Plant-derived Durvalumab variants show efficient PD-1/PD-L1 blockade and therapeutically favourable FcR binding.

Immune checkpoint blocking therapy targeting the PD-1/PD-L1 inhibitory signalling pathway has produced encouraging results in the treatment of a variety of cancers. Durvalumab (Imfinzi®) targeting PD-L1 is currently used for immunotherapy of several tumour malignancies. The Fc region of this IgG1 antibody has been engineered to reduce FcγR interactions with the aim of enhancing blockade of PD-1/PD-L1 interactions without the depletion of PD-L1-expressing immune cells. Here, we used Nicotiana benthamiana to produce four variants of Durvalumab (DL): wild-type IgG1 and its 'Fc-effector-silent' variant (LALAPG) carrying further modifications to increase antibody half-life (YTE); IgG4S228P and its variant (PVA) with Fc mutations to decrease binding to FcγRI. In addition, DL variants were produced with two distinct glycosylation profiles: afucosylated and decorated with α1,6-core fucose. Plant-derived DL variants were compared to the therapeutic antibody regarding their ability to (i) bind to PD-L1, (ii) block PD-1/PD-L1 inhibitory signalling and (iii) engage with the neonatal Fc receptor (FcRn) and various Fcγ receptors. It was found that plant-derived DL variants bind to recombinant PD-L1 and to PD-L1 expressed in gastrointestinal cancer cells and are able to effectively block its interaction with PD-1 on T cells, thereby enhancing their activation. Furthermore, we show a positive impact of Fc amino acid mutations and core fucosylation on DL's therapeutic potential. Compared to Imfinzi®, DL-IgG1 (LALAPG) and DL-IgG4 (PVA)S228P show lower affinity to CD32B inhibitory receptor which can be therapeutically favourable. Importantly, DL-IgG1 (LALAPG) also shows enhanced binding to FcRn, a key determinant of serum half-life of IgGs.

The tobacco GNTI stem region harbors a strong motif for homomeric protein complex formation.

Introduction: The Golgi apparatus of plants is the central cellular organelle for glycan processing and polysaccharide biosynthesis. These essential processes are catalyzed by a large number of Golgi-resident glycosyltransferases and glycosidases whose organization within the Golgi is still poorly understood. Methods: Here, we examined the role of the stem region of the cis/medial Golgi enzyme N-acetylglucosaminyltransferase I (GNTI) in homomeric complex formation in the Golgi of Nicotiana benthamiana using biochemical approaches and confocal microscopy. Results: Transient expression of the N-terminal cytoplasmic, transmembrane, and stem (CTS) regions of GNTI leads to a block in N-glycan processing on a co-expressed recombinant glycoprotein. Overexpression of the CTS region from Golgi α-mannosidase I, which can form in planta complexes with GNTI, results in a similar block in N-glycan processing, while GNTI with altered subcellular localization or N-glycan processing enzymes located further downstream in the Golgi did not affect complex N-glycan processing. The GNTI-CTS-dependent alteration in N-glycan processing is caused by a specific nine-amino acid sequence motif in the stem that is required for efficient GNTI-GNTI interaction. Discussion: Taken together, we have identified a conserved motif in the stem region of the key N-glycan processing enzyme GNTI. We propose that the identified sequence motif in the GNTI stem region acts as a dominant negative motif that can be used in transient glycoengineering approaches to produce recombinant glycoproteins with predominantly mannosidic N-glycans.

The ligand-dependent suppression of TCR signaling by the immune checkpoint receptor LAG3 depends on the cytoplasmic RRFSALE motif.

Lymphocyte activation gene 3 (LAG3) is an inhibitory immune checkpoint receptor that restrains autoimmune and antitumor responses, but its evolutionarily conserved cytoplasmic tail lacks classical inhibitory motifs. Major histocompatibility complex class II (MHC class II) is an established LAG3 ligand, and fibrinogen-like protein 1 (FGL1), lymph node sinusoidal endothelial cell C-type lectin (LSECtin), and Galectin-3 have been proposed as alternative binding partners that play important roles in LAG3 function. Here, we used a fluorescent human T cell reporter system to study the function of LAG3. We found that LAG3 reduced the response to T cell receptor stimulation in the presence of MHC class II molecules to a lesser extent compared with the receptor programmed cell death protein 1. Analysis of deletion mutants demonstrated that the RRFSALE motif in the cytoplasmic tail of LAG3 was necessary and sufficient for LAG3-mediated inhibition. In this system, FGL1, but not LSECtin or Galectin-3, acted as a LAG3 ligand that weakly induced inhibition. LAG3-blocking antibodies attenuated LAG3-mediated inhibition in our reporter cells and enhanced reporter cell activation even in the absence of LAG3 ligands, indicating that they could potentially enhance T cell responses independently of their blocking effect.

Engineered soluble, trimerized 4-1BBL variants as potent immunomodulatory agents.

Targeting co-stimulatory receptors promotes the activation and effector functions of anti-tumor lymphocytes. 4-1BB (CD137/TNFSF9), a member of the tumor necrosis factor receptor superfamily (TNFR-SF), is a potent co-stimulatory receptor that plays a prominent role in augmenting effector functions of CD8+ T cells, but also CD4+ T cells and NK cells. Agonistic antibodies against 4-1BB have entered clinical trials and shown signs of therapeutic efficacy. Here, we have used a T cell reporter system to evaluate various formats of 4-1BBL regarding their capacity to functionally engage its receptor. We found that a secreted 4-1BBL ectodomain harboring a trimerization domain derived from human collagen (s4-1BBL-TriXVIII) is a strong inducer of 4-1BB co-stimulation. Similar to the 4-1BB agonistic antibody urelumab, s4-1BBL-TriXVIII is very potent in inducing CD8+ and CD4+ T cell proliferation. We provide first evidence that s4-1BBL-TriXVIII can be used as an effective immunomodulatory payload in therapeutic viral vectors. Oncolytic measles viruses encoding s4-1BBL-TriXVIII significantly reduced tumor burden in a CD34+ humanized mouse model, whereas measles viruses lacking s4-1BBL-TriXVIII were not effective. Natural soluble 4-1BB ligand harboring a trimerization domain might have utility in tumor therapy especially when delivered to tumor tissue as systemic administration might induce liver toxicity.

In planta deglycosylation improves the SARS-CoV-2 neutralization activity of recombinant ACE2-Fc.

SARS-CoV-2 infects human cells via binding of the viral spike glycoprotein to its main cellular receptor, angiotensin-converting enzyme 2 (ACE2). The spike protein-ACE2 receptor interaction is therefore a major target for the development of therapeutic or prophylactic drugs to combat coronavirus infections. Various engineered soluble ACE2 variants (decoys) have been designed and shown to exhibit virus neutralization capacity in cell-based assays and in vivo models. Human ACE2 is heavily glycosylated and some of its glycans impair binding to the SARS-CoV-2 spike protein. Therefore, glycan-engineered recombinant soluble ACE2 variants might display enhanced virus-neutralization potencies. Here, we transiently co-expressed the extracellular domain of ACE2 fused to human Fc (ACE2-Fc) with a bacterial endoglycosidase in Nicotiana benthamiana to produce ACE2-Fc decorated with N-glycans consisting of single GlcNAc residues. The endoglycosidase was targeted to the Golgi apparatus with the intention to avoid any interference of glycan removal with concomitant ACE2-Fc protein folding and quality control in the endoplasmic reticulum. The in vivo deglycosylated ACE2-Fc carrying single GlcNAc residues displayed increased affinity to the receptor-binding domain (RBD) of SARS-CoV-2 as well as improved virus neutralization activity and thus is a promising drug candidate to block coronavirus infection.

Assessment of transient expression strategies to sialylate recombinant proteins in N. benthamiana.

There is a demand for increasing the current manufacturing capacities for recombinant protein-based drugs. Novel expression systems such as plants are being explored as faster, more flexible, and possibly cheaper platforms. Many of these therapeutic proteins are glycosylated and require terminal sialylation to attain full biological activity. In planta protein sialylation has been achieved by the introduction of an entire mammalian biosynthetic pathway in Nicotiana benthamiana, comprising the coordinated expression of the genes for (i) biosynthesis, (ii) activation, (iii) transport, and (iv) transfer of Neu5Ac to terminal galactose. Here we address technical issues that can compromise the efficacy of protein sialylation and how they can be overcome. We used the same reporter protein to compared three strategies to transiently deliver the sialylation pathway-genes evaluating efficacy, heterogeneity and batch-to-batch consistency. In addition, we assess the ability of the single-step method to sialylated additional recombinant proteins with different complexity and number of glycosylation sites. Finally, we show that efficient protein sialylation can be up-scaled for large-scale production of sialylated proteins in plants.

Reteplase Fc-fusions produced in N. benthamiana are able to dissolve blood clots ex vivo.

Thrombolytic and fibrinolytic therapies are effective treatments to dissolve blood clots in stroke therapy. Thrombolytic drugs activate plasminogen to its cleaved form plasmin, a proteolytic enzyme that breaks the crosslinks between fibrin molecules. The FDA-approved human tissue plasminogen activator Reteplase (rPA) is a non-glycosylated protein produced in E. coli. rPA is a deletion mutant of the wild-type Alteplase that benefits from an extended plasma half-life, reduced fibrin specificity and the ability to better penetrate into blood clots. Different methods have been proposed to improve the production of rPA. Here we show for the first time the transient expression in Nicotiana benthamiana of rPA fused to the immunoglobulin fragment crystallizable (Fc) domain on an IgG1, a strategy commonly used to improve the stability of therapeutic proteins. Despite our success on the expression and purification of dimeric rPA-Fc fusions, protein instability results in high amounts of Fc-derived degradation products. We hypothesize that the "Y"- shape of dimeric Fc fusions cause steric hindrance between protein domains and leads to physical instability. Indeed, mutations of critical residues in the Fc dimerization interface allowed the expression of fully stable rPA monomeric Fc-fusions. The ability of rPA-Fc to convert plasminogen into plasmin was demonstrated by plasminogen zymography and clot lysis assay shows that rPA-Fc is able to dissolve blood clots ex vivo. Finally, we addressed concerns with the plant-specific glycosylation by modulating rPA-Fc glycosylation towards serum-like structures including α2,6-sialylated and α1,6-core fucosylated N-glycans completely devoid of plant core fucose and xylose residues.

The Instability of Dimeric Fc-Fusions Expressed in Plants Can Be Solved by Monomeric Fc Technology.

The potential therapeutic value of many proteins is ultimately limited by their rapid in vivo clearance. One strategy to limit clearance by metabolism and excretion, and improving the stability of therapeutic proteins, is their fusion to the immunoglobulin fragment crystallizable region (Fc). The Fc region plays multiple roles in (i) dimerization for the formation of "Y"-shaped structure of Ig, (ii) Fc-mediated effector functions, (iii) extension of serum half-life, and (iv) a cost-effective purification tag. Plants and in particular Nicotiana benthamiana have proven to be suitable expression platforms for several recombinant therapeutic proteins. Despite the enormous success of their use for the production of full-length monoclonal antibodies, the expression of Fc-fused therapeutic proteins in plants has shown limitations. Many Fc-fusion proteins expressed in plants show different degrees of instability resulting in high amounts of Fc-derived degradation products. To address this issue, we used erythropoietin (EPO) as a reporter protein and evaluated the efforts to enhance the expression of full-length EPO-Fc targeted to the apoplast of N. benthamiana. Our results show that the instability of the fusion protein is independent from the Fc origin or IgG subclass and from the peptide sequence used to link the two domains. We also show that a similar instability occurs upon the expression of individual heavy chains of monoclonal antibodies and ScFv-Fc that mimic the "Y"-shape of antibodies but lack the light chain. We propose that in this configuration, steric hindrance between the protein domains leads to physical instability. Indeed, mutations of critical residues located on the Fc dimerization interface allowed the expression of fully stable EPO monomeric Fc-fusion proteins. We discuss the limitations of Fc-fusion technology in N. benthamiana transient expression systems and suggest strategies to optimize the Fc-based scaffolds on their folding and aggregation resistance in order to improve the stability.

Generation of enzymatically competent SARS-CoV-2 decoy receptor ACE2-Fc in glycoengineered Nicotiana benthamiana.

Human angiotensin-converting enzyme 2 (ACE2) is the primary host cell receptor for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) binding and cell entry. Administration of high concentrations of soluble ACE2 can be utilized as a decoy to block the interaction of the virus with cellular ACE2 receptors and potentially be used as a strategy for treatment or prevention of coronavirus disease 2019. Human ACE2 is heavily glycosylated and its glycans impact on binding to the SARS-CoV-2 spike protein and virus infectivity. Here, we describe the production of a recombinant soluble ACE2-fragment crystallizable (Fc) variant in glycoengineered Nicotiana benthamiana. Our data reveal that the produced dimeric ACE2-Fc variant is glycosylated with mainly complex human-type N-glycans and functional with regard to enzyme activity, affinity to the SARS-CoV-2 receptor-binding domain, and wild-type virus neutralization.