Promotion of CTL epitope presentation by a nanoparticle with environment-responsive stability and phagolysosomal escape capacity

Abstract

Vaccines that induce cytotoxic T lymphocyte (CTL)-mediated immune responses constitute an important class of medical tools to fend off diseases like infections and malignancy. Epitope peptides, as a format of CTL vaccines, are being tested preclinically and clinically. To elicit CTL responses, epitope vaccines go through an epitope presentation pathway in dendritic cells (DCs) that has multiple bottleneck steps and hence is inefficient. Here, we report the development of a strategy to overcome one of these barriers, phagolysosomal escape in DCs. First, we furnished a previously established carrier—an immune-tolerant elastin-like polypeptide nanoparticle (iTEP NP)—with the peptides that are derived from the DNA polymerase of herpes simplex virus 1 (Pol peptides). Pol peptides were reported to facilitate phagolysosomal escape. In this study, while we found that Pol peptides promoted the CTL epitope presentation; we also discovered Pol peptides disrupted the formation of the iTEP NP. Thus, we engineered a series of new iTEPs and identified several iTEPs that could accommodate Pol peptides and maintain their NP structure at the same time. We next optimized one of these NPs so that its stability is responsive to its redox environment. This environment-responsive NP further strengthened the CTL epitope presentation and CTL responses. Lastly, we revealed how this NP and Pol peptides utilized biological cues of phagolysosomes to realize phagolysosomal escape and epitope release. In summary, we developed iTEP NP carriers with a new phagolysosomal escape function. These carriers, with their priorly incorporated functions, resolve three bottleneck issues in the CTL epitope presentation pathway: vaccine uptake, phagolysosomal escape, and epitope release.

1. Introduction

Cytotoxic T lymphocyte (CTL)-mediated immunity is central to many types of immunotherapies for infectious diseases and cancer [1–4]. For example, CTL immunity is the ultimate tumor cell elimination mechanism of the immune checkpoint blockade therapy for cancer [5,6]. CTL immunity may be elicited by vaccines including CTL epitope vaccines. CTL epitopes are the peptides that are presented by professional antigen-presenting cells (APCs) and used to activate CD8+ T cells. CTL epitope vaccines have been developed preclinically and clinically as prophylactic medicine or therapeutics [1,4,7–9]. However, these developments are hindered by the suboptimal potency of the vaccines [8,10]. The deficient potency is attributed to the short half-life of the epitopes, which is due to their small size (8–9-mer peptides) and extracellular enzymatic cleavage [10]. Another factor that dampens the efficacy of epitope vaccines is their lack of intrinsic co-stimulants or adjuvants [8]. Using epitopes directly as vaccines also faces the challenge of uptake by APCs [11,12].

A wide range of strategies is being developed to address the deficiencies of CTL epitope vaccines. Using vaccine carriers to enhance epitope vaccines, with or without adjuvants, is one of such strategies [12–17]. These carriers normally go through a complex epitope cross-presentation pathway before their epitope payloads activate T cells [18,19]. The pathway starts with the internalization of carriers by APCs, followed in a chronological order by the intracellular trafficking of the carriers from phagolysosomes to the cytosol (termed phagolysosomal escape hereafter), the release of CTL epitopes from the carriers, the translocation of the epitopes from the cytosol to the ER, the binding of the epitopes to the MHC class I complexes (MHC-I) in the ER, the transportation and presentation of the epitope/MHC-I on the surfaces of APCs, and the ultimate engagement between the epitope/MHC-I complexes and T cell receptors of CD8+ T cells. There are, however, several barriers in this pathway including internalization, phagolysosomal escape, and epitope release, among others [19–21].

We previously developed nanoparticles (NPs) for CTL epitope delivery to address barriers in the epitope presentation pathway. These NPs are self-assembled from amphiphilic immune-tolerant elastin-like polypeptides (iTEPs) [11,22]. iTEPs are recombinant polypeptide materials that are biocompatible, immune tolerated, and easy to purify, thanks to their reversible phase transition property [22]. Moreover, the production of iTEPs is highly reproducible and scalable [23–27]. iTEP NPs protect CTL epitopes from enzymatic cleavage and promote their internalization by dendritic cells (DCs), a major type of APCs [11,22]. To leverage the above-mentioned advantages of iTEP NPs and facilitate epitope release from the NPs, we further devised iTEP NPs whose stabilities are regulated by their redox environments. These NPs are stable in the extracellular, non-reductive environment, so they are able to deliver vaccines as particles to DCs. Once they reach a reductive environment, such as cytosol, they become unstable and eventually dissolve, which promotes the release of CTL epitopes from their carriers [11].

Here we report a strategy to address another critical barrier of the CTL epitope presentation pathway, phagolysosomal escape. We screened literature for molecules that facilitate phagolysosomal escape and chose two peptides derived from the DNA polymerase of herpes simplex virus 1 (referred as Pol peptides) that met the prerequisite [28–30]. After we inserted these Pol peptides into an iTEP NP carrier of CTL epitopes, we discovered that while promoting the presentation of CTL epitopes, Pol peptides disrupted the formation of the NP. Then, we devised and compared a series of iTEPs and identified iTEPs that contain Pol peptides and are still able to self-assemble into NPs. Next, we optimized one of these NPs so that its stability is responsive to its redox environment. This environment-responsive NP further strengthened the presentation of CTL epitopes and enabled the epitopes to induce stronger T cell responses in vivo. Last, we leveraged this NP and investigated how Pol peptides mechanistically affected the epitope presentation pathway. Phagolysosomal acidification was found to be critical for Pol peptides to function. Putting these together, we successfully incorporated the phagolysosomal escape function into iTEP NP carriers. With this success, we are able to establish a CTL vaccine delivery platform that enhances CTL vaccines from at least three aspects: vaccine uptake, phagolysosomal escape, and epitope release. Yet, the multi-functional platform is based on a simple, homogeneous, and highly scalable polypeptide, iTEP.

2. Results

2.1. The impact of Pol peptides on the CTL epitope presentation and the stability of iTEP NPs

We chose two Pol peptides and tested their capacities to boost the CTL epitope presentation. The first one is AVGAGATAEE, termed as Pol_s. Pol_s was reported to promote the phagolysosomal escape of a protein with which Pol_s was connected [30]. The second one is EKLA-GFGAVGAGATAEE, termed as Pol_CD. Compared to Pol_s, Pol_CD has seven extra residues on its N-terminus, EKLAGFG, that constitute a putative cathepsin D (CD) cleavage site. We included both Pol_s and Pol_CD in this study because while both Pol peptides may facilitate phagolysosomal escape, Pol_CD may confer a further advantage. CD is a major aspartic protease in lysosomes [31–33]. The CD cleavage sequence in Pol_CD could promote epitope release from its carriers.

We inserted the two Pol peptides into a previously reported iTEP carrier, iTEP_B70-iTEP_A28-pOVA, through genetic engineering [22,34]. The two product carriers were iTEP_B70-iTEP_A28-Pol_s-pOVA and iTEP_B70-iTEP_A28-Pol_CD-pOVA (schematic in Fig. 1A). They were both amphiphilic, consisting of a hydrophilic iTEP segment (iTEP_B70)_, a hydrophobic iTEP segment (iTEP_A28), and a CTL epitope (pOVA). iTEP_A28 has 28 repeats of an iTEP building block, GVLPGVG. iTEP_B70 has 70 repeats of an iTEP building block, GAGVPG [22]. pOVA has the sequence, SIINFEKL, and is a hydrophobic peptide derived from ovalbumin [35,36]. The purity and sizes of both Pol-containing carriers were confirmed by SDS-PAGE (Fig. S1). Both Pol-containing carriers retained the thermally-induced, reversible phase transition property found in iTEPs (data not shown, [11,22]). However, neither of them was able to self-assembled into NPs based on dynamic light scattering (DLS) analyses. The mean hydrodynamic diameters of both Pol-containing carriers were less than 10 nm at all measured concentrations (Fig. 1B), suggesting they are soluble. This finding was unexpected given that the parent carrier of these two Pol-containing carriers, iTEP_B70-iTEP_A28-pOVA, formed NPs (Fig. 1B, and [11,22]), and Pol peptides were very small relative to the parent carrier (1.82/2.52 KDa versus 47.19 KDa). The insertion of the small Pol peptides was not expected to disrupt the NP structure.

Fig. 1.

The impact of Pol peptides on the carrier structure and epitope presentation. (A) Schematic showing the design of two Pol-containing iTEP carriers. The amino acid sequences of two carriers are provided in colored text. The amino acid sequence underlined is the cleavage site of cathepsin D. (B) Hydrodynamic diameters of three iTEP carriers at different concentrations. The measurements (number distributions) were obtained in PBS at 37 °C via DLS. (C) The epitope (pOVA) presentation with MHC-I by DC2.4 cells after the cells were incubated with different iTEP carriers (5 μM) or the DC culture medium. The data are presented as MFI means ± SD of DC cells of each carrier/treatment (N = 3). The MFI values were standardized to the MFI of the medium control treatment. The mean MFI of each treatment is provided. The graph represents data collected from three independent experiments. (D) The activation of B3Z T cells after they were incubated with DC2.4 cells. The DCs were pre-incubated with different iTEP carriers (1 μM) and the DC culture medium. The shown data are mean OD_570–635 ± SD of the samples (N = 3). The values of the means are provided. The graph represents data collected from three independent experiments.

We compared the presentation of pOVA by DCs in vitro after DC2.4 cells, a murine DC line, were incubated with these two Pol-containing carriers and their parent carrier. All carriers led to pOVA presentation in comparison to the medium control. Relative to the parent carrier, iTEP_B70-iTEP_A28-Pol_s-pOVA increased the presentation by 3.2 times (P = 0.0002, Fig. 1C); iTEP_B70-iTEP_A28-Pol_CD-pOVA increased the presentation by 2.1 times (P = 0.0077, Fig. 1C). The elevated presentation suggested that both Pol peptides are effective to facilitate the CTL epitope presentation. Interestingly, iTEP_B70-iTEP_A28-Pol_s-pOVA resulted in higher presentation than iTEP_B70-iTEP_A28-Pol_CD-pOVA (P = 0.0002, Fig. 1C), indicating that Pol_s exerts a stronger effect than Pol_CD.

Lastly, we examined whether the above presentation could translate into T cell activation. Here, we used B3Z cells, a CD8+ T cell line restricted to the pOVA/MHC-I (H-2K^b), for a T cell activation assay. All carriers led to stronger B3Z cell activation in comparison to the medium control. The mean value of the B3Z cell activation was 0.67 (an arbitrary value defined by the assay) after the B3Z cells were incubated with DCs that were pretreated with the parent carrier, iTEP_B70-iTEP_A28-pOVA. The mean value resulting from the Pol_s-containing carrier was 1.45; the mean value from the Pol_CD-containing carrier was 0.99. In comparison to the parent carrier, Pol_s and Pol_CD enhanced the B3Z cell activation by 120% and 50%, respectively (P < 0.0001 and P = 0.0023, Fig. 1D). Similar to the results of DC presentation, the effect of Pol_s is stronger than Pol_CD (P = 0.0004, Fig. 1D).

Putting these results together, we found that both Pol peptides increased CTL epitope presentation although they disrupted the NP self-assembly of an iTEP carrier. We also found that Pol_CD did not bestow additional benefits to the CTL epitope presentation as compared to Pol_s. Indeed, the epitope presentation and the T cell activation resulting from the Pol_CD vaccine were significantly weaker. We, therefore, chose Pol_s-containing carriers in the subsequent studies.

2.2. Relationship between the hydrophobic segment of iTEP and the stability of iTEP NPs

Since NP-based CTL vaccines were more potent than soluble CTL vaccines [11], and Pol_s augmented CTL epitope presentation, we strived to obtain an iTEP carrier design that would maintain an NP structure after accommodating Pol_s. The carrier, iTEP_B70-iTEP_A28-Pol_s-pOVA, did not have the design; however, it could serve a starting point for new designs. We modulated the length of the hydrophobic iTEP_A segment of this carrier because the length is critical to the stability of NP self-assembled from amphiphilic molecules [22]. We gradually increased the repeat numbers of iTEP_A and generated a series of iTEP carriers, iTEP_B70-iTEP_A(N)-Pol_s-pOVA, that encompassed seven different iTEP_A repeats, (N = 32, 36, 40, 44, 48, 52, and 56). The purity and sizes of the carriers were confirmed by SDS-PAGE (Fig. S2). The carriers that contained iTEP_A32, iTEP_A36, and iTEP_A40 only self-assembled into NPs when their concentrations were equal or higher than 100, 50, and 25 μM, respectively (Table 1 and Table S2). NPs formed under these conditions had average hydrodynamic diameters ranging from 25 to 35 nm. In contrast, the carriers that contained iTEP_A44, iTEP_A48, iTEP_A52, and iTEP_A56 formed and maintained their NP structures at concentrations as low as 5 μM. The NPs of these carriers had average hydrodynamic diameters ranging from 29 to 47 nm. These data suggested that the stability of these iTEP NPs is dependent on the length of the iTEP_A segment. Consistent with this notion, an iTEP vaccine without any iTEP_A segment, iTEP_B70-Pol_s-pOVA, did not assume an NP structure (its measured diameter was less than 10 nm, Table 1).

Table 1.

Sizes of iTEP carriers with different iTEP_A. The hydrodynamic diameters of these carriers are shown (nm).

Concentration (μM)	5	25	50	100
Carriers
iTEPB70-iTEPA28-Pols-pOVA	8.8 ± 1.3	10.1 ± 1.7	10.4 ± 2.1	11.0 ± 2.1
iTEPB70-iTEPA32-Pols-pOVA	9.7 ± 2.0	9.9 ± 2.2	11.1 ± 2.4	25.6 ± 7.2
iTEPB70-iTEPA36-Pols-pOVA	8.8 ± 2.2	10.1 ± 2.2	30.3 ± 7.6	34.3 ± 8.6
iTEPB70-iTEPA40-Pols-pOVA	10.5 ± 1.3	28.3 ± 7.8	30.7 ± 8.3	35.0 ± 8.8
iTEPB70-iTEPA44-Pols-pOVA	39.2 ± 9.6	29.1 ± 8.3	32.6 ± 8.5	36.9 ± 8.5
iTEPB70-iTEPA48-Pols-pOVA	41.5 ± 9.7	42.1 ± 9.3	41.4 ± 9.5	41.1 ± 9.0
iTEPB70-iTEPA52-Pols-pOVA	39.8 ± 10.4	44.1 ± 11.1	41.2 ± 10.3	40.5 ± 11.0
iTEPB70-iTEPA56-Pols-pOVA	44.1 ± 11.6	47.1 ± 10.4	45.0 ± 9.8	46.8 ± 10.3
iTEPB70-Pols-pOVA	8.2 ± 1.8	8.4 ± 1.8	8.7 ± 1.8	7.4 ± 1.7

Numbers in italic fonts indicate that carriers with these hydrodynamic diameters are not believed to be NPs.

We also compared the pOVA presentation by DC2.4 cells after the cells were incubated with these iTEP carriers. We found that there was a trend of an inverse correlation between the presentation level and the lengths of the iTEP_A segment of these carriers (Fig. 2, Table 2). For example, the presentation that resulted from the carrier with iTEP_A28 was 5 times higher than that of the carriers with iTEP_A52 and iTEP_A56 (P < 0.0001). These results implied there is an inverse correlation between the epitope presentation and the NP stability of the carriers when the presentation is measured in vitro.

Fig. 2.

The pOVA presentation of iTEP carriers having different iTEP_A. The cells were incubated with different iTEP carriers (5 μM) or the DC culture medium. The data are presented as MFI means ± SD of DC cells of each carrier/treatment (N = 3). The MFI values were standardized to the MFI of the medium control treatment. The mean MFI of each treatment is provided. The graph represents data collected from three independent experiments.

Table 2.

The relationship between the NP formation, the pOVA presentation, and the lengths of hydrophobic block of iTEP carriers.

Carriers	Repeat number of iTEPA	pOVA presentation (mean MFI ± SD)	The lowest concentration* to form NPs
iTEPB70-iTEPA28-Pols-pOVA	28	7.67 ± 0.05	NA
iTEPB70-iTEPA32-Pols-pOVA	32	6.61 ± 0.02	100 μM
iTEPB70-iTEPA36-Pols-pOVA	36	5.76 ± 0.02	50 μM
iTEPB70-iTEPA40-Pols-pOVA	40	4.09 ± 0.09	25 μM
iTEPB70-iTEPA44-Pols-pOVA	44	2.88 ± 0.11	5 μM
iTEPB70-iTEPA48-Pols-pOVA	48	1.78 ± 0.12	5 μM
iTEPB70-iTEPA52-Pols-pOVA	52	1.36 ± 0.15	5 μM
iTEPB70-iTEPA56-Pols-pOVA	56	1.47 ± 0.02	5 μM

^*Concentrations used in this study include 5 μM, 25 μM, 50 μM, and 100 μM.

2.3. A Pol_s-containing NP vaccine that has a redox environment-responsive stability

Since Pol-containing carriers that formed stable NPs did not produce a high level of CTL epitope presentation, it became a dilemma whether or not a stable Pol-containing NP is desired as a carrier of CTL epitopes because, on the other hand, NPs was proven advantageous as CTL epitope carriers according to our and others’ previous in vivo studies [11,16,22]. We reasoned that the compromised, in vitro epitope presentation by stable iTEP NPs might be due to the inefficient release of epitopes from the NPs, which is a prerequisite for epitope presentation. For epitopes to be released from iTEP NPs, the NPs need to dissociate into soluble iTEPs, and the epitopes need to be cleaved enzymatically from the soluble iTEP molecules. To resolve the dilemma and meet the seemingly competing needs for a stable and an unstable NP carrier for CTL epitopes, we utilized our previously devised iTEP NPs that manifested different stability in different environments [11]. These NPs, stabilized by disulfide bonds, remain a particle structure in non-reductive environments. However, they lose their particle structure in reductive environments. Based on the design of these previous NPs, we generated two Pol-containing iTEP carriers, iTEP_B70-iTEP_A28-(G₈C)₄-Pol_s-pOVA and iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA. These Pol-containing carriers have a spacer, (G₈C)₄, between iTEP segments and Pol_s. The cysteines on the spacers were included to form inter-iTEP disulfide bonds after iTEPs self-assembled into NPs (schematic in Fig. 3A). We found that iTEP_B70-iTEP_A28-(G₈C)₄-Pol_s-pOVA was not able to assemble into NPs until reaching a concentration of 100 μM (Tables S3 and S4). In contrast, iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA were able to form NPs at all measured concentrations from 5 to 100 μM (Fig. 3B, Tables S3 and S4). The NP structure of iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA was confirmed by TEM having a mean diameter of 24.0 nm ± 6.3 nm (N = 107) (Fig. 3C). We chose iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA for the following studies. This NP, when its disulfide bonds are reduced, has marginal stability.

Fig. 3.

A Pol_s-containing NP that has a redox environment-responsive stability. (A) A schematic diagram showing the assembly of a Pol_s-containing NP that is stabilized by disulfide bonds. The sequence design of the pol-containing carrier that self-assembles in the NP is also illustrated. (B) Hydrodynamic diameters of iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA at various concentrations. The measurements (number distributions) were obtained in PBS at 37 °C via DLS. (C) A representative TEM micrograph of NPs formed by iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA. Three particles are pointed out by arrows. (Scale bar, 100 nm) (D) Hydrodynamic diameters of iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA (5 μM) after it was treated with 1 μM or 1 mM of GSH at 37 °C for overnight. The measurements (number distributions) were obtained via DLS. (E) The reductive status of four iTEP carriers after they were treated overnight with or without 1 mM GSH. A non-reducing SDS-PAGE was used to evaluate the status. 20 μg of each sample was used for the evaluation.

We examined whether the stability of the iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA NP could be diminished in a reductive environment. Here, the NP was pre-formed from iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA, then the disulfide bonds inside the NP were established through oxidization. Further, the NP was incubated with different concentrations of GSH. The NP maintained its structure at 1 μM GSH, a redox environment mimicking extracellular environments in the body (GSH: 1–10 μM) [37–39]. The NP dissociated at 1 mM GSH, a redox environment mimicking cytoplasmic environments (GSH: 1–10 mM) [37,40] (Fig. 3D). As a control, the NP assembled from iTEP_B70-iTEP_A56-Pol_s-pOVA did not respond to the change of environmental GSH concentrations (Fig. 3D). This control NP does not rely on disulfide bonds to maintain its structure. Results of SDS-PAGE confirmed that the disulfide bond-based crosslinking between iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA was diminished by 1 mM GSH. The amount of the protein in its reduced form increased when 1 mM GSH was added into the protein sample (Fig. 3E).

We also compared the epitope presentation in vitro among three vaccine carriers: iTEP_B70-iTEP_A36-pOVA, iTEP_B70-iTEP_A36-Pol_s-pOVA, and iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA. All carriers resulted in pOVA presentation in comparison to the medium control. We also found that the two carriers that contained Pol_s led to at least a 2-times greater presentation than the Pol_s-free carrier (P = 0.0002 and P = 0.0020, Fig. 4A). Between the two Pol_s-containing carriers, the NP carrier having changeable stability, iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA, led to slightly lower presentation than the unstable NP carrier, iTEP_B70-iTEP_A36-Pol_s-pOVA (Fig. 4A). The difference implies that while the vast majority of NPs with changeable stability might have dissolved inside DCs due to a reductive intracellular environment and pOVA was released from the disassociated NPs, the dissolution was not complete. Similar to the observations of the epitope presentation, the DCs pretreated with the two Pol_s-containing carriers also more effectively activated B3Z T cells than the DCs pretreated with the Pol_s-free carrier (Fig. 4B). Similarly, all carriers led to greater B3Z cell activation than the medium control. Interestingly, results from further investigations revealed that the location of Pol_s on iTEP carriers was an important factor of epitope presentation. When Pol_s was not adjacent to pOVA [Pol_s-iTEP_B70-iTEP_A36-(G₈C)₄-pOVA], the effect of Pol_s on the pOVA presentation decreased (P = 0.0005, Fig. 4B), even though the Pol_s position did not affect the NP formation and redox response of the iTEP carriers (Fig. 3D and S4, Tables S3 and S4).

Fig. 4.

CTL epitope presentation by a Pol_s-containing NP that is redox environment-responsive. (A) The epitope (pOVA) presentation with MHC-I by DC2.4 cells after the cells were incubated with iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA (5 μM) or its controls. The data are presented as MFI means ± SD of DC cells of each carrier/treatment (N = 3). The MFI values were standardized to the MFI of the medium control treatment. The mean MFI of each treatment is provided. The graph represents data collected from three independent experiments. (B) The activation of B3Z T cells after they were incubated with DC2.4 cells. The DCs were pre-incubated with iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA (1 μM) or its controls. The shown data are mean OD_570–635 ± SD of the samples (N = 3). The values of the means are provided. The graph represents data collected from three independent experiments. (C) Ex vivo analysis of active, pOVA-restrictive splenocytes from mice (N = 5) immunized with iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA, iTEP_B70-iTEP_A36-(G₈C)₄-pOVA, and pOVA. The activation of the cells was evaluated by an IFNλ-based ELISPOT assay. Data are presented as Spot Forming Units (SFU)/million cells ± SD.

Last, we examined whether the boosted in vitro epitope presentation and T cell activation by Pol_s could translate into elevated CTL responses in vivo. We vaccinated C57BL/6 mice with pOVA and two NPs self-assembled from iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA and iTEP_B70-iTEP_A36-(G₈C)₄-pOVA, respectively. pOVA-restricted T cell responses were quantified by using splenocytes from these immunized mice and an IFN-γ-based ELISPOT assay. Both NP vaccines induced stronger T cell responses than pOVA (averagely 230 and 141 spots versus 78 spots per million splenocytes, P = 0.0367 and 0.0006). Moreover, the NP containing Pol_s elicited greater immune responses than the Pol_s-free NP (230 versus 141, P = 0.03) (Fig. 4C).

2.4. Mechanistic exploration of the effect of Pol_s

Since Pol_s peptides, as a component of CTL epitope carriers, improved the epitope presentation, we went further to explore how the improvement happened mechanistically.

First, we examined how phagolysosomal acidification would affect the function of Pol_s-containing carriers using the aforementioned B3Z cell activation assay. Specifically, we treated DCs with BafA1, an inhibitor of phagolysosomal acidification [41], before incubating the DCs with a Pol_s-containing carrier, iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA. Then, we incubated the DCs with B3Z T cells and quantified the activation of B3Z cells. If the inhibition of phagolysosomal acidification affects the epitope presentation resulting from this carrier, a reduction of B3Z cell activation should be detected. To evaluate the reduction, we used DCs with normal phagolysosomal acidification to perform a parallel experiment and obtained a baseline of B3Z cell activation. Compared to control DCs, the BafA1 treatment (200 nM) lowered the B3Z T cell activation resulting from iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA by 3.43-fold (P = 0.0002, Fig. 5A). We also performed the same study with a similar carrier but without Pol_s, iTEP_B70-iTEP_A36-(G₈C)₄-pOVA. We found the BafA1 treatment lowered the B3Z T cell activation by 2.5-fold (P = 0.0025). These data suggest that phagolysosomal acidification affects the function of both carriers, though the carrier with Pol_s is more dependent on the acidification. It was noteworthy that the effect of BafA was dose-dependent (Fig. 5A). When another inhibitor of phagolysosomal acidification, chloroquine [42], was used for the experiment of the same design, similar observations were obtained. Chloroquine-treated DCs showed a compromised ability to activate B3Z cells inhibition: the activation was reduced by 2.4-fold in the case of iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA (P < 0.0058; Fig. 5B); the activation was reduced by 2.1-fold in the case of iTEP_B70-iTEP_A36-(G₈C)₄-pOVA (P < 0.001). Together, these data suggested that phagolysosomal acidification is important for the function of the two carriers tested and that the dampened acidification negatively impacts the function of Pol peptides.

Fig. 5.

Working mechanisms of Pol_s in iTEP NPs. (A and B) Results of the B3Z T cell activation after the T cells were incubated with DC2.4 cells. The DCs were pretreated with BafA 1 (A) and chloroquine (B) for one hour before they were incubated with iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA (1 μM), iTEP_B70-iTEP_A36-(G₈C)₄-pOVA (1 μM), and the DC culture medium. The shown data are mean OD_570–635 ± SD of the samples (N = 3). The values of the means are provided. The graph represents data collected from three independent experiments. (C) Results of the B3Z T cell activation after the T cells were incubated with DC2.4 cells. The DCs were pretreated with 1,10-phenanthroline before they were incubated with iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA (1 μM), iTEP_B70-iTEP_A36-(G₈C)₄-pOVA (1 μM), and the DC culture medium. The shown data are mean OD_570–635 ± SD of the samples (N = 3). The values of the means are provided. The graph represents data collected from three independent experiments.

We also found that the metallopeptidase activity of DCs supported the function of the above two carriers, iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA and iTEP_B70-iTEP_A36-(G₈C)₄-pOVA. In this study, DCs were first treated by an inhibitor of metallopeptidases, 1,10-phenanthroline, then the cells were incubated with the two carriers and used to activate B3Z cells. The inhibitor reduced the T cell activation to a similar extent in both carriers (Fig. 5C). These data suggested the support of metallopeptidases to the Pol-containing carriers is independent of Pol peptides.

3. Discussion

We developed a CTL epitope vaccine carrier that promotes phagolysosomal escape and possesses an NP structure sensitive to its redox environment. This carrier provides an idea and a model to overcome at least three challenges of CTL epitope vaccines: limited uptake, inefficient phagolysosomal escape, and deficient epitope release [19,20]. Through the carrier development process, we gained insights about how to address these challenges organically. We will recount and discuss these insights in the following paragraphs.

Pol peptides enhance epitope presentation after they are incorporated into epitope carriers; such function of Pol peptides is influenced by carrier structures. We performed paired-comparisons of the in vitro epitope presentation within the three pairs of Pol-containing and Pol-free carriers: iTEP_B70-iTEP_A28-Pol_s-pOVA vs iTEP_B70-iTEP_A28-pOVA, iTEP_B70-iTEP_A28-Pol_CD-pOVA vs iTEP_B70-iTEP_A28-pOVA, as well as iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA vs iTEP_B70-iTEP_A36-(G₈C)₄-pOVA. According to the results of these comparisons, Pol peptides doubled or even tripled the presentation. Further, the Pol-containing carrier, iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA induced stronger CTL responses in vivo than its Pol-free control. It is evident that Pol peptides benefit CTL epitope presentation as one component of the epitope carriers. Based on our design of Pol-containing NPs, Pol peptides are located in the hydrophobic core of the NPs along with CTL epitopes, which protects the epitopes and Pol peptides during extracellular transportation of the NPs. Our observations suggest that Pol peptides need to be exposed from the NPs to function: first, stable Pol-containing NPs are less effective for epitope presentation than those unstable Pol-containing NPs that tend to dissociate (Table 2); second, a Pol-containing NP that dissolves in a reductive environment leads to greater epitope presentation than a Pol-containing NP that is stable in the same environment. Consistent with these observations, it was reported that Pol peptides need to interact with phagolysosomal membranes to exert their functions [43].

There are apparently competing needs on the stability of epitope carrier NPs. NPs with a fluid stability could meet all these competing needs. Stable NP carriers are needed to keep CTL epitopes inside them so that the epitopes will have less enzymatic degradation, longer half-lives, and greater cellular uptake [11,12,22]. However, it is equally important for NPs to dissociate at certain points of the epitope presentation pathway so that epitopes are exposed to enzymatic cleavage and subsequently released for their presentation. Further, Pol peptides in Pol-containing NPs need to be exposed to phagolysosomal membranes to disrupt the membranes. The positive side is that these needs, though competing, do not need to be served simultaneously. Indeed, stable NPs are only required before they reach the phagolysosomes, and after that point, they are preferred to dissociate. In brief, NPs need single-directional dynamic stability. Given this trend, we designed NPs that are self-assembled from amphiphilic iTEPs, e.g. iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA, and mainly stabilized by inter-iTEP disulfide bonds. These redox environment-responsive NPs are stable in non-reductive environments such as extracellular ones; and they are not stable in reductive environments such as cytoplasm that have high concentrations of GSH (> 1 mM). When these NPs reach to the phagolysosomes, they could also be disassembled. These organelles have an accumulation of reductases, such as gamma-interferon-inducible lysosomal thiol reductase (GILT) [44] that can reduce the disulfide bonds of these NPs and initiate their dissembling. The in vitro and in vivo data of this current study prove the success of environment-responsive NPs.

Phagolysosomal acidification is a prerequisite for the success of iTEP NPs in this study. The acidification affects dissociation, phagolysosomal escape, as well as epitope release of NPs. Both of the two tested acidification inhibitors, BafA and chloroquine, lowered epitope presentation in a dose-dependent manner. We reason there are at least two mechanistic bases for the outcome. First, phagolysosomal acidification is required for the dissociation of NPs in phagolysosomes. GILT in these organelles reduces disulfide bonds [45,46] and disassemble environment-responsive NPs. GILT was also reported to facilitate the phagolysosomal escape of antigens [47]. GILT needs an acidic environment (pH 4.5–5.5) to function properly [44]. Therefore, inhibition of phagolysosomal acidification would jeopardize the function of GILT and hence hinder NP dissociation and the subsequent phagolysosomal escape and epitope release. Secondly, an acidic environment is required for Pol peptides to function properly. Lysosomal pH (pH 4.0) is required to neutralize the charges of the two glutamate residues in Pol peptides in phagolysosomes so that the residues and their neighboring residues can participate in an α-helix of Pol peptides. The α-helix is critical for Pol peptides to disrupt phagolysosomal membranes [29,43]. Put together, phagolysosomal acidification is implicated in both the reduction of disulfide bonds and the membrane disruption, the two main functions of the iTEP NPs in this study.

The discoveries of this study combined with previous conclusions of iTEP NPs [11,22] show a mechanism on how self-assembled iTEP NPs with fluid stability and Pol peptides promote epitope presentation. After the NPs enter the phagolysosomes of APCs, the phagolysosomes containing the NPs start acidification. Next, the thiol reductases like GILT in the phagolysosomes reduce the disulfide bonds of the NPs, which transforms the NPs into an unstable form and dissociates them into single, soluble iTEP molecules. Consequently, Pol peptides are exposed and destabilize phagolysosomal membranes. Eventually, the content of the phagolysosomes including soluble iTEP molecules translocates into the cytosol. In the cytosol, CTL epitopes are released from soluble iTEP molecules and passed along to follow the steps of the epitope presentation pathway (Fig. 6). It is acknowledgeable that not all NPs dissociate in the phagolysosomes. Some may translocate to the cytosol and dissolve there, a reductive environment; others may never be dissociated. Indeed, the NPs with fluid stability did not achieve the same high level of in vitro epitope presentation as did the intrinsically unstable iTEP NPs, indicating that some NPs are never dissociated nor do they produce any free CTL epitopes.

Fig. 6.

A schematic diagram on how Pol-containing iTEP NPs function inside APCs and benefit the CTL epitope presentation. The NPs, after uptake by APCs, are transported to phagolysosomes. Then, some NPs have their disulfide bonds (green circles) reduced, and the reduced NPs dissociate into single-molecule iTEP carriers. Pol peptides on these single molecules disrupt and create pores on phagolysosomal membranes. Both single iTEP carriers and iTEP NPs escape to the cytosol through these pores. Inside the cytosol, iTEP NPs dissociate under a reductive environment. Further, single-molecule iTEP carriers are processed by proteasomes in the cytosol and produce free CTL epitopes. These epitopes are eventually presented together with MHC-I on the surface of APCs. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

In summary, we developed an iTEP NP that realized environment-responsive stability and phagolysosomal escape, two capacities important to CTL epitope carriers. The NP also benefits the cellular uptake of CTL epitopes. It is noteworthy that this multifunctional NP is self-assembled from multiple copies of a simple and monodispersed iTEP fusion protein [22]. This composition of the protein can be precisely and easily controlled through genetic engineering. The protein nature of iTEPs also lowers the translational and safety barrier of iTEP fusion proteins and iTEP NPs. iTEP fusion proteins have been used to deliver adjuvant [25], target epitopes to lymphatic organs [48], and directly load epitopes onto the MHC-I on DCs [26,27]. It is appealing to use this simple platform to meet the complex needs of CTL epitope vaccines, which may facilitate clinical translation of the preclinical successes of an iTEP-based CTL vaccine delivery system.

4. Materials and methods

4.1. Mice and cell lines

Female, 6 to 8-week old, C57BL/6 mice were purchased from Charles River Laboratories and used in the in vivo CTL vaccination study. The study follows an approved protocol by the Institutional Animal Care and Use Committee (IACUC) at the University of Utah.

The DC2.4 and B3Z cell lines were generous gifts from Dr. Kenneth Rock (University of Massachusetts, USA) and Dr. Nilabh Shastri (Johns Hopkins University), respectively. The DC 2.4 cells were maintained in RPMI-1640 medium supplemented with 10% inactivated fetal calf serum, 1% non-essential amino acids, 2 mM glutamine, 1% Hepes, 50 μM β-Mercaptoethanol, 100 units/mL penicillin, and 100 μg/mL streptomycin (ThermoFisher Scientific, USA). The B3Z cells were maintained in a medium similar to the medium for DC 2.4 cells, except the non-essential amino acids and Hepes were substituted with 1 mM sodium pyruvate.

4.2. Construction of the expression plasmids of iTEP carriers

The genes encoding all iTEP carriers were synthesized on a modified pET25b(+) vector using a previously described Pre-RDL method [22,34]. First, genes that encoded GVLPGVG (the building block of iTEP_A), GAGVPG (the building block of iTEP_B), SIINFEKL (pOVA), AVGAGATAEESIINFEKL (Pol_s-pOVA), EKLAGFGAVGAGATAEESIINFEKL (Pol_CD-pOVA), and (G₈C)₄ were generated by annealing the sense and antisense oligonucleotides of these genes (Eurofins Genomics. USA). The sequences of these oligonucleotides are listed in Supplementary Table S1. Next, the annealing products were inserted into the modified pET25b(+) vector at the BseR I site. Then, the genes for iTEP_A and iTEP_B with desirable lengths were extended from the genes of their building blocks by the Pre-RDL method on the vector. Genes of other peptides were later fused with genes of iTEPs at the defined position using the Pre-RDL method. Ultimately, the resulting expression pET25b (+) vectors were transformed into DH5α for amplification. The sizes of coding genes were confirmed by an Xba I and BamH I double digestion in combination with an agarose gel analysis. The sequences of coding genes were verified by DNA sequencing (Genewiz, USA).

4.3. Production and purification of iTEP carriers

All iTEP carrier proteins were produced from BL21 cells and purified as previously described [11,22]. One particular modification for the purification of iTEP carrier proteins containing the (G₈C)₄ sequence is that the proteins were purified in PBS with 10 mM TCEP-HCl (pH 7.0) to reduce disulfide bonds. The size and purity of all the proteins were assessed by SDS-PAGE (Figs. S1, S2, and S3). The endotoxin was removed by 1% Triton X-114 3 times as previously described. Then Triton X-114 was removed using Amicon Ultra-15 (10 kDa) centrifugal filters (Millipore, USA). The residual endotoxin in the samples was determined by Limulus Amebocyte Lysate (LAL) PYROGENT Single Test Vials (Lonza, Allendale, NJ, USA). All samples used for in vitro and in vivo immune assays had their endotoxin level below 0.25 EU per mg protein.

4.4. Particle size characterization of iTEP carriers

The size distributions of iTEP carriers were determined by dynamic light scattering (DLS) using a Zetasizer Nano-ZS instrument (Malvern Instruments, Malvern, UK) as previously described [11]. For iTEP carrier samples without cysteine residues, we resuspended them at various concentrations in PBS and keep them at room temperature overnight before conducting measurements. For iTEP carrier samples with cysteine residues, we first fully reduced the disulfide bonds within them. Then, we set their concentrations to 100 μM and confirmed that the samples formed NPs. Next, we oxidized these samples with 0.3% H₂O₂ at 37 °C for 15 min and generated disulfide bonds to stabilize the NPs. Last, we removed the H₂O₂ from samples through centrifugation with Amicon centrifugal filter devices (Millipore, USA). To test the effect of GSH on iTEP NP formation, 5 μM oxidized iTEP carriers in PBS were treated overnight with different concentrations of GSH at 37 °C before measurement. The reported results represented the average particle size by number. Other parameters including Z-average, size by intensity, and size by volume were also measured and reported in Supplementary Table S2 and S4.

4.5. Negative-stain, transmission electron microscopy of iTEP carriers

First, continuous carbon support film grids (Ted Pella) were glow discharged (PELCO easiGlow, 15 mA) for 45 s. Next, iTEP_B70-iTEP_A36-(G₈C)₄-Pol_s-pOVA particles at a concentration of 50 μM in 1XPBS buffer were then applied to the grids and incubated for 60 s. Then, the grids were then washed with distilled water and stained with 1% aqueous Uranyl Acetate (Electron Microscopy Solutions). Last, the particle samples were imaged at the RTSF Cryo-EM Core Facility at Michigan State University using a Talos Arctica operated at 200 keV. Micrographs were collected on a Ceta camera at a nominal magnification of 22,000 (4.71 Å/pixel) with an exposure time of 4 s and an objective lens defocus setting as of 5 μM underfocus.

4.6. In vitro pOVA presentation by dendritic cells

DC2.4 cells were plated at the density of 2.5 × 10⁵ /500 μL/well in 24-well plates and cultured with 5 μM iTEP carriers for 16 h at 37 °C with 5% CO₂. Then, the cells were collected and washed with PBS. The complex of pOVA/MHC-I (H-2K^b) on DC cell surface was labeled with PE-tagged monoclonal antibody 25-D1.16 (Biolegend, USA, 1:100 dilution). The bound antibodies on the cell surface were measured with flow cytometry (collecting 5 × 10⁴ events per sample). The data are presented as the mean fluorescence intensity (MFI) of carrier-treated DCs relative to fluorescence signals of untreated DCs based on flow cytometry data.

4.7. Activation of B3Z hybridoma (CD8+ T) cells

B3Z is a CD8+ T-cell hybridoma line engineered to secrete β-galactosidase when its T-cell receptors are engaged with pOVA:MHC-I (H2K^b) complexes [49]. This assay was done by a previously published protocol with modifications [22]. Briefly, DC2.4 cells were set in 96-well plates at the density of 1 × 10⁵ cells/well. iTEP carriers at indicated concentrations were added into the DC culture and incubated for 4 h before being washed. The DCs were then fixed with 1% of paraformaldehyde for 15 min at room temperature and then extensively washed 5 times with culture media for B3Z cells. After the wash, 1 × 10⁵ B3Z cells/well were added to the DC culture and co-cultured with DCs for 24 h. After the incubation, all cells were washed and lysed with 100 μL of lysis buffer (PBS with 100 mM 2-mercaptoethanol, 9 mM MgCl2, and 0.125% NP-40). Meanwhile, chlorophenol red β-galactoside substrate (Sigma, St. Louis, MO, USA) was added into the lysate at 0.15 mM to start a colorimetric reaction. The reaction was stopped with 50 μL of 15 mM EDTA and 300 mM glycine after 4 h at 37 °C. The OD₅₇₀ of reaction samples was measured using an Infinite M1000 PRO plate reader (Tecan Trading AG, Switzerland), and OD₆₃₅ was used as a reference. In mechanistic studies, BafA 1, chloroquine, and 1,10-phenanthroline at various concentrations were added to the DC culture 1 h before the addition of iTEP carriers.

4.8. Immunization and splenocyte isolation from mice

Mice were randomized into 3 groups for treatments with three different types of iTEP carriers and immunized subcutaneously with 2 nmol iTEP carriers twice, on day 0 and day 7. The mice were sacrificed on day 17 and the spleens were harvested. Single splenocytes were generated and counted using a Countess™ Automated Cell Counter (ThermoFisher Scientific, USA).

4.9. IFN-γ-based Enzyme-linked immunospot (ELISPOT) assay

The assay method was described previously [22]. Briefly, splenocytes collected from immunized mice were reactivated by pOVA (2.5 mg/mL) for 48 h before being loaded into wells of 96-well filtration plates (Millipore, Billerica, MA, USA). The wells were precoated with 5 mg/mL of capture anti-mouse IFN-γ mAb (Clone: R4–6A2, Biolegend, San Diego, CA, USA). Triplicates were set up for each condition. Cells were discarded after 24 h, and the wells were incubated overnight with 2 mg/mL of biotinylated detection anti-mouse IFN-γ mAb (Clone: XMG1.2-Biotin, Biolegend, San Diego, CA, USA). After washing, the bound anti-mouse IFN-γ mAb was detected using horseradish peroxidase (HRP Avidin, Biolegend, San Diego, CA, USA) together with a 3-amino-9-ethyl-carbazole (AEC) substrate (Sigma, St. Louis, MO, USA). The bottoms of the wells were scanned, and the spots on the bottom of the well were automatically counted using ImageJ software.

4.10. Statistical analysis

Data were analyzed for statistical significance using an unpaired Student’s t-test. GraphPad Prism 6 (GraphPad Software, San Diego, CA, USA) was used for statistical analysis and figure construction.