香京julia种子在线播放

Surveying missions to Mars, like the InSight lander¹ (Overview | Mission--NASA’s InSight Mars Lander) launched in 2018 and Perseverance rover² in 2020, directly support the objectives of NASA’s long-term Mars Exploration Program³ : an effort to explore the potential for life on Mars and prepare for human exploration of Mars. The maturation of the program requires redefining the risks to human health as mission architectures transition from the current ‘‘Earth Reliant’’ paradigm used on the International Space Station (ISS) to the cislunar space ‘‘Proving Grounds’’ and finally to deep-space ‘‘Earth Independent’’ mission architectures, as defined in NASA’s report titled, ‘‘Journey to Mars: Pioneering Next Steps in Space Exploration.’’⁴

Human missions to Mars will be ‘‘Earth Independent,’’ meaning there will be very limited emergency evacuation and re-supply capabilities along with substantially delayed communications with the Earth-based mission team. The NASA Human Research Roadmap⁵ currently rates most human health risks, which include “risk of adverse health outcomes and decrements in performance due to inflight medical conditions” and “risk of ineffective or toxic medications during long-duration exploration spaceflight,” as either medium or high risk for a Mars planetary visit/habitat mission. Risk ratings are based on failure mode and effects analysis and on hazard analysis using dimensions of severity, occurrence, and detectability. A recent review highlights the current understanding of the primary hazards and health risks posed by deep space exploration as well as the six types of countermeasures: protective shielding, biological and environmental temporal monitoring, specialized workout equipment, cognition and psychological evaluations, autonomous health support, and personalized medicine (Afshinnekoo et al., 2020).

Of these countermeasures, it could be argued that medicine is the most crucial and least advanced toward mitigating space health hazards. There is very limited information on, and few direct studies of, pharmaceutical usage, stability, and therapeutic efficacy (i.e., pharmacokinetics, pharmacodynamics) in spaceflight or in a Mars surface environment (Blue et al., 2019). Furthermore, flown stores of pharmaceuticals face two additional barriers: (1) radiation-accelerated degradation (Du et al., 2011), and (2) addressing a myriad of low occurrence and high impact health hazards without the ability to fly and maintain potency of therapeutics for all of them. In these circumstances, it is often more beneficial to build robustness to these low occurrence health hazards rather than to try to predict them. It is therefore imperative that on-planet and/or in-flight pharmaceutical production be developed to bridge this risk gap. These pharmaceutical foundry technologies will supplement, not replace, the flown pharmaceutical formulary designed to treat anticipated medical threats during space missions.

Pharmaceuticals are produced either chemically or biologically. A recent review of pharmaceutical production for human life support in space compares these two methods, highlighting the need for biological production in order to address many low occurrence and high impact health hazards (e.g., sepsis, ear infection, and glaucoma) and further comparing different biological production systems (McNulty et al., 2021). One major advantage of biological production is the efficiency in transporting and synthesizing genetic information as the set of instructions, or sometimes the product itself, to meet the therapeutic needs for a variety of disease states. The emerging field of Space Systems Bioengineering (Berliner et al., 2020) encapsulates this need for biological production, of which pharmaceuticals is identified as an important subset.

Biopharmaceuticals must be purified after accumulation with the biological host organism, or cell-free transcription-translation reaction, in order to meet requirements for drug delivery and therapeutic effect (Harrison et al., 2015). The majority of commercial biopharmaceutical products are administered via intravenous and subcutaneous injection (Škalko-Basnet, 2014). Biopharmaceutical formulations for injection requires high purity (>95%) product, as impurities introduced directly into the bloodstream can trigger significant immune responses and reduce efficacy (Haile et al., 2015).

Downstream processing of biopharmaceuticals is therefore usually a resource-intensive section of overall processing, being cited as high as 80% of production costs (and contributions of input mass) for monoclonal antibody (mAb) therapeutics produced using mammalian cell cultures (Rathore et al., 2015; Budzinski et al., 2019). In addition to the processing burden for biopharmaceutical injectables, there are also often substantial storage costs involving complex supply chain and storage management with stability requirements for factors including temperature, time, humidity, light, and vibration (Sykes, 2018). There are several approaches being pursued to overcome the challenges and costs associated with downstream processing and formulation.

First are the tremendous efforts in process intensification (Strube et al., 2018). While the highly sensitive nature of biopharmaceuticals to minor process changes has introduced barriers and complexities to innovation through process intensification that have not been realized in non-healthcare biotechnological industries, there have been significant strides made in the past decade in the areas of process integration (Steinebach et al., 2017), automation (Pollard et al., 2017), and miniaturization (Adiga et al., 2018; Crowell et al., 2018).

Another route that researchers are pursuing to reduce downstream processing costs and resources is a biological solution to processing technology. In the same vein that the biopharmaceutical industry sprung out of researchers leveraging the power of biology to produce therapeutically relevant molecules that were inaccessible or excessively costly by means of chemical synthesis, researchers are now also trying to apply that same principle to purifying therapeutically relevant molecules. The simplicity of production, reagents that can be produced using self-replicating organisms, and potential recyclability of spent consumables are significant advantages of biological purification technology for space or other limited resource applications. Examples of primary biological technologies include fusion tags (Bell et al., 2013), stimuli-responsive biopolymers (Sheth et al., 2014), hydrophobic nanoparticles (Jugler et al., 2020), and plant virus nanoparticles (Werner et al., 2006; Uhde-Holzem et al., 2016).

Lastly, there are vast efforts to establish alternative drug delivery modalities (Anselmo et al., 2018). Other modalities that do not require injection and which might be more compatible to administration in limited resource environments, such as oral consumption, nasal spray, inhalation, and topical application, have long presented challenges in biopharmaceutical stability (e.g., denaturation in stomach acid) and delivery to the active site (e.g., passing the gut-blood barrier) that minimize product efficacy and necessitate costly advanced formulations and chemistries (Mitragotri et al., 2014).

A particularly promising drug delivery technique to circumvent downstream processing burdens is to sequester the active pharmaceutical ingredient in the host cells of the upstream production system as a protective encapsulation in order to facilitate bioavailability through oral delivery (Kwon and Daniell, 2015). It represents an opportunity to greatly lower the cost of in situ production of human medicine for a space mission. This technique presumes that the host system is safe for human consumption, and so naturally lends itself to utility in systems such as yeast and plant production hosts. Oral delivery via host cell encapsulation has been recently established as commercial drug delivery modality with the US Food and Drug Administration approval of Palforzia as an oral peanut-protein immunotherapy (Vickery et al., 2018). However, this solution is not necessarily amenable to the diversity of pharmaceutical countermeasures that may be required, especially for unanticipated needs in which the product may not have been evaluated for oral bioavailability.

In 2011, the space shuttle program was retired due to increasing costs, demonstrating that reduction of economic cost is critical for sustaining any campaign of human exploration (Wall, 2011). Although recent efforts in reducing the launch cost to low earth orbit by commercial space companies have aided in the redefinition of the space economy (Whealan George, 2019), the barrier to longer term missions, such as a journey to Mars, is still limited by the extreme financial cost in transporting resources. Additionally, it has been shown that as the mission duration and complexity increases–as expected for a human mission to Mars–the quantity of supplies required to maintain crew health also increases (Anderson et al., 2018). In the case of meeting the demand for medication, biopharmaceutical synthesis has been proposed as an alternative to packaging a growing number of different medications (Menezes et al., 2015; McNulty et al., 2021). Assuming that both technologies can meet mission demand, selection of the production-based biotechnology platform will be dependent on its cost impact. It is therefore critical that the cost model of biopharmaceutical synthesis accounts for and minimizes the cost of any and all subprocesses, including those for purification.

The current terrestrial biopharmaceutical synthesis cost model does not align with the needs for space exploration environments. For example, the literature highlights the high cost of Protein A affinity chromatography resin ($8,000–$15,000/L) and the need to reduce the price (Ramos-de-la-Peña et al., 2019). However, the purchase cost of chromatography resin is not nearly as critical in space environment applications where the major costs are more closely tied to the physical properties of the object (mass, volume, refrigeration requirements, etc.), as a result of fuel and payload limitations and the crew time required for operation (Jones, 2001). The distinct cost models of space and terrestrial biopharmaceutical production may increase the burden of identifying space-relevant processing technologies and may also limit direct transferability of terrestrial technologies without attention given to these areas.

On the other hand, changing incentives structures relating to sustainability and the advent of new platform technologies are rapidly increasing alignment and the potential for technology crossover. For example, companies like On Demand Pharmaceuticals⁶, EQRx⁷, and the kenUP Foundation⁸, initiatives leading to industry adoption of environmental footprint metrics such as E-factor (Sheldon, 2007) and process mass intensity (PMI) (Budzinski et al., 2019), and diffusion from the adjacencies of green and white biotechnology (Tylecote, 2019) all promote development of accessible and sustainable technologies. As these trends pertain to space-relevant processes, these examples can also be viewed as driving more closed loop systems composed of simpler components.

Techno-economic evaluations are performed using the recently proposed equations for ESM that include calculation of costs at each mission segment (Berliner et al., 2021). ESM for the mission ESM₀ is defined as

where M_i, V_i,P_i,C_i, T_i are the initial mass [kg], volume [m³], power requirement [kW], cooling requirement [kg/kW], and crew-time requirement [CM-h/h], M_eq, V_eq,P_eq,C_eq, T_eq are the equivalency factors for mass [kg/kg] (which is set to 1 in this study), volume [kg/m³], power [kg/kW], cooling [kg/kW], and crew time [kg/CM-h], respectively, L_eq is the location equivalency factor [kg/kg] that accounts for costs associated with mass transport occurring at a particular mission segment (e.g., orbital maneuvers required for the return transit), and D is the duration of the mission segment [day] over a set of subsystems i ∈ A and set of mission segments k ∈ℳ. The mission ESM in this study is specifically defined as the sum of subtotal ESM for each mission segment within the scope of the RMA defined in this study (pre-deployment ESM_0,pd, crewed transit to Mars ESM_0,tr1, Mars surface operations ESM_0,sf, and return crewed transit to Earth ESM_0,tr2).

Key mission and pharmaceutical assumptions are summarized in Table 2. The mission timeline depicted in Figure 2 provides insight into the proposed RMA and downstream crew needs and mAb production horizon. Here we assume a total mission duration of 910 days. First, a crew of 6 will travel from Earth to low Earth orbit, then board an interplanetary craft for a 210-day journey to Martian orbit, where the crew will descend to the surface in a separate craft, allowing the large transit vehicle to remain in orbit. Once on Mars, the crew will perform surface operations for 600 days. Following surface operations, the crew will leave Mars in a fueled ascent craft, board the interplanetary vehicle, and return to Earth orbit in 200 days. The mission timeline, crew size, and ESM equivalency factors are consistent with the recent RMA presented for inclusion of biomanufacturing elements (Berliner et al., 2021).

The mission demand for mAb therapies is assumed to be 30,000 mg over the entirety of the mission (supporting logic detailed in Supplementary Table 1). Pharmaceutical stores and production resources are assumed to be flown with the crew transit (no pre-deployment in order to maximize shelf-life). We assume that the production resources are stable throughout the mission duration. We conservatively assume (in the face of insufficient spaceflight stability data for biologics for a more refined estimate) that the first 600 days of pharmaceutical demand will be met through flown stores (20,000 mg), at which point pharmaceutical ISRU manufacturing is needed (10,000 mg) to alleviate the impact of accelerated pharmaceutical degradation and provide supplementary medication. The pharmaceutical production window opens prior to the ISRU demand timeframe and persists through a portion of the return transit (up to mission day 810) to reflect the expected life support advantage of maintaining capabilities to counter unanticipated needs or threats. We assume that the Protein A-based unit procedures consistently yield 98% recovery of mAb from the input stream.

Deterministic models for each unit procedure were developed in Microsoft Excel (Supplementary Data Sheet 2) using reference protocols cited in Table 1 as a series of executable operations, each containing a set of inputs defined by cost categories (labor, equipment, raw materials, and consumables) that are correspondingly populated with characteristic ESM constituent (mass, volume, power, cooling, and labor time) values (model composition illustrated in Supplementary Figure 1). Unit procedures have been defined as the smallest single execution (i.e., unit) of the secondary purification capture step procedure according to the reference protocol. We define the unit capacity by volume according to the equipment and consumables used (e.g., 2 mL maximum working volume in a 2 mL tube) and by mAb quantity according to the binding capacity for the given method (e.g., 1 mg mAb/mL resin) (Supplementary Table 2). Unit procedures with no explicit working volume constraints (i.e., the liquid solution volume for biotic technologies) have been defined with a maximum unit volume of 2 mL. ESM-relevant characteristics of individual inputs (e.g., equilibration buffer, 2 mL tube) are defined based on publicly available values, direct measurements taken, and assumptions (which are explicitly identified in the Supplementary Data Sheet 2).

There are several model features that we have considered and decided not to include within the scope of analysis. Packing and containers for the inputs are not included for three reasons: (1) the contributions of the container are considered negligible as compared to the input itself (e.g., container holding 1 L buffer as compared to the 1 L of liquid buffer); (2) materials flown to space are often re-packaged with special considerations (Wotring, 2018); and (3) the selection of optimal container size is non-trivial and may risk obscuring more relevant ESM findings if not chosen carefully. We do not consider buffer preparation and assume the use of flown ready-to-use buffers and solutions. Furthermore, refrigeration costs of the input materials and costs that may be associated with establishing and maintaining a sterile operating environment (e.g., biosafety cabinet, 70% ethanol in spray bottles) are expected to be comparable between unit procedures and not considered. Impacts of microgravity on unit procedure execution are not considered for the return transit production. Refrigeration costs associated with low temperature equipment operation (e.g., centrifugation at 4°C) are included in the equipment power costs.

Inputs common across unit procedures are standardized (Supplementary Table 3). One operational standardization is the inclusion of pH neutralization of the product stream following the low pH elution mechanism, which was explicitly stated in some procedures while not in others. Input quantities are scaled from a single unit to determine the number of units required to meet the RMA specifications. The ESM constituent inputs (mass, volume, power, cooling, and labor time) are converted into equivalent mass values using RMA equivalency factors (Supplementary Table 4).

Given the limited granularity of the presented RMA, which was scoped as such to reflect the lack of literature presenting an overarching and validated Concept of Operations for a Transit to Mars (Antonsen et al., 2017), we do not define strict manufacturing scheduling criteria for pharmaceutical production. Construction of a detailed pharmaceutical production RMA is hindered by uncertainty in the number and identity of mAb therapy products that would be included within mission scope, the decay rate of mAb therapy stores in the mission environments, and a reasonable basis for building robustness to unanticipated disease states. Rather, we choose to establish an objective comparison between unit procedures by normalizing for scheduling-associated manufacturing efficiencies. We accomplish this by first identifying the number of batches per mission (and thus batch size) needed to meet the mAb demand (base case of 10,204 mg mAb feed assuming 98% recovery) that minimizes the ESM output for a given unit procedure, and then running the simulation of pharmaceutical production at that number of mission batches, as shown in Figure 3A and tabulated in Supplementary Table 5.

In Figures 3B–E, we visualize a deconstruction of ESM output, using the VIN unit procedure as an example, by key performance metrics that vary with a scheduling dependence in order to illustrate the significance of batch optimization in unit procedure comparison. The processing of a given batch volume and mAb quantity is allocated into a number of units, as determined by the volume and mAb quantity constraints of a given unit procedure, and a number of use cycles per batch, as determined by the capacity of the equipment specified in the given unit procedure. We show how the variation in ESM output over the number of mission batches maps to extent of unit vacancy or underutilization (Figure 3B), extent of operational equipment (e.g., centrifuge) vacancy or underutilization (Figure 3C), and number of required use cycles (Figure 3D). We also show an oscillatory behavior in the scheduling (i.e., total mAb purified per mission, % purified at surface operations) that quickly dampens as number of mission batches increases (Figure 3E). This behavior is a result of the assumption that the mAb feed stream is coming from a discrete upstream production batch (e.g., batch-mode bioreactor) that does not output partial batch quantities, as opposed to a continuous upstream production for which there are no defined batches. Accordingly, partial batch needs are met by the processing of a full batch.

The ESM and output metrics of the base case scenario (10,000 mg mAb demand, 1 mg mAb/mL feed concentration, 98% recovery) for each of the six unit procedures are shown in Figures 4A–F. From this viewpoint of an ESM output for an isolated unit procedure outside the context of a full purification scheme, the ESM ranked from lowest to highest are VIN < SPN < OLE < CHM < MAG < OLE. However, we reason that it is more important to understand the model inputs that influence the ESM output rankings than to use the rankings in this isolated subsystem analysis to make technology selection choices, which requires the context of a full pharmaceutical foundry and of linkages to other mission elements.

We observe that mass costs are generally the primary contributor to ESM output, except for the MAG and ELP procedures in which labor time costs are larger. The mass costs are not closely associated to any given cost category across unit procedures, but rather the breakdown of mass costs varies widely by unit procedure.

Power costs (kW) are disproportionately high given that the static nature of ESM assumes constant usage, and thus energy (kW⋅h) in this context (i.e., the power supply to the equipment is not turned off in this analysis). These costs represent an upper bound assuming that the power supply system capacity is sized to support a maximal power consumption in which all power-drawing elements are simultaneously in operation. Time of power usage as a fraction of duration are as follows: CHM (99%) > MAG (78%) > ELP (48%) > SPN (45%) > OLE (42%) > VIN (30%). The lower use fraction unit procedures are therefore paying a relatively higher cost per unit power demand in this current method. The electrical needs of the equipment used by the unit procedures are within NASA-proposed Mars mission RMA bounds, with energy use across all unit procedures would peak at ∼1% of a proposed Mars transfer vehicle electric capacity (50 kWe) or ∼5% of the habitat capacity (12 kWe) of a reference stationary surface nuclear fission power reactor (Drake et al., 2010).

The mission segment breakdown of ESM illustrates the relatively high costs of pharmaceutical manufacturing capabilities for transit, even for the transit to Mars (tr₁) in which there is no actual production taking place. There is a strong economic incentive to limit the amount of supplies flown on tr₁. Alternatives such as the pre-deployment of reagents and consumables and limiting of production to surface operations on Mars (which has lower RMA equivalency factors for mass and volume than transit operations) must be balanced against the risk to human health posed by removing pharmaceutical production capabilities from a mission segment and potentially exposing the supplies to longer storage times that could challenge shelf lives.

Labor and operation times are important parameters in the broader mission and pharmaceutical foundry context. These unit procedures represent a single step of pharmaceutical production, which if realized in a space mission context, would, in turn, need to be a small portion of a crew member’s time allocation. Assuming 40-h work weeks for crew members, the labor time spans a range of ∼1% (CHM) to ∼14% (ELP) of the available crew time over the 600-day production window. It is not feasible to operationalize with such high labor and operation times at this scale of production, particularly as they stand for MAG and ELP. While strategies such as batch staggering and concurrency can be used to reduce durations, advanced automation will almost certainly need to be built into the core of a pharmaceutical foundry.

A prevailing trend throughout the unit procedures is that the number of unit executions and use cycles required by a given unit procedure are positive correlated with the ESM output value, except for the equipment cost-dominant and higher unit capacity CHM procedure. The equipment modeled in the analysis for CHM and the other unit procedures are almost certainly not space-ready and could be further designed to reduce mass and volume and increase automation to reduce crew labor time. The increased equipment costs in the CHM procedure are primarily due to automation and monitoring hardware for running liquid chromatography, which is reflected in the minimal labor costs of the CHM procedure. Miniaturization efforts, such as those focusing on microfluidic systems (Millet et al., 2015; Rodríguez-Ruiz et al., 2018; Murphy et al., 2019), are emerging as a potential path toward mitigating the high equipment costs associated with highly automated and tightly controlled manufacturing, which are crucial for freeing up valuable crew time.

The number of unit executions is determined by the binding capacity of the technology and the nominal unit size. This indicates that the unit capacity for purification is an important consideration and influential factor. Unit sizing is an important consideration that is valuable to assess more holistically within the broader pharmaceutical production and mission context.

The number of use cycles is determined by the number of unit executions required and by the maximal unit capacity of the equipment items (e.g., if you presume that an 18-slot centrifuge is the equipment bottleneck then the effective number of batches is the number of units required divided by 18). Therefore, it can be understood that the equipment unit capacity is a critical parameter in tuning the number of use cycles and, by extension, the labor costs. For processes with lower labor costs, due to the intrinsic nature of the procedure or through automation of labor, equipment unit capacity will still influence the total duration and production throughout. The MAG and ELP procedures yield both high labor and duration times and are thus particularly sensitive to the equipment capacity.

We analyzed the specific ESM output broken down by cost category for the six unit procedures over a range of input stream mAb concentrations (Figure 6) and mission demand for mAb (Figure 7). Specific ESM, termed cost of goods sold in traditional manufacturing analyses, is the ESM output required to produce 1 mg mAb. This is used in the scenario analyses to normalize ESM output across variation in mission demand for mAb. The optimal number of batches per mission was found and used for each unit procedure and scenario tested (Supplementary Tables 7, 8).

We observe the general and expected trends that specific ESM decreases with an increasing feed stream mAb concentration and mission demand. The CHM procedure exhibits notably limited sensitivity to feed stream mAb concentration, which can be attributed to the equipment-dominated cost profile, fixed column size, and nature of the governing reference protocol that does not specify restrictions on sample load volume. Depending on the pre-treatment of the feed stream, it may be more reasonable to impose constraints on the sample load volume. In contrast, the specific ESM output of the CHM procedure is the most sensitive to mission mAb demand with higher demand increasingly offsetting the fixed capital costs. The CHM procedure is also the largest capacity unit modeled in the analysis (i.e., CHM capacity is 30 mg mAb/unit as compared to 2.7 mg mAb/unit for MAG, the next highest capacity unit) and is accordingly expected to scale well with demand.

The SPN, ELP, OLE procedures exhibit behaviors in which the specific ESM output abruptly plateaus with an increasing feed stream mAb concentration. This observation can be attributed to the unit procedure operating in a mAb binding capacity-limited regime (as opposed to volume-limited for more dilute feeds) which also then controls and maintains unit procedure throughput (e.g., the ELP number of units, 37,044, and use cycles per mission, 2,058, is constant at and above 0.35 mg mAb/mL input stream concentration). This can be de-bottlenecked via technology (e.g., improved chemistry of the capture step unit leading to higher binding capacity) or methodology (e.g., increased concentration of the capture step unit leading to higher binding capacity) improvements.

Low demand scenarios are particularly relevant for examination in a space health context, as small capacity redundant and emergency utility is a likely proving ground for inclusion of a space pharmaceutical foundry. At the lower boundary of the tested range (1,000 mg mAb/mission), we see the ESM outputs from lowest to highest are re-ordered as MAG < VIN < SPN < OLE < ELP < CHM. Minimization of equipment costs are particularly important in this regime, and it is observed that, indeed, the ESM output near completely aligned with the ranking of equipment cost (MAG < VIN < SPN < ELP < OLE < CHM). It is likely that other non-ESM factors such as integration with other flown elements will understandably influence the design and composition of early and low capacity flown pharmaceutical foundries.