, atezolizumab, has addition, Raw264.7 cells were similarly stimulated and been developed and approved by the FDA as an antibody- stained with purified rGFP-PDL1scFv. Flow cytometry drug (Mathieu et al. 2021). By binding to PD-L1 in the showed a statistically significant increase in GFP-positive tumor microenvironment, anti-PD-L1 antibodies inhibit cells (Fig. 5b, c). Furthermore, fluorescence microscopy the interaction between PD-1 and PD-L1, reducing observed GFP fluorescence in LPS-stimulated Raw264.7 immunosuppression (Beyrend et al. 2019; Freeman et al. cells (Fig. 5g). In contrast, in the unstimulated cells, only 2000; Iwai et al. 2002). While these effects make anti-PD- nuclei stained with DAPI were observed (Fig. 5e). L1 antibodies potent antitumor agents, the high cost and Namai et al. AMB Express (2023) 13:4 Page 8 of 11 Fig. 5 Cell surface antigen recognition ability of rGFP-PDL1scFv. Raw264.7 cells (2.0 × 10 cells/well) were stimulated with lipopolysaccharide (LPS) (1 or 10 µg/mL) and stained using PE anti-mouse PD-L1 antibody or purified rGFP-PDL1scFv. Then fluorescence was detected by flow cytometry (a‑ c) and fluorescence microscopy (d‑ g). Representative flow cytometry histograms of staining with PE anti-mouse CD274 (B7-H1, PD-L1) antibody (a) or rGFP-PDL1scFv (b) are shown, respectively. Each white histogram indicates non-stimulated Raw264.7 cells. Gray histograms indicate LPS-stimulated Raw264.7 cells, and orange histograms indicate LPS-stimulated Raw264.7 cells. Black bars indicate the gated region used to calculate PD-L1-positive cells. c The graph shows the measurement results for the gated positive cell ratio in the stimulated or non-stimulated Raw264.7 cells stained with rGFP-PDL1scFv. Data are the mean ± SD (n = 3), and each dot on the plot represents one experiment. Representative fluorescence microscopy images of Raw264.7 cells stimulated with LPS (f, g) or without LPS (d, e) and stained with DAPI and PE anti-mouse CD274 (B7-H1, PD-L1) antibody (d, f) or rGFP-PDL1scFv (e, g) are shown (blue: DAPI, red: PE, green: GFP). Scale bar = 20 μm labor-intensive nature of producing antibody drugs have protein to which a flexible peptide linker attaches to been a bottleneck in their adoption (Samaranayake et al. the antigen recognition site of an antibody and which 2009). Against this background, we focused on scFv, a has a binding ability comparable to that of the original small-molecule antibody, to develop an affordable PD-1/ antibody (Bird et al. 1988; Ma and O’Kennedy 2017). PD-L1-signaling inhibition tool. scFv is a recombinant In addition, because of its simple structure, scFv can be Namai et al. AMB Express (2023) 13:4 Page 9 of 11 produced in bacterial heterologous protein expression Using an ELISA system, we investigated whether systems. This study employed a gmLAB strain as a host rGFP-PDL1scFv exhibits immunoreactivity against for producing PDL1scFv. PD-L1. Specifically, gmLAB cell lysates that produce Previously, we reported several gmLAB strains produc- rGFP-PDL1scFv were added to the target protein PD- ing scFv and verified their properties, such as their anti - L1-immobilized wells to confirm binding. As a result, gen recognition ability (Namai et al. 2020a, b; Shigemori the absorbance increased as the concentration of the cell et al. 2017a). Since gmLAB are developed using LAB, lysate increased. In contrast, no increase in absorbance they do not contain endotoxin. Therefore, scFv-produc - was observed when PD-L1 was not immobilized. Fur- ing gmLAB can be administered directly without the thermore, the control gmLAB did not increase absorb- purification of recombinant protein from the bacterial ance. These results, combined with the fact that the scFv cell (Shigemori and Shimosato 2017). Oral and intranasal was derived from an α-PD-L1 antibody, suggested that administration of gmLAB has been reported to prevent the scFv designed in this study could bind to PD-L1. We and/or alleviate disease in animal models by transport- verified whether rGFP-PDL1scFv could recognize PD-L1 ing recombinant proteins to local mucosal tissues (Namai expressed on the cell surface using cell lines in vitro. et al. 2020c; Yumoto et al. 2020). In this context, we Since preclinical studies, such as those using mouse developed a gmLAB strain that produces PDL1scFv as an models, will be necessary in the future, experiments were affordable tool for preventing and alleviating cancer. We conducted using mouse-derived macrophage cell lines. consider the findings to be significant because not only In addition, atezolizumab, the original antibody for the is PD-1/PD-L1 signaling prevented using this method, α-PD-L1 scFv designed in this study, is known to be a but the presence or absence of PD-L1 expression in the humanized human and mouse cross-reactive antibody tumor microenvironment is also an important consid- (Lesniak et al. 2016). Cells from the mouse macrophage eration for selecting an appropriate treatment regimen cell line, Raw264.7, are known to express PD-L1 protein (Shen and Zhao 2018). Consequently, to make the con- on the cell surface after stimulation with LPS (Xiao et al. structed gmLAB better suited for use as a diagnostic/ 2020). Here, we describe the binding of rGFP-PDL1scFv treatment tool, a fusion protein containing GFP added to to the cell surface PD-L1 using a fluorescence micro - the scFv was incorporated into the gmLAB. scope. Furthermore, green fluorescence attributable to We first designed the scFv based on the amino acid GFP was observed when purified rGFP-PDL1scFv was sequence of atezolizumab, an anti-PD-L1 antibody. The added to LPS-stimulated Raw264.7 cells. amino acid sequence was obtained by connecting the gmLAB based on intracellular production have been Fv regions of the heavy and light chains of atezolizumab reported, and mucosal administration has shown to have with a peptide linker, followed by conformational pre- desirable effects (Liu et al. 2018). In addition, previous diction to confirm whether the resulting product was studies have shown that nasally administered gmLAB in the form of scFv. The resulting DNA sequence was cleared more than 24 h after administration (Garcia inserted into the multiple cloning site of the lactococcal et al. 2018). These results suggest that the administered nisin-induced GFP expression vector (pNZ8148#2:CYT- gmLAB either rupture or are phagocytosed by immune GFP) using restriction digestion to create a GFP-fused cells and leak intracellular proteins. To verify these pro- anti-PD-L1 scFv-producing plasmid (pNZ8148#2:CYT- cesses and to assess the potential further application GFP-PDL1scFv). The resulting vector was introduced of the constructed gmLAB, conducting nasal admin- into NZ9000 to generate a gmLAB strain, and expres- istration experiments using mice will be necessary. In sion analysis of recombinant scFv was performed by WB. conclusion, we created a gmLAB strain that produces The constructed gmLAB produced rGFP-PDL1scFv only GFP-fused anti-PD-L1 scFv that shows immunoreactiv- upon adding nisin, a gene expression inducer, confirm - ity to PD-L1. Since PD-L1 is a target in many antitumor ing that gene expression occurred in a nisin stimulation- treatment strategies, and because rGFP-PDL1scFv can dependent manner. The fluorescence of the gmLAB was recognize PD-L1 and remain fluorescent green, we con - also examined, and green fluorescence was observed sider that the gmLAB strain developed in this study has only in nisin-stimulated NZ-GFP and NZ-GFP-PDL1s- the potential for use as an affordable antitumor agent and cFv, indicating that GFP produced as a fusion protein diagnostic tool. exhibits sufficient fluorescence for use as a transformant Acknowledgements marker. However, the fluorescence intensity for NZ-GFP- We thank Ms. Asami Ueda and Mr. Aito Murakami (Graduate School of Science PDL1scFv was weaker than that for NZ-GFP, suggesting and Technology, Shinshu University) for their technical support. We also thank the Research Center for Support of Advanced Science, Shinshu University, for that fusion with the scFv reduced expression levels and the use of their facilities. affected the 3D structure. Namai et al. AMB Express (2023) 13:4 Page 10 of 11 Author contributions Hill D, Sugrue I, Arendt E, Hill C, Stanton C, Ross RP (2017) Recent advances in FN, SS, and NN performed the experiments. FN, TSA, and TSH wrote the microbial fermentation for dairy and health. F1000Res 6:751. doi:https:// manuscript. FN, TSA, and TSH designed the research. TSA and TSH supervised doi. org/ 10. 12688/ f1000 resea rch. 10896.1 the work. All authors read and approved the final manuscript. Ishida M, Namai F, Shigemori S, Kajikawa S, Tsukagoshi M, Sato T, Ogita T, Shimosato T (2020) Ribosome-engineered Lacticaseibacillus rhamnosus Funding strain GG exhibits cell surface glyceraldehyde-3-phosphate dehydroge- This study was supported by a JSPS KAKENHI Grant (No. 19KK0208) to TSA nase accumulation and enhanced adhesion to human colonic mucin. and an Education and Research Support Fund from Shinshu University Appl Environ Microbiol. https:// doi. org/ 10. 1128/ AEM. 01448- 20 (2019–2022) to TSH. Iwai Y, Ishida M, Tanaka Y, Okazaki T, Honjo T, Minato N (2002) Involvement of PD-L1 on tumor cells in the escape from host immune system and tumor immunotherapy by PD-L1 blockade. Proc Natl Acad Sci U S A Declarations 99(19):12293–12297. doi:https:// doi. org/ 10. 1073/ pnas. 19246 1099 Jiang X, Wang J, Deng X, Xiong F, Ge J, Xiang B, Wu X, Ma J, Zhou M, Li X, Li Y, Li Ethics approval and consent to participate G, Xiong W, Guo C, Zeng Z (2019a) Role of the tumor microenvironment This article does not contain any studies with human participants or animals in PD-L1/PD-1-mediated tumor immune escape. Mol Cancer 18(1):10. performed by any of the authors. doi:https:// doi. org/ 10. 1186/ s12943- 018- 0928-4 Jiang Y, Chen M, Nie H, Yuan Y (2019b) PD-1 and PD-L1 in cancer immuno- Competing interests therapy: clinical implications and future considerations. Hum Vaccin The authors declare no competing interests. Immunother 15(5):1111–1122. doi:https:// doi. org/ 10. 1080/ 21645 515. 2019. 15718 92 Author details Jimenez M, Langer R, Traverso G (2019) Microbial therapeutics: new opportuni- Department of Biomolecular Innovation, Institute for Biomedical Sciences, ties for drug delivery. J Exp Med 216(5):1005–1009. doi:https:// doi. org/ 10. Shinshu University, 8304 Minamiminowa, Kamiina, Nagano 399-4598, Japan. 1084/ jem. 20190 609 Lesniak WG, Chatterjee S, Gabrielson M, Lisok A, Wharram B, Pomper MG, Nim- Received: 22 November 2022 Accepted: 8 January 2023 magadda S (2016) PD-L1 detection in tumors using [(64)Cu]Atezolizumab with PET. Bioconjug Chem 27(9):2103–2110. doi:https:// doi. org/ 10. 1021/ acs. bioco njchem. 6b003 48 Liu L, Zhang W, Song Y, Wang W, Zhang Y, Wang T, Li K, Pan Q, Qi X, Gao Y, Gao L, Liu C, Zhang Y, Wang Y, He G, Wang X, Cui H (2018) Recombinant References Lactococcus lactis co-expressing OmpH of an M cell-targeting ligand and Ai L, Xu A, Xu J (2020) Roles of PD-1/PD-L1 pathway: signaling, cancer, IBDV-VP2 protein provide immunological protection in chickens. Vaccine and beyond. Adv Exp Med Biol 1248:33–59. https:// doi. org/ 10. 1007/ 36(5):729–735. doi:https:// doi. org/ 10. 1016/j. vacci ne. 2017. 12. 027 978- 981- 15- 3266-5_3 Ma H, O’Kennedy R (2017) Recombinant antibody fragment production. Meth- Arbabi-Ghahroudi M, Tanha J, MacKenzie R (2005) Prokaryotic expression of ods 116:23–33. doi:https:// doi. org/ 10. 1016/j. ymeth. 2016. 11. 008 antibodies. Cancer Metastasis Rev 24(4):501–519. doi:https:// doi. org/ 10. Mathieu L, Shah S, Pai-Scherf L, Larkins E, Vallejo J, Li X, Rodriguez L, Mishra- 1007/ s10555- 005- 6193-1 Kalyani P, Goldberg KB, Kluetz PG, Theoret MR, Beaver JA, Pazdur R, Singh Beyrend G, van der Gracht E, Yilmaz A, van Duikeren S, Camps M, Hollt T, H (2021) FDA approval summary: atezolizumab and durvalumab in Vilanova A, van Unen V, Koning F, de Miranda N, Arens R, Ossendorp combination with platinum-based chemotherapy in extensive stage F (2019) PD-L1 blockade engages tumor-infiltrating lymphocytes to small cell lung cancer. Oncologist 26(5):433–438. https:// doi. org/ 10. 1002/ co-express targetable activating and inhibitory receptors. J Immunother onco. 13752 Cancer 7(1):217. doi:https:// doi. org/ 10. 1186/ s40425- 019- 0700-3 Namai F, Shigemori S, Sudo K, Sato T, Yamamoto Y, Nigar S, Ogita T, Shimosato Bird RE, Hardman KD, Jacobson JW, Johnson S, Kaufman BM, Lee SM, Lee T, T (2018a) Recombinant mouse osteocalcin secreted by Lactococcus lactis Pope SH, Riordan GS, Whitlow M (1988) Single-chain antigen-binding promotes Glucagon-Like Peptide-1 induction in STC-1 cells. Curr Micro- proteins. Science 242(4877):423–426. doi:https:// doi. org/ 10. 1126/ scien biol 75(1):92–98. doi:https:// doi. org/ 10. 1007/ s00284- 017- 1354-3 ce. 31403 79 Namai F, Yamamoto Y, Sato T, Ogita T, Shimosato T (2018b) Recombinant Cano-Garrido O, Seras-Franzoso J, Garcia-Fruitós E (2015) Lactic acid bacteria: mouse calcitonin gene-related peptide secreted by Lactococcus lactis reviewing the potential of a promising delivery live vector for bio- inhibits lipopolysaccharide-induced inflammatory response in mac- medical purposes. Microb Cell Fact 14:137. doi:https:// doi. org/ 10. 1186/ rophages. Anim Sci J 89(12):1707–1711. doi:https:// doi. org/ 10. 1111/ asj. s12934- 015- 0313-6 13115 de Ruyter PG, Kuipers OP, Beerthuyzen MM, van Alen-Boerrigter I, de Vos WM Namai F, Murakami A, Ueda A, Tsukagoshi M, Shigemori S, Ogita T, Sato T, Shi- (1996) Functional analysis of promoters in the nisin gene cluster of Lacto- mosato T (2020a) Construction of genetically modified Lactococcus lactis coccus lactis. J Bacteriol 178(12):3434–3439. doi:https:// doi. org/ 10. 1128/ producing Anti-human-CTLA-4 single-chain fragment variable. Mol Bio- jb. 178. 12. 3434- 3439. 1996 technol 62(11–12):572–579. https:// doi. org/ 10. 1007/ s12033- 020- 00274-8 Dermani FK, Samadi P, Rahmani G, Kohlan AK, Najafi R (2019) PD-1/PD-L1 Namai F, Shigemori S, Ogita T, Sato T, Shimosato T (2020b) Construction of immune checkpoint: potential target for cancer therapy. J Cell Physiol genetically modified Lactococcus lactis that produces bioactive anti- 234(2):1313–1325. doi:https:// doi. org/ 10. 1002/ jcp. 27172 interleukin-4 single-chain fragment variable. Mol Biol Rep. doi:https:// doi. Francisco LM, Sage PT, Sharpe AH (2010) The PD-1 pathway in tolerance and org/ 10. 1007/ s11033- 020- 05765-0 autoimmunity. Immunol Rev 236:219–242. doi:https:// doi. org/ 10. 1111/j. Namai F, Shigemori S, Ogita T, Sato T, Shimosato T (2020c) Microbial thera- 1600- 065X. 2010. 00923.x peutics for acute colitis based on genetically modified Lactococcus lactis Freeman GJ, Long AJ, Iwai Y, Bourque K, Chernova T, Nishimura H, Fitz LJ, hypersecreting IL-1Ra in mice. Exp Mol Med 52(9):1627–1636. doi:https:// Malenkovich N, Okazaki T, Byrne MC, Horton HF, Fouser L, Carter L, Ling V, doi. org/ 10. 1038/ s12276- 020- 00507-5 Bowman MR, Carreno BM, Collins M, Wood CR, Honjo T (2000) Engage- O’Toole PW, Marchesi JR, Hill C (2017) Next-generation probiotics: the ment of the PD-1 immunoinhibitory receptor by a novel B7 family spectrum from probiotics to live biotherapeutics. Nat Microbiol 2:17057. member leads to negative regulation of lymphocyte activation. J Exp doi:https:// doi. org/ 10. 1038/ nmicr obiol. 2017. 57 Med 192(7):1027–1034. doi:https:// doi. org/ 10. 1084/ jem. 192.7. 1027 Samaranayake H, Wirth T, Schenkwein D, Raty JK, Yla-Herttuala S (2009) Chal- Garcia PC, Paillavil BA, Scioscia N, Dale JB, Legarraga P, Salazar-Echegarai FJ, lenges in monoclonal antibody-based therapies. Ann Med 41(5):322–331. Bueno SM, Kalergis AM, Wozniak A (2018) Clinical and microbiological doi:https:// doi. org/ 10. 1080/ 07853 89080 26988 42 response of mice to intranasal inoculation with Lactococcus lactis express- Schwede T, Kopp J, Guex N, Peitsch MC (2003) SWISS-MODEL: an automated ing group A streptococcus antigens, to be used as an anti-streptococcal protein homology-modeling server. Nucleic Acids Res 31(13):3381–3385. vaccine. Microbiol Immunol 62(11):711–719. https:// doi. org/ 10. 1111/ doi:https:// doi. org/ 10. 1093/ nar/ gkg520 1348- 0421. 12657 Namai et al. AMB Express (2023) 13:4 Page 11 of 11 Shen X, Zhao B (2018) Efficacy of PD-1 or PD-L1 inhibitors and PD-L1 expres- sion status in cancer: meta-analysis. BMJ 362:k3529. doi:https:// doi. org/ 10. 1136/ bmj. k3529 Shigemori S, Shimosato T (2017) Applications of genetically modified immunobiotics with high immunoregulatory capacity for treatment of inflammatory bowel diseases. Front Immunol 8:22. https:// doi. org/ 10. 3389/ fimmu. 2017. 00022 Shigemori S, Yonekura S, Sato T, Nakanishi M, Otani H, Shimosato T (2012) Expression of a biologically active GFP-α(S1)-casein fusion protein in Lactococcus lactis. Curr Microbiol 64(6):569–575. doi:https:// doi. org/ 10. 1007/ s00284- 012- 0111-x Shigemori S, Ihara M, Sato T, Yamamoto Y, Nigar S, Ogita T, Shimosato T (2017a) Secretion of an immunoreactive single-chain variable fragment antibody against mouse interleukin 6 by Lactococcus lactis. Appl Microbiol Biotech- nol 101(1):341–349. doi:https:// doi. org/ 10. 1007/ s00253- 016- 7907-8 Shigemori S, Namai F, Yamamoto Y, Nigar S, Sato T, Ogita T, Shimosato T (2017b) Genetically modified Lactococcus lactis producing a green fluorescent protein-bovine lactoferrin fusion protein suppresses proin- flammatory cytokine expression in lipopolysaccharide-stimulated RAW 264.7 cells. J Dairy Sci 100(9):7007–7015. doi:https:// doi. org/ 10. 3168/ jds. 2017- 12872 Xiao Z, Su Z, Han S, Huang J, Lin L, Shuai X (2020) Dual pH-sensitive nanodrug blocks PD-1 immune checkpoint and uses T cells to deliver NF-kappaB inhibitor for antitumor immunotherapy. Sci Adv 6(6):eaay7785. doi:https:// doi. org/ 10. 1126/ sciadv. aay77 85 Yumoto K, Sato T, Nakashima K, Namai F, Shigemori S, Shimosato T, Kaneko T (2020) Nasally administered Lactococcus lactis secreting heme oxyge- nase-1 attenuates murine emphysema. Antioxid (Basel). https:// doi. org/ 10. 3390/ antio x9111 049 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations.
Journal
AMB Express – Springer Journals
Published: Jan 12, 2023
Keywords: gmLAB; Next-generation probiotics; PD-L1; scFv; Tumor microenvironment