Plasmids and cell lines

The construction and characterization of Chinese hamster ovary (CHO) cells expressing human C5a receptor 1 (C5aR1) and human C5aR1-green fluorescent protein (GFP) have been described previously [43]. pCMV6 mouse C5aR1 was purchased from OriGene Technologies (catalog number MC208206; GenBank Accession NM_007577). The mouse C5aR1-GFP construct was created by deleting the C5aR1 stop codon and inserting the GFP-coding cDNA into the Mlu I-Not I restriction sites in the pCMV6 vector. Human CD177 cDNA in the pDNR-Dual plasmid (clone HsCD00000952; GenBank Accession AY888472) was obtained from the DNASU Plasmid Repository. The cDNA was cloned into the constitutive expression vectors pBGSA and pBZSA to allow for selection in CHO cells with G418 and Zeocin (or Bleocin), respectively [44]. Mouse (Mus musculus) CD177 cDNA was obtained from OriGene Technologies in the pCMV6-Kan/Neo plasmid (catalog number MC201519; GenBank Accession BC027283). Since antibodies against mouse CD177 were not available at the start of this project, two different tags were tested: A c-myc tag (EQKLISEEDL) placed either two amino acids downstream from the predicted signal peptide cleavage site, or 4 or 21 amino acids upstream from the predicted single transmembrane domain, resulted in misfolded proteins that were mostly retained in the endoplasmic reticulum, based on immunofluorescence microscopy. Similarly, an HA tag (YPYDVPDYA) two amino acids downstream from the predicted signal peptide cleavage site caused partial ER retention. An HA tag four amino acids upstream from the predicted single transmembrane domain was successfully expressed on the cell surface with no evidence of intracellular retention. CHO cells expressing HA-tagged formyl peptide receptor 1 (HA-FPR1) were used in the phage display peptide library screening to eliminate HA-binding peptides [45]. The CHO cell line was obtained in 1987 by HMM while working in the laboratory of Dr. Ira Mellman at Yale University. This cell line is not listed in the Database for Cross-Contaminated or Misidentified Cell Lines by the International Cell Line Authentication Committee (ICLAC).

Antibodies and peptides

Mouse monoclonal anti-human CD177 antibody, clone MEM-166, catalog # 551899, was from BD Biosciences. Mouse monoclonal anti-HA-tag antibody, HA.11 clone 16B12, catalog # MMS-101R, was from BAbCO/Covance Research Products. Rabbit polyclonal anti-mouse CD177 antibody, clone 1171A, catalog # MAB8186, was from R&D Systems. Mouse monoclonal anti-human LAMP2 antibody, clone H4B4, catalog # sc-18822 was from Santa Cruz, Inc. Mouse monoclonal anti-hamster LAMP2 antibody (UH3) was a generous gift from Dr. Bruce L. Granger [44]. The antibodies against human and mouse CD177 and against the HA-tag showed no binding to wild-type CHO cells by immunofluorescence microscopy, validating their target specificity. The mouse anti-hamster LAMP2 antibody has been previously validated by immunocytochemistry and western blot analysis [44]. Secondary Alexa Fluor 488 goat anti-mouse monoclonal antibody, catalog # A-21121, and Alexa Fluor 488 goat anti-rabbit antibody, catalog # A-11008, were from Molecular Probes. The human CD177-binding peptide (Peptide H) and the scrambled control peptide were synthesized by GenScript, and the mouse CD177-binding peptide was synthesized by Peptide 2.0 (sequences are listed in List A in S1 File).

siRNAs and antisense oligonucleotides

All siRNAs were purchased from Dharmacon (ThermoScientific): ON-TARGET plus non-targeting siRNA#1 (catalog # D-001810-01) was used as negative control siRNA. GFP Duplex I (catalog # P-002048-01) was used as positive control siRNA (ref# 4303). ON-TARGET plus human C5aR1 siRNA pool consisting of four different 19-mers (catalog #J-005442) was used for silencing of human C5aR1-GFP. Due to relatively poor results using pooled mouse C5aR1 siRNAs (catalog # J-043176), each mouse C5aR1 siRNA was also purchased and tested separately (-5, -6, -7, -8) with only -6 resulting in significant knockdown (S2 Fig and S3 Fig). Antisense mouse C5aR1 LNA™ GapmeR oligonucleotides with a phosphorothioate backbone were produced by Exiqon, Inc., with design based on proprietary software (Design ID # 580726–1 and 580726–2). An ASO produced by Integrated DNA Technologies to knockdown mouse C5aR1 had no effect and was used as the non-target control (NTC). The siRNA and ASO sequences can be found in List B and List C in S1 File.

HPLN materials and other reagents

Hybrid Polymerized Liposomal Nanoparticles (HPLNs) were synthesized by NanoValent Pharmaceuticals, Inc. (Bozeman, MT). The components comprising the HPLN liposomes are hydrogenated soy L-α-phosphatidylcholine (“hydrogenated soy PC”), cholesterol and 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-2000] (“m-Peg₂₀₀₀-DSPE”) (all from Avanti Polar Lipids, Alabaster, Alabama), N-(5’-hydroxy-3’-oxypentyl)-10-12-pentacosadiynamide (“h-Peg₁-PCDA”), and N-(methoxy(polyethylene glycol)-2000)-10-12-pentacosadiynamide (“h-Peg₂₀₀₀-PCDA”).

Phage display peptide library screening

Two libraries, PhD-7 and PhD-12 (New England Biolabs), were used to screen for peptides that bind human CD177. To eliminate peptides that bind to CHO cells, 1 x 10⁹ phage in PBS containing calcium and magnesium (PBS⁺⁺), and supplemented with 0.5% bovine serum albumin (BSA), were incubated three rounds with wild-type CHO cells plated on 10 cm tissue culture plates, each for 1 h at room temperature on a rocker. Unbound phage were then incubated with CHO cells expressing human CD177 on a 10 cm tissue culture plate for 3 h at room temperature. Cells were rinsed ten times with PBS⁺⁺ with 0.5% BSA and 0.1% Tween-20. Cell-bound phage were eluted with 0.1 M glycine-HCl, pH 2.2, for 10 min on ice with rocking and removed from the plate and neutralized with Tris-HCl, pH 8.8. To release internalized phage, cells were then incubated with 1% TritonX-100, 0.1 M glycine-HCl, pH 2.2, for 10 min on ice, removed from the plate and neutralized. Cell debris was removed by centrifugation. The eluted phage were amplified according to the instructions by New England Biolabs and the peptide library screening was repeated two times as above. The library screening for mouse CD177-binding peptides was carried out as above, except CHO cells expressing HA-FPR were used instead of wild-type CHO cells to eliminate peptides that bound to the HA-tag (in addition to endogenous CHO molecules). The DNA sequences of the phage peptides bound and internalized by CHO human CD177 cells and by CHO mouse CD177-HA cells were determined after the third round of selection. The best peptide for each CD177 was selected by comparing binding and internalization by CHO cells expressing CD177 to cells without CD177.

Synthesis of CD177-binding lipid (peptide H- or M-Peg₂₀₀₀-DSPE) and scrambled peptide lipid

The CD177-binding dodecapeptide lipid was prepared by first exposing 1.0 ml of an 8.8 mg/ml aqueous solution of cysteine-terminated dodecapeptides (Peptide H, Peptide M or scrambled) to resin-supported TCEP reducing agent (Sigma-Aldrich) by shaking for 15 minutes. The solution was filtered directly into a flask containing 17.6 mg of powdered 1, 2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[maleimide(polyethylene glycol)-2000] (“mal-Peg₂₀₀₀-DSPE”) (Avanti Polar Lipids, Alabaster, Alabama) and stirred under nitrogen. The powdered DSPE lipid dissolved completely after several hours and was allowed to react overnight at ambient temperature. Thin layer chromatography (TLC) analysis on silica gel plates (Kieselgel 60 F₂₅₄, Merck KGaA, Darmstadt, Germany) with chloroform/methanol solvent mixture (4/1) as eluent showed that all the mal-Peg₂₀₀₀-DSPE was consumed and converted to a lower Rf spot. The solution was dried to a waxy solid and redissolved in 1/1 chloroform/methanol and filtered through Celite yielding Peptide H, Peptide M or scrambled peptide-Peg₂₀₀₀-DSPE.

Production of HPLN particles

Non-targetable HPLNs were prepared from h-PEG₁PCDA, hydrogenated soy PC, cholesterol, m-Peg₂₀₀₀-DSPE and m-PEG₂₀₀₀-PCDA at a molar proportion of 14:43:32:10:1, and targetable HPLNs were prepared from h-PEG₁PCDA, hydrogenated soy PC, cholesterol, Peptide H or Peptide M-Peg₂₀₀₀-DSPE and m-PEG₂₀₀₀-DSPE at a molar proportion of 14:43:32:10:1, according to the method previously described [46]. Briefly, lipids were mixed and evaporated in vacuo to a film. Deionized water was added to the films so as to give a 25 mM (total lipid and cholesterol) suspension. The suspension was heated via sonication at 70–80°C with a probe-tip sonicator (Fisher sonic dismembrator model 300) for 10 min. The resulting nearly clear solution was then cooled to 5°C for 12 h. After warming to ambient temperature, the liposomes were polymerized by UV light irradiation (254 nm) with a Spectrolinker XL-1000 UV Crosslinker (Spectronics Corp.) for 1 min. The resulting blue HPLNs were heated to 65°C for 5 min to convert them to the red (fluorescent) form. The colored solutions were syringe filtered through 0.2 μm cellulose acetate filters in order to remove trace insoluble contaminants. The HPLN concentration was determined using the ammonium ferrothiocyanate assay [47]. The particle size (average ~96 nm) was determined on Malvern ZetaSizer in water.

Immunofluorescence microscopy

CHO cells stably expressing human CD177 or HA-tagged mouse CD177, grown on glass coverslips, were incubated with various HPLN nanoparticles at 75 μg/ml, as indicated in the Figure legends. Cells were washed to remove unbound HPLNs, fixed with 2.5% paraformaldehyde in PBS, permeabilized and blocked with 0.01% saponin/0.2% gelatin in PBS, and stained with primary and secondary fluorescent antibodies, as indicated in the Figure legends. In experiments examining internalization of HPLNs, cells were incubated with or without subtilisin (1 mg/ml) in PBS containing 1 mM DTT and 5 mM EDTA. Since this treatment resulted in loss of cell adherence, the cells were centrifuged at 300 rpm for 2 min onto glass slides using a Cytospin 4™ Cytocentrifuge. Human or mouse anticoagulated blood was incubated with HPLN nanoparticles at 50 μg/ml for the indicated times. Red blood cells were separated from buffy coat and plasma by 5 min centrifugation at 420 x g. Most of the plasma was discarded and the buffy coat was collected into a new tube. The remaining red blood cells in the buffy coat were lysed with ice cold water (3x), NaCl was added to a final concentration of 150 mM, and the neutrophils were pelleted and suspended in RPMI-1640 supplemented with 10% fetal bovine serum. The neutrophils were centrifuged at 300 rpm for 2 min onto glass slides using a Cytospin 4™ Cytocentrifuge. Cells were fixed 5 min in cold methanol, blocked in 0.2% gelatin/PBS and stained with primary and secondary fluorescent antibodies, as indicated in the Figure legends. Collection of human blood was approved by the Montana State University Institutional Review Board for the Protection of Human Subjects (FWA00000165 and Protocol Number: HM100909). Healthy volunteers gave written informed consent. Mouse blood was obtained from the Montana State University Animal Resources Center, a registered facility with the U.S. Department of Agriculture that complies with the Public Health Service Center Policy of Use of Laboratory Animals (AAALAC #713 and Animal Welfare #A3627-01.

Flow cytometry

Human leukocytes were isolated from whole blood as described above. Cells were incubated on ice for 30 min in the presence or absence of mouse anti-human CD177 antibody, then centrifuged for 5 min at 420 x g and suspended in PBS supplemented with 10% fetal bovine serum. Cells were incubated for 30 min on ice in the presence or absence of Peptide H-HPLN (37.5 μg/ml) and Alexa Fluor 488 goat anti-mouse IgG (1:100). Leukocytes (and remaining red blood cells) were analyzed with a Becton-Dickinson Accuri C6 flow cytometer with a 488 nm excitation laser and 533/30 nm FL1 filter for detecting Alexa Fluor 488 and a 585/40 nm FL2 filter for detecting the HPLNs. The neutrophil population was gated in the FSC and SSC channels.

For siRNA studies, 5x10⁴ CHO cells expressing human C5aR1-GFP or mouse C5aR1-GFP were plated in 500 μl medium in 24-well plates one day prior to transfection. Cell were transfected with 25–100 nM siRNA using Dharmafect 1 (Dharmacon) or TransIT-X2 (Mirus Bio LLC) transfection reagents, according to the manufacturer’s protocol. For ASO studies, transfections were carried out with no ASO (mock transfected), 50 nM or 100 nM ASO (C5aR1 LNA GapmeR_1) or 100 nM non-targeting control (NTC) using Xfect™ RNA transfection reagent (Takara Bio USA) or TransIT®-Oligo transfection reagent (Mirus Bio, LLC) using the manufacturer’s protocol. Culture medium was replaced every 24 h. 72 h post transfection, cells were removed from tissue culture plates with trypsin-EDTA, centrifuged and suspended in PBS containing 10% FBS and 1 μg/ml propidium iodide (PI). Cell suspensions were analyzed with a Becton-Dickinson Accuri C6 flow cytometer with a 488 nm excitation laser and 533/30 nm FL1 filter for detecting GFP and a 585/40 nm FL2 filter for detecting PI fluorescence. Cells were gated in the FSC and SSC channels to exclude large/clumped cells and highly vacuolar cells (indicating poor health), and in the FL2 channel to exclude dead cells based on their PI fluorescence.

Reverse transcription quantitative PCR (RT-qPCR)

Total RNA was purified 72 h post transfection from the same cells as had been analyzed in flow cytometry experiments using the Qiagen RNeasy Plus micro kit (catalog # 74034) and 100 ng total RNA was reverse transcribed using the Qiagen miScript II RT kit (catalog # 218160). qPCR was carried out in a Roche LightCycler® using 2x QuantiTect SYBR green mix (Qiagen; catalog # 204143), 200 nM of each primer, and cDNA equivalent of 5 ng reverse transcribed RNA. Primers for CHO housekeeping genes were selected from the study carried out by Bahr and coworkers [48]. Two different sets of mouse C5aR1 primers were used to control for possible annealing and experimental variability. Primer sequences are shown in List D in S1 File and cycling conditions are shown in Table A in S1 File. The data were analyzed using the ΔΔCq calculation using the technical notes provided by Dharmacon [49] based on the MIQE guidelines [50].

Identification of peptides binding to mouse and human CD177

As outlined in the Introduction, one challenge in developing neutrophil-specific therapeutics is finding neutrophil-specific targets that do not activate those cells. We chose as our target CD177, a GPI-anchored glycoprotein with no known signaling function and limited expression in other cell types [31, 32]. Bacteriophage peptide libraries were used to select peptides that bind human and mouse CD177 stably expressed on the surfaces of CHO cells. Whereas selection with the PhD-7 library did not result in phage peptides with specific binding to human or mouse CD177, the PhD-12 library resulted in several promising dodecapeptides (Table 1). Further quantitative binding analysis identified the best peptides for mouse and human CD177 (Table 2). Neutrophil targeting molecules have previously been identified by two research groups: 1) Two human neutrophil-binding peptides were discovered using a random phage peptide display library [51], but the identities of the receptors remain unknown. The first peptide activated cell signaling through an unknown G protein coupled receptor and is therefore not ideal for therapeutic targeting because of this uncertainty [52]. The second peptide was found to target liposomal nanoparticles into monocytes as well as neutrophils, and therefore lacks cell type specificity [53]. 2) Wang et al. showed that albumin-coated nanoparticles could be targeted to activated neutrophils bound to vascular endothelial cells; these nanoparticles did not contain a targeting peptide, but relied on endocytosis mediated in part by the FcγIII receptor on the neutrophils and by other unknown receptor(s) [54, 55]. The advantages of our peptides over those listed above are three-fold: 1) the receptor is known; 2) the peptide binding does not cause neutrophil activation; 3) the receptor is neutrophil-specific.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Amino acid sequences of peptides that bind human and mouse CD177.

https://doi.org/10.1371/journal.pone.0200444.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Quantitative phage peptide binding assay.

https://doi.org/10.1371/journal.pone.0200444.t002

Generation and testing of HPLNs with targeting peptides

Initially, the targeting peptides were expressed as genetic fusions on the coat protein of Salmonella typhimurium bacteriophage P22 viral-like particle and on the phage decoration protein, Dec [56, 57]. However, binding experiments with CHO cells expressing human or mouse CD177 gave poor and inconsistent results. Next, we explored HPLNs that have been previously used for cell targeting [58]. The peptides were synthesized with a glycine spacer and a C-terminal cysteine for chemical conjugation to mal-Peg₂₀₀₀-DSPE. These hybrid polymerized liposomal nanoparticles (HPLNs) share many structural attributes of conventional liposomes. They are self-assembling, unilamellar spheres whose surfaces can be modified using the same chemical coupling strategies as employed for liposomes. Unlike liposomes, HPLNs can be manufactured to be intrinsically fluorescent. Ultraviolet irradiation leads to cross-linking of diacetylene residues present in their acyl chains, leading to highly colored blue particles, and heat treatment of the HPLN vesicles leads to color change and fluorophore formation [59–61]. The fluorescence emission spectrum peaks at 550–570 nm with a broad and complex excitation spectrum from 470–540 nm. As a result, HPLNs converted into their fluorescent form can be readily traced from the time they bind to target cells until they are deposited and compartmentalized into subcellular structures.

Binding and internalization was first examined using CHO cells stably expressing human CD177 or mouse CD177 with an HA-tag (CD177-HA). Incubation of the human peptide (Peptide H) HPLN particles with CHO human CD177 cells showed binding and apparent internalization of the HPLNs, whereas HPLN particles with a scrambled peptide showed no binding or internalization (Fig 1). Since delivery to lysosomes could result in nucleotide degradation by lysosomal enzymes, we examined the co-localization of the HPLN particles with the lysosomal marker, LAMP2. No co-localization was apparent up to 3 h of continuous incubation with the nanoparticles, but 17 h incubation resulted in some co-localization, suggesting that at least some of the nanoparticles are eventually routed to lysosomes (Fig 2). Similarly, HPLNs with the mouse CD177 binding peptide (Peptide M), were shown to bind CHO cells expressing mouse CD177-HA (Fig 3). To further confirm that the HPLN particles become internalized, we examined the distribution of the Peptide M-containing particles after treatment with subtilisin, a protease that will cleave HPLN particles from the cells unless endocytosis has taken place. Cells were incubated on ice with Peptide M-HPLN particles and either moved to 37°C to allow internalization, or kept on ice. Cells were then incubated with subtilisin, followed by fixation and processing for immunofluorescence microscopy. As shown in Fig 3, subtilisin treatment not only removed the surface bound HPLN particles, but also mouse CD177-HA, whereas internalized particles and mouse CD177-HA (37°C) were protected from protease activity.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. HPLNs displaying Peptide H bind human CD177 on cells.

CHO cells expressing human CD177 were incubated for 3 h at 37°C with 75 μg/ml Peptide H-HPLNs. Cells were fixed and permeabilized, and CD177 was detected using a mouse anti-human CD177 antibody and an Alexa Fluor 488 goat anti-mouse secondary antibody. The HPLN particles are fluorescent red. As a negative control, HPLN particles displaying a scrambled peptide were used. The experiment was carried out twice, with at least 100 cells examined in each experiment. A direct correlation between the expression level of CD177 (based on fluorescence intensity) and the Peptide H-HPLN signal could be observed. The cell in the lower panel shows a typical cell distribution of CD177 and Peptide H-HPLNs after 3 h incubation. In contrast, the two cells shown in the upper panel are characteristic of the CD177 staining in the absence of Peptide H-HPLNs and in the presence of Scrambled Peptide H-HPLNs. Scale bar, 50 μm.

https://doi.org/10.1371/journal.pone.0200444.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Some Peptide H-HPLN particles co-localize with LAMP2 after prolonged incubation, suggesting delivery into the lysosomes.

CHO cells expressing human CD177 were incubated for 3 h and 17 h at 37°C with 75 μg/ml Peptide H-HPLN particles. Cells were fixed and permeabilized, and the lysosomes were stained with a mouse anti-hamster LAMP2 antibody and an Alexa Fluor 488 goat anti-mouse secondary antibody. The experiment was carried out once. The selected micrographs are representative of more than 100 cells with similar labeling patterns. Scale bar, 50 μm.

https://doi.org/10.1371/journal.pone.0200444.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. HPLN particles displaying Peptide M become internalized by CHO cells expressing mouse CD177-HA.

CHO cells expressing mouse CD177 containing an HA-tag were incubated with 75 μg/ml Peptide M-HPLN particles for 2 h on ice to allow binding. Cells were then washed and kept on ice or warmed to 37°C for 1 h. Cells were then treated with subtilisin to remove surface-bound Peptide M-HPLN particles, or treated with buffer alone. After fixation and permeabilization, mouse CD177-HA was stained with a mouse anti-HA antibody and an Alexa Fluor 488 goat anti-mouse secondary antibody. The experiment was carried out once. Most cells out of more than 100 cells visualized on the slides had a similar staining pattern as those seen in these micrographs. Scale bar, 50 μm.

https://doi.org/10.1371/journal.pone.0200444.g003

HPLNs show specificity for neutrophils when incubated with whole blood from humans and mice

Therapeutic targeting of neutrophils requires that the binding and internalization of the HPLNs is cell-type specific. We therefore incubated Peptide H-HPLNs with whole human blood to examine the distribution of human CD177 and the HPLNs by immunofluorescence microscopy of white blood cells (after removal of most of the red blood cells). As shown in Fig 4, only cells that stained with anti-human CD177 antibody and had a lobed nucleus—typical of neutrophils—were associated with Peptide H-HPLNs. Similarly, when mouse whole blood was incubated with Peptide M-HPLNs, only cells positive for mouse CD177 showed staining with the Peptide M-HPLN particles. HPLNs containing the human CD177 binding Peptide H did not bind to mouse CD177, suggesting species specificity (Fig 5). The specificity of Peptide H-HPLN was confirmed by flow cytometry—only CD177-positive neutrophils were positive for Peptide H-HPLN (Fig 6). A time course immunofluorescence experiment using purified human neutrophils showed a homogeneous surface distribution of CD177 and Peptide H-HPLN after a 15 min incubation at 37°, whereas incubations for 30, 60 and 120 min showed an increasing amount of redistribution, suggestive of endocytosis (Fig 7). Our attempts to confirm endocytosis using subtilisin failed because of neutrophil sensitivity to the treatment. Incubation with Peptide H-HPLN for 120 min did not show colocalization with the lysosomal marker, LAMP2, in agreement with the results using the CHO transfectants (S1 Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Peptide H-HPLN particles bind CD177-expressing human neutrophils in whole blood.

Peptide H-HPLN particles were incubated with whole human blood for 1 h at 37°C. The white blood cells were collected from the buffy coat and the red blood cells were lysed. The white blood cells were centrifuged onto glass slides, fixed in methanol and stained with mouse anti-human CD177 antibody and an Alexa Fluor 488 goat anti-mouse secondary antibody. Scrambled Peptide H-HPLN particles were used as a negative control. The experiment was carried out four times using blood from four different donors. In each case, slides with several hundred cells were examined and the micrographs are representative of these results. Scale bar, 10 μm.

https://doi.org/10.1371/journal.pone.0200444.g004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Peptide M-HPLN particles bind CD177-expressing mouse neutrophils in whole blood.

Peptide M-HPLN particles were incubated with whole mouse blood for 1 h at 37°C. The white blood cells were collected from the buffy coat and the red blood cells were lysed. The white blood cells were centrifuged onto glass slides, fixed in methanol and stained with rabbit anti-mouse CD177 polyclonal antibody and Alexa Fluor 488 goat anti-rabbit secondary antibody. Peptide H-HPLN particles were used as a negative control. The experiment was carried out twice. The micrographs represent the results from >100 cells. Scale bar, 10 μm.

https://doi.org/10.1371/journal.pone.0200444.g005

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Flow cytometry confirms Peptide H-HPLN binding to CD177-positive neutrophils in a pool of purified human leukocytes.

Leukocytes were incubated on ice in the presence or absence of anti-CD177 antibody, washed and incubated on ice in the presence or absence of Peptide H-HPLN and Alexa Fluor 488 goat anti-mouse antibody. A. Forward and side scatter plot of human leukocytes (and remaining red blood cells) with the neutrophil population shown in a circle. B. FL-1 histogram of CD177 expression. Left panel: Total cell population in the absence and presence of anti-CD177 antibody (1° Ab) and Alexa Fluor 488 goat anti-mouse antibody (2° Ab). Right panel: Gated neutrophil population in the absence or presence of anti-CD177 antibody (1° Ab) and Alexa Fluor 488 goat anti-mouse antibody (2° Ab). C. FL2 plot showing Peptide H-HPLN binding. Total cell population in the absence or presence of Peptide H-HPLNs (left panel) and gated neutrophil population in the absence and presence of Peptide H-HPLNs (right panel). D. FL1-A and FL2-A plot showing fluorescence of the total cell population (left panel) and gated neutrophil population (right panel). The experiment was carried out twice with similar results. The blood sample shown had a higher percentage of CD177-positive neutrophils than the blood sample from the other donor.

https://doi.org/10.1371/journal.pone.0200444.g006

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. Time course of CD177 and Peptide H-HPLN redistribution in human neutrophils.

Purified human neutrophils were incubated with Peptide H-HPLNs at 37°C with aliquots taken at 0, 15, 30, 60 and 120 min. Cells were rinsed with PBS, centrifuged onto glass slides, fixed in methanol and stained with mouse anti-human CD177 antibody and an Alexa Fluor 488 goat anti-mouse secondary antibody. The experiment was carried out once using blood from a donor with a high CD177 expression level. Hundreds of cells were examined and the micrographs are representative of these results. Scale bar, 10 μm.

https://doi.org/10.1371/journal.pone.0200444.g007

Knockdown of mouse and human C5aR1 using siRNA and antisense oligonucleotides (ASOs)

In vitro knockdown analysis of human and mouse neutrophils is not possible because of the relatively short half-life of these cells after isolation from blood (~8 hours in vitro compared to ~5.4 days in vivo) [62]. To examine C5aR1 knockdown, we generated stable cell lines that express various combinations of human or mouse C5aR1, C5aR1-GFP and CD177. Protein knockdown was measured by flow cytometry, and mRNA knockdown was measured by RT-qPCR. For protein knockdown, we used human and mouse C5aR1 with a C-terminal fusion of green fluorescent protein (GFP). The benefits of measuring GFP rather than C5aR1 are three-fold; 1) avoiding any potential background labeling due to non-specific binding of primary and secondary antibodies; 2) avoiding antibody titrations to ensure binding site saturation; 3) providing an internal positive control for knockdown using previously tested and commercially available GFP siRNA [63]. Transfection utilized a variety of proprietary, commercially-available lipid formulations (see Materials and methods), with our ultimate goal being utilization of our targeted HPLNs for this purpose. Our initial experiments using a pool of four different mouse C5aR1 siRNAs showed very poor knockdown compared to our positive GFP siRNA control (S2 Fig). We therefore tested the siRNAs individually and found that only one out of four siRNAs (siRNA-6) resulted in knockdown comparable to GFP siRNA (S3 Fig). This siRNA was used in the subsequent experiments. As shown in Fig 8, the mouse C5aR1 siRNA resulted in a 91% decrease in mouse C5aR1-GFP fluorescence compared to the negative control siRNA, and the GFP siRNA positive control resulted in an 85% decrease. The knockdown of human C5aR1-GFP was somewhat less efficient at 67% for both the human C5aR1 siRNA pool and the GFP siRNA (Fig 8).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 8. Mouse C5aR1 siRNA and human C5aR1 siRNA pool result in receptor knockdown.

CHO cells expressing mouse C5aR1-GFP were transfected with 100 nM mouse C5aR1 ON-TARGETplus SMART siRNA–6, 100 nM GFP siRNA (positive control), or 100 nM negative control siRNA. 72 h post transfection cells were analyzed by flow cytometry to measure the relative expression of mouse C5aR1-GFP (left panel). CHO cells expressing human C5aR1-GFP were transfected with 100 nM human C5aR1 ON-TARGETplus SMARTpool siRNA, 100 nM GFP siRNA (positive control), or 100 nM negative control siRNA (right panel). Relative knockdown is based on the percentage of the cells that are to the left of the gate relative to the negative control sample. The experiment was carried out twice with similar results.

https://doi.org/10.1371/journal.pone.0200444.g008

As a more stable alternative to siRNA, we also examined knockdown using antisense oligonucleotides from Exiqon. These third generation LNA GapmeR™ ASOs contain a phosphorothioate backbone and locked nucleotides in each end making them far superior to the previous ASOs in potency, specificity, and stability [64, 65]. Our first test compared the knockdown effect of two mouse C5aR1 ASOs. Based on these results (S4 Fig), all future experiments were carried out using ASO1 (sequence shown in List C in S1 File). Flow cytometry 72 h after transfection with ASO1 showed about 85% decrease in the fluorescence of mouse C5aR1-GFP; but with a somewhat larger cell-to-cell variation in receptor knockdown compared to the siRNAs (Fig 9A). To analyze the knockdown further, we carried out RT-qPCR. Reference (housekeeping) genes for RT-qPCR were selected based on their stable expression levels under different conditions of CHO cell growth [48]. These reference genes were also ideal since the mRNA levels were higher (Eif3i) or lower (Vetz) than the mouse C5aR1 mRNA levels, and their primer pairs had been previously been validated (S5 Fig) [46]. We confirmed that the expression levels of the reference genes were not affected by ASO1 transfection (S6 Fig). The results showed about 89–92% knockdown of mouse C5aR1-GFP with 100 nM ASO1, in good agreement with the flow cytometry results (Fig 9B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 9. Mouse C5aR1 ASO results in knockdown of mouse C5aR1-GFP in CHO transfectants.

CHO cells expressing mouse C5aR1-GFP were transfected with 50 nM or 100 nM LNA GapmeR ASO. The cells were analyzed for C5aR1-GFP expression and mRNA levels 72 h post transfection. A. Relative receptor knockdown was measured by flow cytometry. The percentage knockdown was calculated based on the number of cells to the left of the gate relative to the negative control ASO. B. Relative gene expression was calculated from quantification cycle (Cq) values obtained by RT-qPCR using the ΔΔCq method. To control for possible experimental variation, the qPCR was carried out using two sets of mouse C5aR1 primers (C5aR1 208–402 and 221–430), and two sets of reference primers. The results in the left panel show the relative quantity of C5aR1 mRNA normalized to Eif3i, and the results in the right panel show the relative quantity of C5aR1 mRNA normalized to Vezt. Mock transfected cells received no ASO and non-targeting control (NTC) cells were transfected with a non-targeting ASO. The RT-qPCR was carried out with triplicate samples ± SD. One-way analysis of variance at 95% confidence interval showed that the relative mRNA expression levels were significantly lower in the ASO treated cells compared to the mock transfected and non-targeting ASO cells (p value <0.0001; ***).

https://doi.org/10.1371/journal.pone.0200444.g009

Future research

Cell-specific targeting and cytosolic (or nuclear) delivery of peptides, proteins, and nucleic acids remains a well-known obstacle in drug delivery. In many cases, biotherapeutic agents are sensitive to the low pH and degradative enzymes found in lysosomes [66–68]. To achieve efficient cytosolic delivery of therapeutic cargo that has been loaded into our HPLNs, additional modifications of these nanoparticles may be necessary, and is currently being explored. Future experiments may investigate previously described mechanisms for drug delivery into cells. One strategy is based on a pH-dependent fusogenic peptide (diINF-7) that resembles the N-terminal domain of influenza virus hemagglutinin HA-2 and allows membrane fusion and cytosolic access of endosome-entrapped molecules [69–71]. Another possible approach is the use of a pH-responsive fusogenic peptide, GALA. GALA is a 30 amino acid synthetic peptide that converts from a water-soluble random coil at pH 7.0 to an amphipathic, membrane binding α-helix at pH 5.0 [72]. GALA forms a transmembrane peptide pore comprised of about 10 GALA α-helical monomers arrayed perpendicular to the plane of the membrane, and has been shown to greatly enhance endosomal escape and gene silencing of siRNA-containing nanoparticles [73]. The diINF-7 or GALA fusogenic peptide can be incorporated into the HPLNs in the same way that the targeting peptides are incorporated. Once inside cells, release of the siRNA or ASO cargo is also critical to optimal efficacy. In addition to the fusogenic peptides, the polymer content in the HPLN shell can be optimized to promote opening and nucleotide release [58].