Carrier-Protein-Free Pneumococcal Glycoconjugate Vaccines Enabled by SPAAC: Serotype 15C CPS–PADRE Conjugates and the Impact of an RR Cleavage Motif

TL;DR

Researchers tested a new vaccine design that avoids using carrier proteins by linking a pneumococcal polysaccharide to a helper epitope called PADRE, finding that it can work but with some limitations.

Contribution

A novel carrier-protein-free vaccine design using PADRE as a helper epitope is proposed and tested for pneumococcal conjugate vaccines.

Findings

CPS–PADRE conjugates elicited lower antibody levels compared to traditional CPS–protein conjugates in mice.

Incorporating a cleavable RR motif reduced humoral responses compared to non-cleavable designs.

The study shows carrier-protein-free vaccines are feasible but require careful optimization.

Abstract

Background/Objectives: Polysaccharide-protein conjugate vaccines have proven highly effective, yet they remain limited by manufacturing complexity, cost, and variable performance across serotypes, while carrier proteins can add unwanted immunological and production burdens. To address these constraints, we explored a carrier-protein-free conjugate vaccine concept in which a broadly MHC class II-binding helper epitope (PADRE) replaces the conventional protein carrier to provide T-cell help for a pneumococcal capsular polysaccharide antigen. Methods: Using serotype 15C CPS as a model, we generated CPS–PADRE conjugates and compared designs with or without a putative cleavable motif (RR) at the junction, alongside a conventional protein conjugate as a benchmark. Results: In mice, the CPS–protein conjugate induced the strongest CPS-specific IgG response, whereas CPS–PADRE conjugates elicited clear but overall lower antibody levels. Notably, incorporation of the cleavable motif did not improve immunogenicity and instead reduced humoral responses relative to the non-cleavable design. Conclusion: These findings support the feasibility of carrier-protein-free polysaccharide-peptide conjugate vaccines, while highlighting that cleavable junctions are not universally advantageous and must be empirically optimized for polysaccharide-helper epitope architectures.

Figures

Full text

1. Introduction

Streptococcus pneumoniae remains a leading bacterial cause of pneumonia, sepsis, and meningitis worldwide, with the greatest burden in young children, older adults, and individuals with underlying comorbidities [1,2]. The capsular polysaccharide (CPS) is the major virulence determinant of pneumococci and the principal antigenic target of licensed vaccines. Although polysaccharide-only vaccines such as PPSV23 are effective in adults, they are T cell-independent, provide limited immunological memory, and perform sub optimally in infants/young children and certain high-risk populations. Polysaccharide-protein conjugate vaccines (PCVs) have been exceptionally successful because covalent coupling converts CPS antigens into T cell-dependent immunogens, enabling class switching, affinity maturation, and immunological memory [3]. Consistent with this immunological rationale, current Advisory Committee on Immunization Practices (ACIP) and Centers for Disease Control and Prevention (CDC) guidance increasingly centers pneumococcal vaccination on conjugate vaccines. PCVs (PCV15 or PCV20) are routinely recommended for all children <5 years, and ACIP recommends that PCV-naïve adults aged ≥50 years receive a pneumococcal conjugate vaccine (PCV15, PCV20, or PCV21), with PPSV23 used primarily as a follow-on dose when PCV15 is selected.

However, despite this clinical impact, the current PCV paradigm continues to face significant scientific and translational constraints that motivate alternative conjugate designs. A central limitation is manufacturing complexity. Conventional PCVs are produced through multistep, serotype-specific workflows that typically include polysaccharide activation/derivatization, conjugation to a carrier protein, removal of unconjugated components, and extensive characterization to control heterogeneity and ensure consistency [4,5]. These steps can be difficult to standardize because CPS structures differ widely in charge, stability, and chemical reactivity, and excessive derivatization risks damaging protective epitopes. The complexity increases further for multivalent formulations, in which each serotype may require distinct process conditions. Together, these factors contribute to high production costs and can slow the development or broader deployment of expanded-valency vaccines.

Carrier proteins, while essential for providing T-cell help [6], also impose constraints. Repeated exposure to the same carrier across multiple vaccines can bias immune responses and has been associated with carrier-related immune interference in certain settings [7]. Carrier proteins additionally increase product heterogeneity due to multiple potential conjugation sites and variable loading, complicating analytical release and quality control. These considerations have driven sustained interest in “carrier-sparing” strategies [8] that can deliver cognate CD4^+^ T-cell help with reduced reliance on large protein carriers, potentially enabling more modular manufacturing and improved control over molecular composition.

One conceptual solution is to replace the carrier protein with a defined T-helper epitope peptide that can be covalently linked to CPS. In such a design, the polysaccharide provides the B-cell epitope, while the attached helper peptide supplies T-cell help through linked recognition: CPS-specific B-cells capture the conjugate, internalize it, and present peptide-derived epitopes on MHC class II to recruit CD4^+^ T-cell support [6,9,10]. Among candidate helper epitopes, PADRE (Pan-DR epitope) has been widely used as a broadly MHC class II–binding sequence in humans and represents an attractive module for carrier-protein-free conjugate vaccine architectures [11,12]. Compared with protein carriers, peptide-based helpers offer the prospect of reduced heterogeneity, more precise molecular design, and simplified quality control.

To realize this approach in a practical vaccine format, the conjugation chemistry must be efficient, mild, and broadly compatible with structurally diverse polysaccharides. Notably, while polysaccharide–protein conjugation is well established, polysaccharide–(poly)peptide conjugate vaccines remain far less explored, with comparatively limited methodological and immunological datasets available.

Polysaccharide–peptide conjugates have most often been assembled via condensation/amide-forming reactions after introducing amine or carboxyl handles onto the polysaccharide (e.g., EDC/NHS- or CDI-mediated coupling) [13,14,15]. While effective in some settings, these routes can become difficult to control unless orthogonal protection/site-specific handle design is implemented, because peptides often contain multiple nucleophilic residues (e.g., Lys, N-terminus) and/or carboxylates (Asp/Glu), leading to product heterogeneity. As an alternative to direct polysaccharide–peptide coupling, some groups have genetically fused helper epitopes into a recombinant protein scaffold and then conjugated the polysaccharide to that protein [13,16,17,18], which can improve control over epitope placement and copy number but, in practice, reverts to a protein-carrier-based conjugate paradigm with the associated manufacturing and immunological considerations.

In contrast, bioorthogonal click reactions provide a compelling platform because they form stable covalent bonds under aqueous conditions with minimal side reactions and without harsh reagents that could compromise CPS integrity. In particular, copper-free strain-promoted azide–alkyne cycloaddition (SPAAC) enables rapid coupling of azide- and cyclooctyne-functionalized partners without metal catalysts, simplifying purification and reducing concerns about residual catalyst toxicity. Importantly, copper-free SPAAC has already been used to construct well-defined polysaccharide–peptide/polypeptide conjugates under fully aqueous, catalyst-free conditions. For example, Jain et al. reported copper-free SPAAC coupling of azide-functionalized adhesive peptides onto cyclooctyne-modified alginate, enabling efficient peptide grafting without metal catalysts [19]. More recently, Neves et al. demonstrated SPAAC ligation of azide-functionalized RGD and protease-sensitive peptide motifs (PVGLIG) to BCN-modified alginate, highlighting high chemoselectivity and compatibility with sensitive peptide domains in water [20]. Beyond short peptide motifs, metal-free azide–alkyne cycloaddition has also been used to generate polysaccharide–polypeptide architectures (e.g., grafting poly(L-lysine) or poly(L-glutamic acid) side chains onto polysaccharide backbones such as hyaluronic acid, chitosan, or heparin) [21,22]. There are various copper-free click reactions can also be used for biocompatible conjugations [23].

Notably, these precedents are predominantly biomaterials- or delivery-oriented and mainly involve non-bacterial polysaccharide backbones (e.g., alginate, hyaluronic acid, chitosan/heparin derivatives) rather than native bacterial CPS used as vaccine antigens. Moreover, systematic immunological evaluation—particularly direct comparison between peptide-based helper modules and traditional protein carriers using the same CPS antigen—remains limited.

In this work, we propose a carrier-protein-free pneumococcal conjugate vaccine strategy that couples a pneumococcal CPS to PADRE using a modular, copper-free click-chemistry framework. By combining mild polysaccharide functionalization with catalyst-free SPAAC, this design aims to (i) reduce dependence on carrier proteins, (ii) simplify and modularize conjugate assembly, and (iii) enable systematic optimization of conjugate architecture. Using pneumococcal CPS as a representative polysaccharide antigen, we outline a practical route to CPS–PADRE conjugates and establish an experimental foundation for evaluating peptide-based helper modules as alternatives to conventional carrier proteins in next-generation pneumococcal glycoconjugate vaccines.

2. Materials and Methods

2.1. Reagents, Materials and Animals

All chemical solvents were purchased from Sigma-Aldrich (St. Louis, MO, USA), Acros (Geel, Belgium), Adamas-beta^®^ (Shanghai, China), and Thermo Fisher (Waltham, MA, USA). DMAP (4-(Dimethylamino) pyridine) was obtained from Aladdin (Shanghai, China). CDAP (1-cyano-4-dimethylaminopyridinium tetrafluoroborate) and D_2_O were purchased from Energy Chemical (Shanghai, China) (or equivalent). Unless otherwise specified, reagents were used as received.

CPS was purchased from SSI Diagnostica (Copenhagen, Denmark), received as a white solid, and is readily soluble in water. The proton NMR spectra of the CPS were recorded (400 MHz, 25 °C, 5 mg/mL in D_2_O). The molecular weight of serotype 15C CPS was estimated by high-performance size-exclusion chromatography with refractive index detection (HPSEC–RI) using Shodex OHpak SB-803 and SB-806 columns with 0.1 M phosphate buffer (pH 7.2) as the mobile phase. A calibration curve was generated using dextran standards (6, 20, 40, 70, 500, and 2000 kDa), where the retention time of the polysaccharide peak was plotted against log(MW) (see the Figure S6). The 15C CPS showed a peak retention time of 31.12 min; based on the calibration curve, its molecular weight was calculated to be ~1096.48 kDa (Figures S5–S7).

Azide-functionalized PADRE and RR–PADRE peptides (purity ≥95%, HPLC) were purchased from (Jier Biochemical Co., Ltd., Shanghai, China). CRM197 was purchased from FinaBio (Rockville, MD, USA).

BALB/c mice (6–8 weeks old) were purchased from (SiPeiFu Biotechnology Co., Ltd., Beijing, China) and randomly assigned to groups (n = 5). All animal procedures were approved by School of Pharmaceutical Sciences of Hainan University Animal Care and Use Committee, HPIACUC2025057.

2.2. Preparation of 15C CPS–BCN (15C–BCN) by CDAP Activation

2.2.1. Solution Preparation

CPS solution (5 mg/mL, 150 mM NaCl): Serotype 15C CPS was dissolved in water to 10 mg/mL. An equal volume of 300 mM NaCl was added to yield 5 mg/mL CPS in 150 mM NaCl. The solution was pre-cooled at 4 °C.

DMAP solution (2.5 M, pH 8.0): DMAP was dissolved in water and titrated with 10 M HCl until clarified and pH 8.0 was reached; volume was adjusted to final concentration.

CDAP solution (100 mg/mL): CDAP was freshly dissolved in acetonitrile immediately before use.

Amino-BCN linker solution (2 mg/mL): NH_2_–BCN was dissolved in water.

2.2.2. CDAP-Mediated Coupling

All solutions were pre-chilled on ice. Under stirring, DMAP solution (DMAP stock solution: 2.5 M; added at 1:10 v/v vs. CPS solution; final DMAP concentration: 227 mM.) and CDAP (e.g., 10 mg, the mass ratio of CDAP to CPS is 1:1.) were added to the CPS solution and allowed to activate for 15 min while maintaining pH 8–9. NH_2_–BCN (Functional-group equivalents: the reaction was set such that 1 equiv. of 15C CPS corresponded to 0.2 equiv. of NH_2_–BCN in the input mixture.) was then added and the pH was kept at 8–9 using 0.1 M NaOH as needed. The reaction proceeded at room temperature for ~1 h and was then stirred overnight at 4 °C.

2.2.3. Purification

The crude reaction mixture was purified by dialysis using a (100 kDa MWCO) membrane sequentially against 1 M NaCl, 0.15 M NaCl, and water (two rounds; buffer replaced every 12 h). The retentate was recovered and lyophilized to afford 15C–BCN 73.5%~89.7%

2.3. SPAAC Conjugation of 15C–BCN with Azide-Peptides

2.3.1. 15C–PADRE Conjugate

The 15C–BCN (e.g., 5 mg,10 mg/mL) and azide–PADRE (e.g., 2 mg,4 mg/mL) were dissolved separately in 500 μL D_2_O, Functional-group equivalents: 1 equiv. of 15C–BCN corresponded to 0.2 equiv. of PADRE (or RR–PADRE) in the input mixture. The peptide solution was added dropwise to the polysaccharide solution and stirred at room temperature. Reaction progress was monitored by ^1^H NMR until the disappearance of BCN diagnostic resonances (~2 h). Excess free peptide was removed using a (5 kDa MWCO) centrifugal filter or by dialysis (10 kDa MWCO) as appropriate. The purified conjugate was lyophilized.

2.3.2. 15C–RR–PADRE Conjugate

The same procedure was followed using azide–RR–PADRE. The RR motif is described here as a dibasic “putative protease-cleavable” sequence positioned between the linker attachment site and the PADRE.

2.3.3. Preparation of 15C-CRM197 Conjugate (Benchmark)

15C CPS (5 mg) was dissolved in 2.5 mL of 150 mM NaCl (2 mg/mL). DMAP was added using the same proportional dosing strategy as in Section 2.2.2 (DMAP stock: 2.5 M; pH 8.0;1:10 v/v vs. CPS solution; final DMAP: 227 mM). DMAP and CDAP (100 mg/mL) were added to activate the CPS for 15 min while maintaining pH 8–9. CRM197 was then added at a CPS: protein mass ratio of 1:1 (e.g., 142.9 µL of 35 mg/mL CRM197). The mixture reacted at room temperature for 1 h and was stirred overnight at 4 °C. Purification was performed using a (100 kDa MWCO) centrifugal filter with repeated washes in 150 mM NaCl (4000× g, 20 min, 3–4 cycles) (Figure S1). Final concentration was determined by phenol-sulfuric acid method. Ultimately, we conducted a comprehensive characterization of the vaccine using SDS/PAGE (Figure S4).

2.4. NMR Characterization and DOSY Validation

^1^H NMR spectra were acquired in D_2_O on a Bruker NMR, 400 MHz spectrometer. The BCN degree of substitution was estimated by integrating BCN-specific resonances relative to characteristic CPS resonances, yielding a substitution level of ~17.9%.

DOSY NMR was used to confirm covalent conjugation. In principle, covalent linkage forces peptide-derived resonances and CPS resonances to share a similar diffusion coefficient (co-diffusion). In contrast, free peptide diffuses substantially faster and appears with a markedly higher diffusion coefficient. Co-diffusion of peptide and CPS resonances in DOSY, together with MWCO-based removal of free peptide, was taken as evidence of true conjugation.

2.5. Vaccine Formulation and Immunization

The adjuvant was Al(OH)3, (Maikekang Biotechnology Co., Ltd., Chengdu, China) (10 mg/mL stock). Mice received 100 µL intraperitoneal injections on days 0, 14, and 28. Each dose contained 3 µg CPS-equivalent antigen (15C, 15C–PADRE, 15C–RR–PADRE, 15C–CRM197, and two physical mixture controls (15C + PADRE and 15C + RR–PADRE) where applicable) and 100 µg adjuvant. Sera were collected on days 0, 13, 27, and 41. Tolerability was monitored by body weight and general appearance (e.g., grooming, ruffled fur, activity).

2.6. ELISA for CPS-Specific Antibodies

High-binding plates were coated overnight at 4 °C with biotinylated 15C CPS (1 µg/mL in PBS, 100 µL/well). Plates were washed with PBST and blocked with PBS containing 2% BSA (200 µL/well, 2 h at room temperature). Sera were serially diluted in PBST containing 1% BSA starting at 1:400 to 1:12,800. After incubation (2 h, room temperature), plates were washed and incubated with HRP-conjugated goat anti-mouse secondary antibodies (total IgG, IgM, and IgG subclasses IgG1, IgG2a, IgG2b, IgG3) were diluted to working concentrations of 1:5000;1:5000;1:1000;1:2000;1:2000;1:200, respectively, according to the manufacture’s recommended dilution ratios. TMB substrate was developed for 10–15 min and stopped with 1 M HCl. OD450 was read within 10 min.

Calculation of Antibody Titer: The titers of IgG, IgM, and relevant IgG subclasses were defined as the reciprocal of the highest serum dilution that yielded a positive reaction in the enzyme-linked immunosorbent assay (ELISA). A positive result was determined as an optical density (OD) value ≥ the cut-off, which was set at twice the mean OD value of the negative control group. For samples that tested positive at all dilutions, the antibody titer was recorded as ≥ [reciprocal of the highest dilution]; for those that tested negative at all dilutions, the antibody titer was recorded as ≤ (reciprocal of the lowest dilution). The optimized standard curve Y = 1.0305X − 1.0053 (coefficient of determination R^2 = 0.9999) was adopted, where X represented the logarithmic value of the obtained antibody titer. Subsequently, the titer was multiplied by the dilution factor, and the geometric mean titer was finally calculated based on the valid endpoint data (Figure S5).

2.7. Microarray-Based Antibody Binding Assay

Serotype 15C CPS was prepared in 20% DMSO at four concentrations (0.5, 0.1, 0.02, and 0.004 mg/mL) and printed onto photoreactive slide chemistry—diazirine-functionalized substrates. After overnight drying, slides were UV-irradiated for 12 min to covalently immobilize CPS, washed, dried, and blocked with TBS containing 1% BSA (100 µL/well, 1 h at 25 °C). Sera were diluted 1:100 and incubated for 5 h at 25 °C, followed by PBST washes and incubation with fluorescent secondary antibody (1:1000, 2 h, protected from light). Slides were scanned and analyzed using Mapix software version 8.2.7 (Innopsys Inc., Carbonne, France) (background subtraction and quantification). Outliers were removed by Grubbs’ test (α = 0.05). Graphs were generated using GraphPad Prism 9.0.

2.8. Opsonophagocytic Killing Assay (OPKA)

OPKA was performed to evaluate the functional opsonophagocytic activity of vaccine-induced serum antibodies. Fetal bovine serum (FBS) was heat-inactivated at 56 °C for 30 min. Culture media were prepared as follows: CM1 (IMDM medium 45 mL + heat-inactivated FBS 5 mL + penicillin/streptomycin 0.5 mL + GlutaMAX-1 0.5 mL), CM2 (IMDM medium 45 mL + heat-inactivated FBS 5 mL + GlutaMAX-1 0.5 mL + DMF 0.4 mL), and CM3 (IMDM medium 40 mL + heat-inactivated FBS 10 mL + penicillin/streptomycin 0.5 mL + GlutaMAX-1 0.5 mL). OBB buffer was freshly prepared as 1 × HBSS (with Ca^2+^ and Mg^2+^),1% gelatin solution, and heat-inactivated FBS (85:10:5, v/v/v) and was not stored overnight. Cell washing solution: 1× HBSS (without Ca^2+^ and Mg^2+^). Complement was stored at −80 °C and thawed at room temperature before use; heat-inactivated complement was prepared by incubating complement at 56 °C for 30 min.

HL-60 cells were revived from cryopreservation and expanded in CM3 medium at 37 °C under 5% CO_2_. To induce differentiation, HL-60 cells were first expanded in CM3/CM1 medium while maintaining a density below 5 × 10^5^ cells/mL. Cells were then washed twice with PBS, resuspended in CM2 medium at 4 × 10^5^ cells/mL, and cultured for 5–6 days. Cell differentiation status was evaluated by flow cytometry: apoptosis was evaluated using FITC-conjugated 7AAD, and surface marker expression was analyzed using PE-conjugated anti-CD35 and anti-CD71 antibodies. Differentiated cells were deemed suitable for the opsonophagocytic killing assay (OPKA) if they met the following criteria: ≤25% apoptosis, ≥55% CD35^+^ cells, and ≤20% CD71^+^ cells.

Target bacteria were prepared in OBB buffer at a working concentration of 1 × 10^7^ CFU/mL. Differentiated HL-60 cells were washed sequentially with HBSS lacking Ca^2+^/Mg^2+^ and then with complete HBSS, followed by resuspension in fresh OBB buffer at 5 × 10^6^ cells/mL. The assay was performed in U-bottom 96-well plates. Serum samples (1.6 µL/well) were first opsonized by incubation with the bacterial suspension (10 µL/well) at 37 °C with shaking (200 rpm) for 30 min. Subsequently, a mixture of complement (or heat-inactivated complement) and differentiated HL-60 cells (1:9 ratio) was added (50 µL/well), bringing the final reaction volume to 80 µL. The plates were then incubated at 37 °C with shaking (700 rpm) for 45 min to facilitate phagocytic killing. Reactions were terminated by transferring the plates to ice for 20 min. Aliquots (10 µL) from each well were plated onto agar plates in triplicate. After incubation at 37 °C for 18 h, bacterial colonies were enumerated.

The percent killing was calculated as: (1 − CFU sample/CFU control) × 100%, where CFU control represents the colony count from wells without serum.

2.9. Statistics

Specify the statistical test based on your final dataset and distribution (commonly one-way ANOVA with Tukey’s post hoc test for multiple comparisons; or Kruskal–Wallis with Dunn’s test for non-parametric data). Report exact p values where possible.

3. Results

3.1. Design Rationale and Overall Workflow

To reduce dependence on carrier proteins and simplify manufacturing, we implemented a carrier-protein-free glycoconjugate strategy that combines two key steps: (i) CDAP-mediated installation of a bicyclononyne (BCN) handle on native capsular polysaccharide (CPS) under mild aqueous conditions, and (ii) catalyst-free SPAAC (strain-promoted alkyne–azide cycloaddition) to couple a defined T-helper epitope peptide (PADRE) to the CPS. PADRE (Pan-DR Epitope) is a synthetic peptide designed to bind to multiple MHC class II alleles, thus enhancing immune response through helper T-cell activation

For this strategy, serotype 15C CPS was selected as a stable model antigen available from in-house stocks. Two peptide designs were compared: PADRE without a cleavable motif, and RR–PADRE, which incorporates a cleavable dibasic RR motif positioned between the linker attachment site and the PADRE sequence [24,25,26].

This strategy minimizes the need for traditional carrier proteins, potentially reducing manufacturing costs while enhancing the consistency and efficacy of the resulting glycoconjugate vaccines.

3.2. CDAP Chemistry Enables Controlled BCN Installation on 15C CPS

Activation of 15C CPS with CDAP followed by coupling to NH_2_–BCN yielded 15C–BCN in high isolated yield (Figure 1, 73.5%~89.7% by dry mass) after dialysis. ^1^H NMR spectra of 15C–BCN displayed diagnostic BCN resonances absent in the native CPS, confirming successful installation of click handles. Integration-based analysis indicated an average BCN substitution level of ~17.9% (Figure 2). This substitution was considered well balanced: sufficiently high to provide multiple peptide attachment sites while avoiding excessive modification that could compromise CPS B-cell epitopes.

To exclude the possibility that BCN signals originated from incompletely removed small molecules, DOSY NMR was performed (Figure 3). BCN-associated resonances co-diffused with the CPS resonances, consistent with a single macromolecular species and supporting covalent incorporation of BCN into the CPS scaffold.

3.3. SPAAC Proceeds Efficiently by Simple Mixing and Yields Bona Fide CPS–Peptide Conjugates

Mixing 15C–BCN with azide-PADRE or azide–RR–PADRE in D_2_O resulted in rapid SPAAC conjugation at room temperature Figure 4 and Figure 5. ^1^H NMR monitoring showed progressive loss and eventual disappearance of BCN diagnostic peaks within ~2 h, indicating near-complete consumption of the strained alkyne. After MWCO-based purification (ultrafiltration, 100 kD molecular weight cut-off), peptide-associated resonances remained detectable in the product. The peptide loading ratio was calculated based on NMR data. Both PADRE and RRPADRE contain phenylalanine (F) and tryptophan (W), contributing a total of 10 aromatic protons in the phenyl region. In the ^1^H NMR of 15C-PADRE, the integral of the polysaccharide’s N-acetyl group is 1, and the integral of the peptide’s phenyl region is 0.59. Therefore, the peptide conjugation rate is calculated as follows: (0.59/10)/(1/3) = 0.177, which is very close to the BCN level (17.9%). In the ^1^H NMR of 15C-RRPADRE, the integral of the polysaccharide’s N-acetyl group is 1, and the integral of the peptide’s phenyl region is 0.57. Thus, the peptide conjugation rate is calculated as: (0.57/10)/(1/3) = 0.171 (Figures S8 and S9). Importantly, DOSY NMR demonstrated co-diffusion of peptide resonances with CPS resonances (Figures S2 and S3), consistent with covalent conjugation rather than residual free peptide. Together with the ability of the MWCO purification to remove free peptide in principle, these data support successful formation of 15C–PADRE and 15C–RR–PADRE conjugates.

3.4. Vaccination Study

To enable a PCV-benchmarked, head-to-head comparison of conjugate architectures (CPS–PADRE ± RR versus the CPS–protein benchmark) under controlled conditions, we employed a standardized prime–boost immunization regimen commonly used for pneumococcal conjugate vaccine evaluation in mice. Immunizations were performed via the intraperitoneal (i.p.) route, which is widely used in early-stage murine vaccine screening because it provides highly reproducible delivery, robust serological readouts, and facilitates sensitive discrimination between closely related conjugate designs. Importantly, the same route was used for all groups, ensuring that comparisons reflect conjugate architecture rather than differences in administration. A CPS-equivalent dose of 3 µg per immunization was selected as an intermediate, non-saturating dose within the range commonly used for mouse evaluation of pneumococcal glycoconjugates, with the aim of preserving sensitivity to detect design-dependent differences (PADRE vs. CRM197; ±RR) under a single controlled condition. Mice (n = 5 per group) were immunized with a three-dose schedule at 14-day intervals (Figure 6), and sera were collected on days 0, 14, 28, and 42. All groups were administered the same CPS-equivalent antigen dose (3 µg per immunization) and adjuvant dose (100 µg per immunization) to ensure that observed differences reflect conjugate design rather than differences in dosing or formulation.

Across all immunization groups, animals maintained stable body weight and exhibited no overt adverse signs (e.g., ruffled fur or reduced activity), indicating acceptable tolerability under the tested dosing regimen (3 µg CPS-equivalent antigen + 100 µg adjuvant per dose).

CPS-specific ELISA revealed that the conventional 15C–CRM197 conjugate elicited the highest total IgG responses (Figure 7A). 15C-PADRE conjugate induced measurable IgG, demonstrating that PADRE can provide functional T-cell help in a carrier-protein-free format. However, overall antibody levels induced by 15C-PADRE conjugate were clearly lower than those elicited by the protein conjugate.

Notably, 15C–PADRE (no RR) consistently produced higher total IgG and IgG subclass responses than 15C–RR–PADRE. Microarray-based antibody profiling (Figure 7C) supported the same trend and provided an orthogonal platform for comparing binding signals across immobilized CPS concentrations.

To examine whether CPS-specific antibody levels translate into functional antibacterial activity, we performed an opsonophagocytic killing assay (OPKA) using sera collected at day 42 (Figure 8). Overall, opsonophagocytic activity tracked closely with CPS-specific IgG levels measured by ELISA. The 15C–CRM197 conjugate induced the highest functional activity, yielding an average of 80% killing with 1:50 serum dilution, which was significantly greater than the carrier-protein-free constructs (59% and 40% for 15C-PADRE and 15C-RR-PADRE respectively).

Sera from the 15C–PADRE group showed clear, above-background killing activity, supporting the feasibility of the carrier-protein-free CPS–helper epitope architecture. In contrast, incorporation of the putative cleavable junction (15C–RR–PADRE) did not enhance function and instead resulted in reduced opsonic activity relative to the non-cleavable design, consistent with the diminished humoral responses observed in the binding assays. Physical-mixture controls (15C + PADRE and 15C + RR–PADRE) remained at or near baseline levels, indicating that covalent linkage is required to elicit robust functional activity in this setting.

Although pneumococcal sepsis challenge studies can provide direct evidence of protection, we did not include such experiments in the present work. This study was designed primarily as a feasibility comparison (carrier substitution and junction architecture) against a protein–conjugate benchmark under matched dosing and formulation conditions. Because the carrier-protein-free constructs did not exceed the CRM197 benchmark under the tested regimen, a resource- and risk-intensive challenge model—requiring specialized biosafety infrastructure, substantial animal numbers for adequate power, and extended timelines—was unlikely to yield proportionate additional insight at this stage. We therefore used OPKA as a mechanistically relevant functional correlate to link antibody responses to antibacterial activity and will pursue in vivo challenge studies after further optimization of construct architecture and formulation.

4. Discussion

This study presents a practical route to carrier-protein-free pneumococcal CPS–peptide conjugate vaccines that combines CDAP-based handle installation with catalyst-free SPAAC.

Copper-free bio-orthogonal ligation—most commonly strain-promoted azide–alkyne cycloaddition (SPAAC)—has been increasingly used to construct polysaccharide–peptide/polypeptide conjugates under mild aqueous conditions, enabling chemoselective bond formation while avoiding metal-associated toxicity and minimizing damage to carbohydrate epitopes. Representative reports typically rely on (i) installing azide or cyclooctyne handles onto model polysaccharides (e.g., alginate, hyaluronic acid, chitosan/heparin derivatives), synthetic glycopolymers, or defined oligosaccharide scaffolds via carbodiimide/NHS-type coupling or related activation chemistries, followed by (ii) catalyst-free SPAAC grafting of cyclooctyne- or azide-functionalized peptide/polypeptide segments in buffered media at ambient temperature [19,20,21,22]. In these systems, practical conjugation efficiency is often governed by steric congestion and polymer viscosity, frequently necessitating excess click partner and extended reaction times (hours to days); accordingly, efficiency is commonly quantified indirectly using a combination of NMR integration, UV/fluorophore readouts, and/or compositional analysis. Against this backdrop, our CDAP-enabled handle installation coupled with SPAAC aligns with the established operational advantages of metal-free click chemistry, while extending its application to a native pneumococcal capsular polysaccharide (CPS) in a vaccine-relevant format. Notably, the CPS–peptide coupling was completed within a few hours. Although faster click reactions have been reported, they typically require additional synthetic steps and/or suffer from limited reagent stability [23]. The loading efficiency achieved here can be quantitatively cross-validated: peptide loading derived from ^1^H NMR integration is consistent with the independently determined cyclooctyne (BCN) substitution level (∼17–18%), supporting bona fide covalent conjugation and efficient ligation on CPS backbone. This ensures meaningful comparisons in downstream biological evaluation.

In mice, CPS–PADRE conjugates were immunogenic and well tolerated yet remained inferior to the conventional CRM197 conjugate in terms of CPS-specific IgG magnitude. Unexpectedly, inclusion of a dibasic RR motif intended as a “cleavable” element diminished, rather than improved, antibody responses.

4.1. Why Might the RR “Cleavable” Motif Reduce Responses in This CPS–PADRE System?

Although cleavable linkers have been reported to enhance responses in some conjugate contexts [22,23,24], their benefit is not necessarily universal—particularly when the B-cell epitope is a polysaccharide and the helper epitope is a short peptide. Several non-mutually exclusive mechanisms could explain our observations.

Suboptimal cleavage timing or compartment: RR cleavage may occur inefficiently (or in non-ideal cellular compartments), leading to processing products that do not favor robust PADRE presentation or Tfh engagement processing.

Bias and epitope competition: RR may create alternative proteolytic hotspots or shift peptide processing toward non-productive fragments, reducing the display of the desired PADRE core epitope.

Differences in peptide solubility: Even with identical CPS–BCN substitution, the two peptides differ in solubility: RR-PADRE has a better water solubility. It may interfere with the antigenicity.

Cleavable designs may be more beneficial for protein/peptide B epitopes than for polysaccharide B epitopes: For polysaccharides, maintaining the physical integrity of the conjugate during key steps of BCR-mediated uptake and antigen presentation may be more important than releasing the helper epitope.

4.2. Why Are CPS–PADRE Conjugates Still Weaker than CPS–CRM197?

Several factors likely contribute to the superior performance of CRM197 conjugation:

Higher T-helper capacity: CRM197 provides multiple T-helper epitopes, enabling broader and stronger Tfh recruitment compared to a single PADRE.

Size and trafficking advantages: Protein conjugation increases molecular complexity and may enhance lymph node retention and antigen uptake by APCs.

MHC restriction considerations: PADRE is optimized for broad human HLA–DR binding; its functional dominance in BALB/c (murine I-A/I-E) may be less optimal than in human settings.

Need for optimized formulation: Carrier-protein-free peptide carriers may require additional engineering (higher epitope density, multiepitope designs, delivery systems) to match the immunogenicity of established protein carriers.

4.3. Limitations and Future Directions

This work focused on a single CPS serotype and a limited set of peptide/linker designs. In addition, immunizations were performed via the intraperitoneal route, which—while commonly used for early-stage murine screening—may amplify responses to weaker antigens and is not the clinical route for human pneumococcal vaccination. Future studies should (i) quantify peptide loading on each conjugate, (ii) assess functional antibody activity (e.g., opsonophagocytic killing and challenge study), (iii) explore alternative protease-sensitive motifs better matched to endosomal cathepsins, (iv) evaluate clinically relevant administration routes (e.g., intramuscular or subcutaneous) and dosing regimens, and (v) evaluate delivery strategies (e.g., nanoparticles/liposomes) to increase lymph node targeting and germinal center responses.

5. Conclusions

In summary, this study outlines a practical “carrier-protein-free” pneumococcal glycoconjugate vaccine concept in which a defined CD4^+^ T-helper epitope (PADRE) is used to substitute for conventional carrier proteins. By leveraging mild polysaccharide functionalization and a modular, copper-free click-chemistry coupling strategy, the platform enables straightforward construction of CPS–peptide conjugates while preserving the integrity of the polysaccharide antigen and providing a clear path toward scalable, standardized assembly. In vivo evaluation demonstrates that CPS–PADRE conjugates can elicit CPS-specific IgG responses with good tolerability, supporting the feasibility of supplying T-cell help in a peptide-based format. Importantly, introducing an RR “cleavable” junction between the polysaccharide and PADRE did not enhance humoral immunity in this architecture and instead reduced antibody responses relative to a non-cleavable design, underscoring that linker cleavability is not universally beneficial and must be empirically optimized for polysaccharide–helper epitope vaccines. Although a conventional CPS–protein conjugate remained superior under the conditions tested, these findings provide a rationale and a modular framework for further refinement of carrier-protein-free conjugates through optimization of helper-epitope density, junction design, and formulation/delivery. Collectively, this work advances a scalable design strategy for next-generation pneumococcal glycoconjugate vaccines that may ultimately reduce manufacturing complexity while enabling tunable immunological performance.