Contrasts between White’s proposed applications, and contemporary notion of vegetated architecture as a sustainable technology is especially salient, as language to describe sustainability or environmental performance had not yet been established. The field of vegetation-bearing architecture was rapidly expanded by two contingent patents filed on 4th April and 28th May 1938 by William M. MacPherson and Elmer Hovendon Gates, respectively. These patents reinterpret the structural system proposed by White in which the vegetation bearing units assemble like masonry or bricks, and are essentially self supporting through stacking and repetition . Consecutive to this, patents by Gates and MacPherson envision vegetation-bearing units supported within a structural and load-bearing framework, and vegetation bearing units secured to a self-supporting wall as a veneer or surface system. Although the structural iterations presented by each patent alter the load bearing relationship of the vegetation to the structural element, the new art, as described by White, remains remarkably resilient to these adaptations and contemporary permutations.Although the structural modifications proposed by Gates and MacPherson reorient the relation of the vegetation-bearing unit to the underlying structural system, the basic technology and principles remains intact. The resilience of White’s new art is found in its aptitude for reinterpretation, as he defines a relationship between plants and structure that elucidates their interdependence and interstices. Triangulations between the building sciences, horticultural arts,vertical grow and landscape theory are seminal to White’s conception of the Vegetation Bearing Architectonic Structure and System.

White received a Bachelor of Science in agriculture from Cornell in 1912, his Master of Landscape Architecture from Harvard in 1915.His background in science and design undoubtedly contributed to a synthetic approach to plants, structure, and garden form, yet it is a convergence of material innovations and new scientific discoveries that make the vertical garden possible, and timely, in the 1930s. Innovation in building materials occurred rapidly in early years of the twentieth century, as industrialization and wartime research fueled experiments in architecture and the sciences alike. Material Sciences flourished during World War I and material product lines expanded through the roaring twenties, radically impacting architecture and the production of buildings.New building materials were readily tested, and exhaustive volumes of literature were published to disseminate information in an ever-changing marketplace for architects and designers. Building material bibliographies from the era document rapid integration and research on materials such as armorphy, clay tile, plywood, glass block, masonite, rostone, vinylite, celotex, porcelain, zonolite, to name just a few.Concurrent to the expansion of new materials in architecture, was an expansion of the role of popular science and a belief that technology would improve modern living. In agriculture, new technologies such as hydroponics, popularized in the 1930s, promised to create new productive systems that would increase yields and feed a growing population with increased efficiency. White’s invention exists at the intersection of these spheres of innovation, where the differentiation between horticultural system and building system begins to blur. Efforts towards new and efficient forms of building assembly paralleled new and efficient forms of agriculture.Plant-based pharmaceutical production is appealing given its inexpensive facility and production cost, linear scale-up, the absence of animal pathogens, and capability to produce complex proteins and perform post-translational modifications, which overcomes one or more drawbacks of traditional recombinant protein expression systems such as animal cell culture and bacterial fermentation.

Much work has been carried out using stably transformed plants, but the significantly reduced development and production timeline makes transient expression of proteins in whole plants a particularly attractive option, cutting the time to bring critical medications to the market during a pandemic. Vacuum agroinfiltration is the most widely used method for uniformly introducing agrobacterium harboring an expression cassette containing a gene of interest into plant tissue given its natural ability to transfer T-DNA into plant cells, which is ideal for transient protein production in plants. Although a plant-based recombinant protein production system provides distinct advantages over traditional systems, the differences between N-glycosylation of proteins produced in plants and humans could limit the use of plant systems for the production of glycoprotein-based pharmaceuticals. In higher eukaryotes, the initial steps of N-glycosylation processing are well conserved between plants and human, resulting in oligomannose-type N-glycosylation. However, late N-glycosylation maturation in the Golgi apparatus is kingdom-specific, and thus results in different N-glycosylation on proteins produced in plants compared with human. These plant-specific glycans may lead to potential safety issues such as hypersensitivity or allergy, as plant-specific α-fucose and β-xylose are known to be important IgE binding determinants of plant allergens. Thus, if these plant-specific glycans are present in an injected pharmaceutical product, the glycoprotein could trigger immunological response, or at least, result in a short circulation half-life. N-glycans of proteins produced from mammals are often terminated in β-galactose and sialic acid; sialic acid is particularly important as it typically increases the circulation half-life of proteins.

There are a number of ways to avoid incorporating plant-specific glycans in the product such as adding a signal sequence at the C-terminus of the target protein to retain it within ER, or RNAi-mediated knock-down of α-fucose and β-xylose. These methods require modification to either the protein sequence or to the expression system, which can potentially affect protein structure and require a long developmental time. As an alternative, the use of small molecule inhibitors of intracellular glycosidases is a highly flexible bio-processing approach for controlling protein N-glycosylation patterns in transient agroinfiltration processes, and it is the approach investigated in this study. Here, we report an easy and fast way to modify N-glycosylation of recombinant proteins produced transiently in N. benthamiana through the addition of kifunensine in the agrobacterium suspension prior to vacuum agroinfiltration, which avoids modification to protein sequence or expression system while producing recombinant protein with oligomannose-type N-glycans that are similar between plant and human. Oligomannose-type N-glycan is preferred for the HIV-1 viral vaccine development as a vast majority of broad and potent neutralizing antibody responses during HIV-1 infection target mannose-glycan-dependent epitopes. In addition, monoclonal antibodies with oligomannose N-glycans show increased ADCC activity and affinity for FcγRIIIA. Protein N-glycosylation starts in the endoplasmic reticulum , where N-glycan precursors Glc3Man9GlcNAc2 are first synthesized, followed by the removal of terminal Glc residues, resulting in Man9GlcNAc2 structures. Then, a single α1,2 linked mannose is removed by ER class I α-mannosidase, producing Man8GlcNAc2 structures. The trimming of α1-2 mannose residues continues with the action of Golgi class I α-mannosidases in cis-Golgi to give Man5 structures. Kifunensine is a highly selective inhibitor of class I α-mannosidases in both plants and animals, and it has been used in cell cultures to produce recombinant proteins with oligomannose-type N-glycans. Although the general effects of kifunensine and other alkaloid-like processing glycosidases inhibitors are well understood, for the most part this information comes from cell culture system studies. Meanwhile, the study of kifunensine on whole-plant transient protein expression through agroinfiltration is new. There are only two published papers on whole-plant kifunensine treatment,farming vertical where kifunensine was supplied hydroponically throughout the whole incubation period, which requires larger quantities of kifunensine, a more expensive hydroponic system and constant monitoring especially at large scale as compared to our method. In addition, it was also shown that hydroponic kifunensine treatment resulted in dramatic decrease of protein expression level which was not observed with our method. In this study, Fc-fused capillary morphogenesis gene-2 , an anthrax decoy protein, served as a model protein, which contains single N-glycosylation site within its Fc domain . CMG2-Fc is a potent anthrax decoy protein as shown previously, where the CMG2 domain binds to anthrax protective antigen and prevents the anthrax toxin from entering the cell. Meanwhile, the presence of Fc domain significantly increases the serum half-life, which prolongs therapeutic activity owing the slower renal clearance for larger sized molecules and interaction with the salvage neonatal Fc-receptor. CMG2-Fc thus can be used as potent anthrax therapeutic and prophylactic without frequent redosing. The expression levels of CMG2-Fc produced transiently in wild type N. benthamiana under kifunensine treated and untreated conditions were measured with a sandwich ELISA, and protein N-glycosylation profiles were evaluated with mass spectrometry for kifunensine treated and untreated conditions.

The findings in this study can be applied for N-glycosylation modification of other plant recombinant proteins when oligomannose-type N-glycans lacking core fucose are preferred, without the need to modify protein sequence and/or subcellular targeting. The CMG2-Fc expression levels in crude leaf extract were quantified through a sandwich ELISA to confirm the expression of CMG2-Fc, and to evaluate the effect of kifunensine on protein expression. The ELISA relies on binding of CMG2-Fc through the Fc region to protein A coated on a 96-well plate. A secondary anti-Fc polyclonal antibody linked to a horseradish peroxidase enzyme binds to the CMG2-Fc allowing colorimetric detection. The potential interference of plant host cell proteins and nonspecific binding were determined to be negligible. Twenty wild-type 5–6-week old N. benthamiana plants were divided equally into experimental and control groups, agro-infiltrated and incubated for 6 days, then whole leaves were extracted under identical conditions to determine protein expression. Kifunensine at a concentration of 5.4 µM was included in the agrobacterium suspension in the Kifunensine group. This kifunensine concentration was chosen as a starting point by taking the average of concentrations used in a previous CHO cell culture study, as no reference concentration is available for vacuum infiltration of kifunensine. The average mass of CMG2-Fc per kg leaf fresh weight was 717 and 874 mg/kg leaf FW for the Kifunensine and Kifunensine samples, respectively .These data suggest that the addition of kifunensine in the agro-infiltration process was not detrimental to transient protein production, and in this case, it resulted in a 22% increase in CMG2-Fc yield. This allows the use of kifunensine for modification of glycosylation profiles without compromising protein expression. Total soluble protein content of whole leaf extract of Kifunensine and Kifunensine groups were similar as shown in Figure 1, which indicates that kifunensine does not have significant impact on plant protein synthesis in general.In this study, the influences of one-time kifunensine vacuum infiltration on the expression level, N-glycan profile of a recombinant protein, namely CMG2-Fc, produced transiently in N. benthamiana plants were evaluated in both whole-leaf extract and AWF. We found that kifunensine had a positive impact on protein production when supplied in the agroinfiltration solution; specifically, we observed a 22% increase of protein expression with kifunensine treatment condition, presumably owing to its suppressing effects on ER-associated degradation pathway. This finding is consistent with previous observations in multiple mammalian cell culture systems, and there is no reason to suspect that this will not be the case for other eukaryotic systems, including plant systems. Plants were monitored visually throughout the incubation period, and there were no significant phenotypical differences between kifunensine-treated and control groups. In the case of a whole-plant study, Roychowdhury et al. have shown that the yield of recombinant cholera toxin B subunit dropped by 30% and 75% when kifunensine was supplied at 5 µM hydroponically for 3 days and 5 days post agroinfiltration, respectively. In contrast, we observed slight increase in protein yield when kifunensine was infiltrated in leaf tissue instead of being supplied hydroponically. Thus, the lower protein yield they observed may have resulted from continued application in the hydroponic solution, and it is eliminated when kifunensine was supplied directly to leaf tissue through vacuum infiltration. In addition, in the hydroponic study, the target protein was retained in ER, while the model protein in our study and other cell culture studies were targeted for secretion. This difference in protein targeting may also play a role in protein yield changes upon kifunensine treatment. Since kifunensine targets class I α-mannosidases and N-glycosylation is an enzyme-directed site specific process, it was expected that the kifunensine treated plants will produce CMG2-Fc with predominantly oligomannose-type N-glycans. It was not obvious, however, that one-time vacuum infiltration provided enough kifunensine and it remained active for the whole protein production period, resulting in a complete shift of N-glycan profile from plant-specific complex-type to oligomannose-type. This N-glycan profile shift suggested that either kifunensine remains active throughout the entire incubation/production period, or that the plants only express protein for a briefperiod during which the kifunensine is active.