Whole protein or long peptides in therapeutic vaccination strategies, does it make a difference?

1. Introduction

The cancer vaccine field is expanding as a result of the increasing use of checkpoint inhibitors targeting programmed death-1 (PD-1) or programmed death receptor ligand-1 (PD-L1). With the limited numbers of tumor-associated surface antigens that can be targeted by antibody-based therapeutics, the ligandome, presented on the cell surface by human leukocyte antigens (HLA), represents a unique opportunity to broaden the therapeutic targets within the cancer field. Still, therapeutic development has so far been oriented to epitopes derived from tumor-associated antigens (TAA) or tumor-specific antigens (TSA), capturing the aberrant expression of cancer testis antigens, tissue differentiation antigens, viral antigens, and neoantigens [1], ignoring the remaining ligandome signature in the tumor micro-environment. Stromal cells, vessels, and immune cells represent alternative targets pursued in the clinic [2]. These more unconventional targets are still plausible therapeutic risks, if the induced immune responses can lead the harm of healthy tissue, but also represent a new and alternative drug class opportunity if harnessed correctly. Today, we witness the assessment of peptide-based cancer vaccines incorporating immunologically relevant targets such as PDL1 and indoleamine 2, 3-dioxygenase (IDO), proteins with an upregulated expression pattern in the TME, but also targets not exclusively expressed in the TME [3,4]. So how shall we consider these targets, as opportunities or risks, in cancer vaccine drug development? I wish to argue that it depends on vaccine design, or more specifically, if peptides or whole proteins are used to induce a directed vaccine-induced immune response.

2. Immunogenicity risks of proteins drugs

Traditional pharmaceutical development of biological drugs has stumbled over the class effect that immunogenicity represents. The risk of inducing an immune response in the form of anti-drug antibodies (ADAs) against endogenous proteins [5] can lead to lifelong and life-threatening diseases and is a risk associated with most protein-based drugs. An example of such impact is the known risk of ADA-induced anemia as a result of synthetic erythropoietin (EPO) therapy [6]. In cancer vaccines, these risks are mainly avoided by maintaining a strict selection of protein targets (focus on TAA, TSA) within cancer vaccine platforms, avoiding proteins present in the immune cell, stroma, and vessel compartment. But less discussed in the light of ADA risks associated with protein-based drugs is the opportunity of synthetic long peptide (SLP) vaccines to enable an immunological reaction to non-tumor cell-produced proteins, without the associated risk of developing a polyclonal antibody response leading to unexpected adverse events for patients. The purpose of such opportunity would thus be to broaden the number of pharmaceutical targets, while avoiding ADA, ultimately enabling improved endogenous immune-based regulation/control of tumor-promoting factors in the TME.

3. The HLA ligandome repertoire

The use of whole protein always represents an opportunity to capture the extent of the immunological repertoire that can be induced against a tumor antigen. As such, efforts of using cell-based or whole protein TAAs in cancer vaccines have been a valid choice in the design of cancer vaccines. But as processing of whole proteins can lead to an unpredictable HLA ligandome repertoires depending on the protein and the degradation capability of the antigen-presenting cell [7,8] along with the competition of peptides to be loaded on the MHC molecules [9], it is not obvious that more and diverse epitope patterns improve efficacy. Many advantages have been described for peptide-based vaccines relating to the ease of synthesis, modification, and storage, which also enable efficient drug development and use [10].

Intriguingly, SLP-based vaccines present a focused peptide/HLA ligandome on the presenting APC, enabling a stringent T cell response. If relevant and important epitopes for an effective immune response are shunned, due to the uneven processing and presentation of the whole protein or by intracellular ligandome competition, a limited and unpredictable HLA ligandome repertoire will be the end game. To avoid this, peptides trimmed to fit the HLA groove have been used in vaccine development as a means to improve selectivity and boost specific T cell expansion with a trial in melanoma displaying clinically effective immune responses when combined with high-dose IL2 therapy [11] but not with anti-CTLA-4 therapy [12]. The field thus turned to the development of SLPs to ensure uptake and processing of peptides by APCs and avoid antigen presentation by nonprofessional APCs. Both trimmed peptide vaccines and SLPs are considered safe and are generally not troubled by immunogenicity issues in the form of ADA responses. And even if ADA responses can occur against peptides, the risk of this response being cross-reacting with the same peptide imbedded as a structural epitope in the whole protein is limited. This is due to the lack of native folding by a short peptide stretch, in their shorter synthetic form [13], and as native structure is not required for the efficacy of the HLA/peptide presentation to the T cells, the functionality when it comes to T cell priming is still intact.

4. Safety of peptide vaccines in light of ADA

The inherited low risk of peptide vaccine to be associated with an ADA response against the naturally folded protein is a benefit that is rarely discussed, still the plausible avoidance of ADA responses can actually broaden the HLA ligandome repertoire outside the focus of the TAAs and TSAs. If peptide vaccines represent a means to avoid protein binding ADA responses, it can also explain why targeting of endogenous proteins such as PDL1 and IDO can be performed without association of adverse events greater than the checkpoint therapies alone [2]. If we would have the possibility to target the whole TME HLA ligandome, there can be unconventional non-tumor cell-derived targets presented by APCs, accumulating in high concentrations due to increased protein production and degradation in the inflammatory TME, that could represent targets enabling T cell activation and homing to the tumor area. How this can promote tumor killing remains to be established, e.g. if this represents a means to improve the control of promoting inflammatory or immune suppressive cells. Alternatively, the strategy may promote epitope spread as a way to control tumor growth.

5. Proteins or peptides in the light of the mRNA era

In light of the successful introduction of mRNA into the vaccine platform opportunities, it is important to revisit the choice of the coding sequence and the resulting peptide or protein generated as target antigen of choice for vaccine development. As the mRNA delivery platform allows for the incorporation of longer sequences, proteins can and will be used ahead, but can face the same risk of ADA induction and potential skewness of the HLA/peptide ligandome as prior whole protein-based platforms. Hence, also for mRNA vaccines, the coding of elements representing parts of a protein can and should be evaluated. Of course, with the intricate knowledge of the actual HLA ligandome pattern within the TME as a means to also secure T cell homing after vaccination.

6. Expert opinion

SLP-based vaccines aim to induce a targeted T cell response against the tumor cell, and the combination of these novel vaccines with checkpoint therapies is under evaluation in multiple randomized phase II trials. New unconventional vaccines embedding peptides from proteins derived from non-tumor cell sources, with high abundance in the TME, attract attention. The somewhat untraditional approach can be viewed as a potential drug development risk for unwanted adverse events in the form of immunogenicity, but should they be viewed as such? In one way yes, as the induced T cell response can lead to unwanted T-cell-mediated autoimmune responses in non-tumor-related sites, similar to therapies aiming to work via T effector cells. But, as the HLA ligandome is likely different in non-inflamed healthy tissue compared to the TME, it warrants further investigations.

An uncontrolled ADA response to a whole protein can lead to unpredictable and rapid systemic effects causing cell-killing, hypersensitivity reactions, or immune-complex-mediated inflammatory responses as a risk for essential organs. This is, however, rarely a risk when it comes to peptide-based vaccines. As peptide vaccines in trials are now assessing the induction of T cell responses against immune-related factors in the TME, with, so far, no toxicity reported to exceed conventional CPI therapy, we should therefore dare to explore alternatives to targets via SLP-based therapeutics, or nucleotide-based vaccines coding for the same peptides. This fear of ADA, known to be associated with whole protein exposures, can, with careful peptide selection based on unique HLA ligandome expression signatures in the TME, lead us to a unique and safe way to bolster immune responses toward suppressive or tumor promoting TME signatures, thereby harnessing the full capacity of our powerful immune surveillance program. As with mRNA, novel formulation platforms designed to optimize peptide vaccine efficacy can also transform the field ahead.

Declaration of interest

S Mangsbo is Chief Innovation Officer at Ultimovacs AB and Chief Scientific Officer at Strike Pharma AB and owns stocks in the respective companies. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This paper was funded by Vinnova (Grant No. 2019-00048).