Proceedings of a Workshop—in Brief

Successes and Challenges in Biomanufacturing

Proceedings of a Workshop—in Brief

INTRODUCTION

The use of living organisms and biological components in manufacturing processes is increasing across manufacturing sectors. However, biomanufacturing faces several bottlenecks and challenges to continued growth. To share practices and potential solutions, the National Academies of Sciences, Engineering, and Medicine hosted a workshop titled Successes and Challenges in Biomanufacturing on October 24-25, 2022.¹ The workshop brought together biomanufacturing stakeholders across industry, academia, and government with expertise across diverse fields, including U.S.-based and international speakers. Discussions spanned the breadth of biomanufacturing contexts and applications, including bioindustrial and biopharmaceutical manufacturing. This Proceedings of a Workshop—in Brief provides the rapporteurs’ high-level summary of the topics addressed at the workshop. It should not be viewed as consensus conclusions or recommendations of the National Academies.

Throughout the workshop, participants discussed in breakout groups the key challenges in biomanufacturing and their takeaways from workshop sessions. Highlights from these breakout discussions are included in the associated resource document² for this proceedings. Some examples and cross-cutting themes—like the importance of coordination and communication—were discussed by participants in the context of multiple topics (e.g., workforce development and biomanufacturing scale-up capacity) and were discussed as relevant to multiple sectors of biomanufacturing.

Workshop Context and Opportunities in Biomanufacturing

In kicking off the workshop, Kristala Prather (Massachusetts Institute of Technology [MIT]) noted that leveraging the power of biology and biomanufacturing may enable society to develop solutions to the serious issues affecting planetary health and the health of all who inhabit the planet. She explained that the workshop would discuss technical issues, the current landscape of regulations and standards, gaps in the biomanufacturing workforce, and how to enhance the public’s understanding of biomanufacturing. Prather noted that while this moment of action and excitement in the field and of increasing U.S. government interest in biomanufacturing presents many opportunities for

__________________

advancement, there is an increased responsibility to not waste this opportunity in time. “We really want to be able to leverage the collective wisdom and talent to make sure that we are doing the right things in the right way to see the kinds of advances and innovations that we need,” she said.

Elizabeth McNally (Schmidt Futures) said her organization’s Biofutures Program focuses on repurposing sustainable waste biomass; addressing biomanufacturing’s difficult engineering challenges, particularly those involving scaling production; and talent mobilization for the U.S. bioeconomy. Andrea Hodgson (Schmidt Futures) said there is a growing recognition among policy makers that biological research is an increasingly important U.S. economic driver. At the same time, said Mary Maxon (Schmidt Futures), other nations are developing their own biomanufacturing capabilities, with more than 60 countries issuing bioeconomy strategy documents and several recognizing the role a circular bioeconomy³ can play in addressing carbon management goals. Maxon added that U.S. bioeconomy stakeholders can learn from these international policies and actions, which include industry incentives for biogenic carbon capture and business models for sustainable mobilization and valorization of agricultural residues, including through industrial symbiosis.

Federal Government Perspective

Paula Hammond (MIT and President’s Council of Advisors on Science and Technology [PCAST]) said estimates for the future value of the global bioeconomy range from $4 to $30 trillion by the end of the decade.⁴ Currently, she said, the United States lacks the biomanufacturing infrastructure and workforce required to scale biology-based prototypes to products and remain globally competitive, and a PCAST working group identified three key gaps: a lack of manufacturing capacity; regulatory uncertainty; and an outdated national strategy. Hammond said PCAST will make recommendations to address these gaps in an upcoming report.⁵

The recent Executive Order on Advancing Biotechnology and Biomanufacturing Innovation for a Sustainable, Safe, and Secure American Bioeconomy;⁶ directs the National Science Foundation (NSF) to lead the federal government in identifying high-priority, fundamental, and use-inspired basic research goals, along with education and workforce training needs, to advance domestic biotechnology and biomanufacturing capacity, said Susan Margulies (NSF). For example, NSF is partnering with the Department of Energy’s Agile BioFoundry to accelerate biomanufacturing and decarbonize the economy. NSF’s Innovation Corps is an entrepreneurial training program that features an experiential training program for scientists and engineers to ready concepts for commercialization, including those pertaining to the bioeconomy. NSF’s America’s SEED Fund program is funding startup biomanufacturing companies and its Regional Innovation Engines program aims to catalyze a regional innovation ecosystem that includes biotechnology and biomanufacturing.

Hammond and Margulies noted the importance of financial investments from both the federal government and venture capital to bring innovations to market. Hammond said the federal government’s role can include developing shared testbed facilities. Margulies suggested creating NSF–venture capital partnerships to move products through the final stages of development leading to commercialization, and also noted the importance of partnering with communities where biomanufacturing facilities are located.

BIOMANUFACTURING FOR SUSTAINABILITY AND A CIRCULAR BIOECONOMY

Today’s economy is linear, built on a philosophy of take-make-dispose, explained Jim Philp (Organisation for

__________________

Economic Co-operation and Development [OECD]). In a circular economy, waste is a raw material, as products and raw materials are designed to be reused or recycled. In a circular bioeconomy, said Corinne Scown (Lawrence Berkeley National Laboratory [LBNL]), biomanufactured products are designed to either decompose back to reusable starting materials in a composting facility or be recycled.

Scown said a portfolio of solutions is needed to address climate change. Biomanufacturing can contribute to the strategy of becoming independent from fossil carbon sources by developing biologically derived transportation fuels and producing the other products currently made from petroleum. She added that biomanufacturing also has the potential to produce alternative protein sources to replace some of the more greenhouse gas–intensive and land-intensive food products.

To become fossil fuel independent by 2040, Brian Fahie’s (Biogen) company established three principles of sustainable product development as a foundation for its efforts: set a baseline to determine where it can improve sustainability for maximum effect; acquire and share knowledge of best green practices from industry and academia to drive change that will compound itself; and design products for safety, sustainability, health equity, and efficiency. He added that in Biogen’s experience, there is no example where it costs extra to be more sustainable.

Dina Nielsen (Novo Nordisk Foundation Center for Biosustainability) said challenges her organization has encountered include retaining talent, insufficient scale-up capacity to conduct all of the projects it could given laboratory successes, and conducting mission-driven, high-level research spanning many disciplines across an academic setting. She noted that if biomanufacturing is going to make microbial foods, the fermentation capacity needed will be in the billions of liters, compared to the tens of millions of liters currently available.

One advantage of biomanufacturing is that agriculturally produced carbon sources are local, creating the potential to address supply chain issues, said Guillaume Lamy (ARD). His company’s goal is to streamline the use of raw material intermediates to get from one point to an economical solution at the end. He noted that ARD has been part of the Pomacle-Bazancourt integrated biorefinery that is designed for circularity by leveraging waste streams from food processing as inputs for other products’ biomanufacturing processes. Reacting to Lamy’s remarks, Philp opined that the Pomacle-Bazancourt integrated biorefinery is a remarkable example of effective industrial symbiosis already in operation.

Discussion

Scown said sometimes it is possible to develop molecules that enable products to be recycled into the same product or similarly valued products. In other cases, biomanufacturing can convert waste products into value-added products. She added that circularity is not just about carbon, as it also includes nutrients. Nielsen said biomanufacturing and circularity are not yet completely overlapping. For example, biomanufacturing injectable drugs or stem cells requires pure substrates and feedstocks that cannot yet be produced using sustainable starting materials.

Scown said even with the electrification of transportation, there is room for biofuels in hard-to-electrify sectors, like aviation and long-haul freight. She added that there is also an opportunity to make ethanol production carbon negative by capturing the relatively pure carbon dioxide stream emitted for sequestration or use in another biomanufacturing process. She also mentioned that carbon-negative biomanufacturing could start with addressing limited-size markets for high-value products, which could provide flexibility to try new technologies, iterate to optimize the biomanufacturing processes, and then eventually transfer the technologies to larger commodity markets. To move toward an economically efficient industry, Scown referenced the principles of circularity and industrial symbiosis, suggesting starting in places with inexpensive and accessible waste feedstocks as a carbon source, especially if an offset credit is available.

Fahie noted that there has been good work establishing guidelines and metrics for measuring and reporting

sustainability improvements, and government could play a helpful role by harmonizing regulations globally. Nielsen added that there is room for cooperation in the precompetitive space for innovation, and that platforms exist to bring industry, academia, and venture capital together to collaborate before intellectual property (IP) development.

REGULATION AND STANDARDS

Regulation of the bioeconomy is fragmented, both globally and by discipline, said Jeffrey Baker (National Institute for Innovation in Manufacturing Biopharmaceuticals [NIIMBL]), due to the diverse technologies and biosources employed. He added that detailed, prescriptive regulation might not be as appropriate as well-articulated principles of practice because of this diversity and rapid innovation in the field. He noted that new and established biopharmaceuticals need to be safe, efficacious, and reproducibly manufactured to be fit for purpose. In his view, relying on these principles allows flexibility in execution and rapid technical development, which might not be anticipated when a regulation is written.

Agricultural biotechnology, said Anastasia Bodnar (U.S. Department of Agriculture [USDA]), faces a regulatory challenge arising from a dramatic change in the technology used to develop plants, animals, and microorganisms with desired traits. The current regulations were developed with transgenic organisms in mind, but genome editing alters an organism’s genome without introducing foreign DNA. In 2020, USDA’s Animal and Plant Health Inspection Service issued modernized regulations to facilitate innovation and remove unnecessary burdens that impeded small developers and prevented new products from entering the marketplace.

Bodnar noted that according to USDA data small- to medium-size developers accounted for 25 percent of all regulatory status review requests before the new regulations, but that number jumped to 79 percent after the new regulations went into effect. She continued that more than 20 countries have revised or are revising their agricultural biotechnology policies or regulations to accommodate genome editing, though there is no international consensus around regulating genome editing agricultural products.

Sheng Lin-Gibson (National Institute of Standards and Technology [NIST]) said standards can help accelerate research, development, and commercialization; maintain quality and consistency; promote trade; instill confidence among consumers; enable a common understanding and common practices; serve as platforms, common parts, and reference materials; and serve as the basis for quality and risk management. The right standards can promote research and manufacturing innovations; expand manufacturing capacity; secure supply chains, cybersecurity, and other infrastructure; and streamline regulatory review and enable international harmonization. She added that the wrong standards will slow development activities and increase costs. As such, she said, NIST has formed several public–private partnerships (PPPs) and consortia to develop precompetitive solutions to standards development.

Developing standards is difficult, particularly for a fast-moving field such as biomanufacturing, Lin-Gibson said. It involves building consensus that a proposed standard is technically sound, has a justified business need, benefits the entire sector, promotes innovation as opposed to being overly prescriptive, and enables interoperability and integration. She noted that effective standards require a measurement infrastructure to ensure that they are scientifically sound and have high-quality protocols, control specifications, and reference material requirements. Beyond their role in regulatory approvals, standards should promote business-to-business interactions and support academia-to-industry translation, said Lin-Gibson.

Manuel Porcar (University of Valencia/Darwin Bioprospecting Excellence) thinks about standardization in terms of DNA parts, genes, promoters, and other genomic components, but also genetic circuits, engineered cells, and human practices, and each of those levels can be standardized. Currently, synthetic biology is progressing toward having standards. For example, the Standard European Vector Architecture, though lacking a

strong regulatory effort behind it, has become a de facto set of standards that work well and are widely shared at no cost. Another developing standard is the Synthetic Biology Open Language, a free and open-source standard for representing biological designs.

Discussion

Baker noted that one challenge to modernizing the biopharmaceutical industry is industry’s increasing risk aversion in deploying new manufacturing technologies and control strategies. He added that this comes from a perception that questions from regulatory authorities might delay regulatory review and approval. During the discussion of standards, Baker directed the audience to public databases on commonly accepted standards in biopharmaceutical manufacturing and the extensive guidance to industry from the Food and Drug Administration (FDA) and international organizations.

Porcar said the attitude in industry and academia is that developing standards is not rewarding. Lin-Gibson said companies generally focus on solving problems rather than developing standards as the primary end goal. She said a process should be encouraged; for example, in PPPs, where best practices can be captured while problems are being solved, and these best practices can become the starting point for standards development. To Lin-Gibson, standards should allow innovation with robust analytical capabilities and tools instead of defining a minimum performance criterion.

Bodnar said using standards in regulations is a good idea in theory, but in practice, regulations are largely based on laws. Regulations and laws are hard to change, which makes international harmonization of regulations difficult. She proposed looking for commonalities across nations’ systems and developing a common understanding of concepts. Standards have to provide value if the expectation is for compliance with a standard.

Ashley Williams (Office of Senator Christopher Coons [D-DE]) asked the panel to comment on the potential role that mid-sized innovation institutes, such as those NSF is establishing, could play in developing standards. Lin-Gibson said she sees that as a significant opportunity because standards development and related precompetitive efforts would be integrated with technology innovation.

BIOMANUFACTURING WORKFORCE DEVELOPMENT AND EDUCATION

Tom Tubon (BioMADE) said BioMADE has three focus areas in its workforce development program: building awareness of bioindustrial manufacturing careers, preparing the future workforce with innovative education, and supporting the growth of the current workforce with world-class professional development. BioMADE’s engagement strategy includes vertical integration, which includes mentoring internships, articulation pathways, and traditional pathways to achieving credentials at all degree levels, and horizon integration, which includes skills retooling and retraining; working with incumbent workers, displaced workers, and veterans; and industry-based training. Challenges include getting the entire educational community working together and breaking down disciplinary silos.

Tubon and Natalie Kuldell (BioBuilder) noted that biotechnology has advanced in many ways since 1987, when the first community college biotechnology programs started, but biotechnology education has not changed as dramatically. Kuldell said BioBuilder is moving biotechnology education away from memorizing and regurgitating content to an approach focusing on hands-on experiences that inspire students to learn and appreciate the life sciences. She described a BioMADE-funded program that allows students to graduate from high school with a certificate of achievement along with high school and college credits in high-demand fields. The program includes teacher training in an industry-supported learning laboratory.

One challenge, said Kuldell, is finding instructors who are cross-trained and experienced in biomanufacturing and willing to work after school for minimal pay. Another challenge is connecting with parents who might then encourage their high schoolers to enroll in the program. Scaling the program to other schools and for college credit has been a slow process and prone

to failure, in part because it is expensive for schools to adopt and run.

BioNetwork, explained Erica Monique Vilsaint (North Carolina Community College System and BioNetwork), is a life science training initiative to supply talent for area life science companies. Her program works with K-12 institutions to spark interest in the life sciences and overcome the fact that students are unaware of the employment opportunities in the life sciences. She added that many students do not see themselves as a technician, engineer, or scientist, in part because they do not see others who look like them in those roles. The program coordinates and communicates with area community colleges, 4-year institutions, and industry to map industry demand for upskilling and reskilling talent.

Jason Ryder (University of California, Berkeley, and Joywell Foods) said many students graduating from biotechnology programs today do not meet training benchmarks that the biomanufacturing industry sets for associate scientists or process engineers. With colleagues from the biomanufacturing industry, Ryder created a Master of Bioprocess Engineering program that includes hands-on training with bench-scale and pilot-scale bioprocessing equipment common to biopharmaceutical, industrial biotechnology, and food technology processes.

One limitation to scaling this kind of program, Ryder said, is the limited number of modern bioprocessing tools available to meet students’ needs. In his program, one capstone class gives students the opportunity to work on 300-liter bioreactors, pilot-scale chromatography equipment, and pilot-scale protein purification equipment at LBNL’s Advanced Biofuels and Bioproducts Process Development Unit (ABPDU). The program also brings industry scientists and engineers into the classroom to talk about their processes, products, and pathways as bioprocess scientists and engineers. The program is highly successful, Ryder said, with every student in recent cohorts hired into process engineering jobs upon graduation.

Discussion

Tubon said one approach for addressing the shortage of modern biomanufacturing tools for students would be for community colleges to partner with 4-year institutions and industry, which have the equipment and expertise. Vilsaint emphasized communication across these regional partnerships to share resources and ensure students are not missing out on opportunities.

Kuldell said programs should help students see a place for themselves in biomanufacturing. Tubon’s program relies on subject-matter experts to develop curricula that prepare students to work with emerging technologies. The panelists noted the importance of bringing together people from K-12, community colleges, 4-year institutions, and industry to develop curricula to suit regional needs and form partnerships to benefit students and industry.

Tubon noted the importance of taking a multigenerational approach to engagement to attract more underrepresented individuals to training opportunities in biomanufacturing and biotechnology. Engaging multiple generations makes this a sustainable and continuing conversation in the household, he noted. Vilsaint suggested bringing industry representatives to speak to underrepresented communities so community leaders, parents, students, and educators can learn about the opportunities and start conversations that can spark a student’s interest in pursuing a biomanufacturing career.

ECONOMIC CONSIDERATIONS AND CHALLENGES IN BIOMANUFACTURING

Stephen Sofen (Abata Therapeutics) said he would not hesitate to invest in a promising new technology because most new technologies at one time seemed impossible and experience failures along the way to commercialization. He does worry, however, about factors that would slow development, like the potential for cumbersome legal and IP constraints in PPPs. Sofen also noted potential commercialization delays after partnering with an academic core facility to manufacture clinical trial material, if that core facility is not able to support the pivot to commercial manufacturing on account of jeopardizing its parent’s tax-exempt status.

When looking at a new investment opportunity, Pulakesh Mukherjee (Imperative Ventures) tries to understand if the new technology has a structural advantage over existing technologies, if it is scalable, and whether life-cycle analysis shows that a biological system is better

for the environment. In his view, customers do not care about technology, just the product, its cost, whether it meets specifications, performs as expected, and passes regulatory review. He cautions developers to define the technical and economic advantages at the earlier stages of research and to not start scaling before meeting those metrics. Everything being equal, he suggested that a new process to make an existing chemical should cut its price by 20-25 percent.

Jenny Rooke (Genoa Ventures) said there are common themes for why companies in this space fail, starting with the fact that while a technology might be exciting, technology is a feature, not a complete solution to what the marketplace needs. The second theme is that scaling is hard, and many ways that scaling fails are not found in textbooks and are knowable only through experience. She added that many entrepreneurs do not have the experience to incorporate failure into their planning, risk mitigation, and capital needs, or know when it is reasonable to seek venture funding versus focusing on grants or partnerships. A third theme, she said, is a lack of market data and technical transparency in biomanufacturing, where incumbents closely hold process details. She noted that as a result, it can be difficult for startups to get the necessary data to make important planning decisions and bring innovations in biomanufacturing to the market.

Discussion

Mukherjee noted that in the industrial chemical space, the challenge is not a lack of available market and cost data but getting the right talent and connecting with the customer. In the biopharmaceutical world, information disconnects occur in understanding a disease, said Sofen; most biotechnology companies go after diseases with no treatment, so an understanding of a disease may be lacking.

Rooke said the barrier to success for biofuels was the additional large capital expenditure needed to create the physical infrastructure and delivery system that was separate from the capital needed to achieve the necessary technical innovations. She sees the same thing in the alternative protein space, where capital expenditures will need to be significant. As a result, she said, it is not a space she has invested in because while she and her team have considered, it does not make economic sense yet. Mukherjee agreed, noting challenges related to the unit economics for both biofuels and alternative proteins.

Sarah Richardson (MicroByre) asked if there are problems that are not addressable with today’s or foreseeable technologies. An obvious non-starter, said Sofen, is using a living organism to make a toxic substance that would kill the organism. Mukherjee said, in principle, he sees no constraints from a technology perspective regarding industrial chemicals aside from the perspective of cost and life-cycle analysis. Rooke said any limitations have to do with intended uses. For example, using crop plants to make and deliver bioactives is a clever use of biology, but it can run up against the variabilities of life when it comes to quality control and quality assurance in the manufacturing process. Each panelist noted the importance of considering a product’s cost structure and conducting a techno-economic analysis, underscoring the importance of accessing talent with the experience in that type of analysis.

All three panelists thought that predictive biology and artificial intelligence (AI) will advance biomanufacturing in time, particularly if they can help narrow the parameters to explore. Mukherjee said AI can be a useful tool for the development of multicomponent systems, but not as a stand-alone basis for a company that promises to optimize other firms’ processes.

BIOMANUFACTURING ECOSYSTEMS AND PARTNERSHIPS

Maureen Toohey (BioFabUSA) said when BioFabUSA opened in 2017, she hypothesized that gaps in good engineering and manufacturing processes limited the transition of tissue engineering technology from academic research to commercial biomanufacturing. In the tissue engineering space, she said, one challenge is going from a few cells to the trillions needed to fashion an organ and developing the different types of scaffolding and bioreactors that could produce an engineered tissue with the appropriate functionality. The lack of standards and fragmented regulatory environment are other obstacles, she noted.

Toohey noted several successes, including BioFabUSA’s “Tissue Foundry” prototyping. Using a scalable, modular, automated, and closed system, it developed production processes for cell, tissue, and organ constructs. Another success is its deep tissue characterization center using a big data approach to compare analytical data to observed clinical outcomes from manufactured tissues. This allowed BioFabUSA to identify critical quality attributes that its consortium of more than 180 members used in regulatory discussions with FDA. Toohey added that BioFabUSA helps its small company members with aspects of commercialization, such as acquiring funding, and developed a certificate and training program for biotechnicians that has largely attracted women and less advantaged students.

Kelvin Lee (NIIMBL) said NIIMBL projects use lessons learned from antibody manufacturing to accelerate the development of standards and to scale the ability to manufacture cell and gene therapies. To address the biopharmaceutical industry’s conservative approach to adopting new manufacturing technologies, NIIMBL created testbeds for generating a shared understanding that can lead to a regulatory strategy. Lee said NIIMBL has created a culture around shared manufacturing technology innovation that enables companies to share their experiences and knowledge and innovate collaboratively.

Among various workforce programs, Lee described how NIIMBL collaborates with Historically Black Colleges and Universities and Minority Serving Institutions to identify individuals interested in the life sciences, but who had not considered working in biomanufacturing, in order to expose them to a variety of career options. Through this program, NIIMBL member companies offer dedicated internships and hire program graduates.

Lee and Douglas Friedman (BioMADE) discussed opportunities to advance biomanufacturing that they suggested at a recent White House meeting. Their suggestions included investing in infrastructure for manufacturing scale up; investing in collaboration across institutes, centers, and programs to identify synergies and create new network effects; and creating better coordination within the federal government to support the nation’s bioindustrial strategy.

BioMADE, said Friedman, focuses on biomanufacturing low-margin, high-volume products. Despite billions of dollars invested in developing modern biotechnology in the laboratory, he said, commercialization has not been a focus outside of the health space. He added that as a result, the number of successes has not been sufficient to maximize the value of industrial biotechnology. BioMADE has been identifying economic and commercialization challenges and determining what it can do to address those challenges. One area of emphasis Friedman described is techno-economic analysis to determine whether there can be an economically viable bio-based solution for a commodity chemical.

Friedman also noted that BioMADE helps companies understand the opportunities in the federal procurement market to provide more sustainable products at competitive prices. BioMADE also advocates for building more domestic biomanufacturing scale-up infrastructure, particularly for producing kilograms of a product that potential customers can use and validate. Friedman said having customer validation, a positive economic analysis, and robust technology opens the door to private capital investment and a promising business model.

Discussion

Toohey noted gaps associated with collaborations across Manufacturing USA institutes, federal research agencies, national laboratories, and engineering disciplines. One key for successful collaboration is to look at potential collaborators as customers and identify their needs, Friedman noted. Along with stronger collaborations, Lee said, adjacent disciplines like electrical, mechanical, and automation engineering are essential for biomanufacturing as they contribute a wide range of skillsets to process development.

When asked about how biomanufacturing can better serve underserved, rural, and diverse communities, Friedman noted that biomass is distributed across the nation. Instead of transporting biomass across the country, biomanufacturing facilities can be close to their feedstock sources, which could enable more equitable

distribution. Regarding serving patients with rare diseases, Lee said NIIMBL aims to address a significant challenge—the lack of a viable commercial market—by developing open platform processes for gene therapy vectors that FDA could approve and developers could use to deliver a specific gene of interest. Automated production platforms for new modalities such as CAR-T immunotherapies could reduce the cost of today’s expensive manufacturing processes. Toohey agreed that focusing on common process and platform technologies and reducing manufacturing costs will enable cures for more rare diseases.

TRANSLATING LESSONS FROM DIFFERENT BIOMANUFACTURING SECTORS

Don Parsons (Moderna Therapeutics) said messenger RNA (mRNA) technology has the potential for developing a production infrastructure that is smaller and less capital intensive than some conventional biomanufacturing processes. He described how in an uncommon move, the company diverted resources from fundamental R&D at the beginning of its work on mRNA to solve manufacturing issues early and accelerate the clinical evaluation of mRNA therapies. As a result, the company produced its COVID-19 mRNA vaccine for clinical trials in 6 weeks after the SARS-CoV-2 sequence was announced. The company also had ongoing discussions with regulators and formed partnerships with the National Institutes of Health before the pandemic.

Parsons noted that the company had to scale up and scale out to produce a COVID-19 mRNA vaccine at the scale needed to address the pandemic. To ensure that outside manufacturers could produce the vaccine with the required consistency, the company crafted a kit with rigorous standards for core equipment, procedures, and training that it could reproduce globally. Establishing corporate and governmental partnerships was critical for funding and to the scale out process, Parsons said, and Moderna’s partnership with Lonza Group was particularly important for creating a global supply chain. One important lesson was the need to think deeply about how to articulate core principles to regulators who lacked the time to become familiar with this new technology.

Jennifer Holmgren (LanzaTech) described how her company captures waste carbon and transforms it into sustainable products already on the market using biology, another example of industrial symbiosis discussed at the workshop. Holmgren described how they do this using bacteria that metabolizes carbon monoxide and/or carbon dioxide and hydrogen, and produces ethanol in a continuous bioreactor, with ethanol serving as a building block for an array of products, including sustainable aviation fuel, fabrics, detergents, and more. An advantage of using biology, Holmgren said, is that it can capture and use the chaotic inputs that are inherent to waste feedstocks. She added that biology also shines in its ability to selectively make a product, which enables distributed manufacturing. She noted that the company did as much as it could to validate its technology at smaller scales before going to commercial production scale, and that PPPs and strategic investors played a pivotal role in the 15-year journey to develop a commercial process.

Synthetic biology will be an important enabler for the future of biomanufacturing, said Holmgren. For example, LanzaTech reprogrammed bacteria to use gases rather than sugar as its source of energy. This was only possible by working with leaders in the field who had developed technology to modify sugar-consuming organisms. Another advantage of biology is the flexibility it provides, on one hand using an array of different feedstocks and on the other hand producing many products with one manufacturing plant and the same hardware. Holmgren noted the need for policy to be technology neutral to enable disruptive technologies to reach the market.

Discussion

Parsons said one advantage of working internationally is access to talent. Holmgren agreed and added that legislation is more favorable internationally. As one example, she noted that her company has not built ethanol production facilities from steel mill gasses in the United States because the Renewable Fuel Standard program mandates the use of ethanol made from sugar or biomass for obligated parties. She next noted that some venture capital firms have started 20-year funds to make longer-term investments in breakthrough green technologies. Both Holmgren and Parsons pointed

to the role government can play by creating market pull for products such as vaccines and sustainable aviation fuel. Holmgren said a whole-of-government approach to procurement will be critical for success in biomanufacturing.

MODELING, DATA, ANALYSIS, AND PROCESS CONTROL

Richard Braatz (MIT) said biopharmaceutical companies are increasing their use of mechanistic models to design their systems. He said models enable conducting multiple experiments in silico before conducting a single laboratory experiment, saving time and money. He noted that experimental data can then be used to estimate initial model parameters and iteratively improve a model.

Data quality and reproducibility are critical for predictive analytics, said Theresa Kotanchek (Evolved Analytics), and having more data is less important than having a diverse dataset. Additionally, early identification of when a process is not working is as important as predicting when it is. In various biomanufacturing contexts, she described how symbolic regression can be used in predictive analytics. She suggested that managing and accessing data as a reusable asset for full system integration is key.

Chong Wing Yung (Agilent) said while multi-attribute methods analysis is powerful and can consolidate separate assays, it requires complex hardware and software and presents difficulties in aligning instrument performance across multiple laboratories to ensure data integrity. One solution, he said, is to develop automated and robust analytical methods requiring minimal analyst training and with acceptable assay variation for quality control testing.

Stephen Balakirsky (Georgia Institute of Technology) described his work on automated control systems and machine learning algorithms to improve the reproducibility and yield of autologous cell biomanufacturing and decrease cost. He described how the closed-loop system can detect and quantify essential metabolites, model how to change those metabolites to optimize cell production, and change the operating conditions in the bioreactor using the system’s automated control features. Balakirsky noted that improving model-based control will require a better understanding of the basic biology of the manufactured cells and that more automation early might help solve challenges related to having consistent starting conditions.

Discussion

Using open-source software for intelligent bioreactors and biopharmaceutical processes could be challenging because of the extensive testing required for regulatory approval, said Braatz, but it is proving useful for offline process and workflow development. Balakirsky said a robust, open-source software community in the cell therapy manufacturing space is working on planning and control algorithms and communication frameworks in the laboratory. Kotanchek added that open-source software will be useful in chemical infrastructure control.

One challenge for automating process control is sampling a bioreactor aseptically, particularly in the cell and gene therapy space, said Yung. Non-contact, spectroscopic methods may work, and that will require knowing what parameters are important to monitor. Whatever techniques are deployed, the sensors and analytical tools must be reproducible and robust, said Kotanchek. Balakirsky added that biofouling is an issue with sensors inside a bioreactor.

BIOMANUFACTURING PLATFORM DEVELOPMENT

Nili Ostrov (Cultivarium) said most biomanufacturing processes today use one of a few microorganisms, yet a strong bioeconomy should use a diversity of microorganisms to make a wide range of products. Her organization focuses on identifying microorganisms amenable to biomanufacturing. This is difficult, Ostrov said, in part because of the hard challenge of culturing new organisms and the time required to make them genetically tractable. For example, she and her Cultivarium co-founder took more than 10 years to make a marine bacteria genetically tractable and useful as a platform organism for biomanufacturing. She noted that when her organization develops a useful assay for a particular organism, it puts the protocol on its website. She also mentioned the need for organism readiness

levels, similar to technology readiness levels, to define the information and technologies for working with an organism. Her organization is working to establish those readiness levels.

Rahul Singhvi (National Resilience) said his company is developing manufacturing platforms to address pain points in biomanufacturing. One project focuses on the scalability and cost structure of autologous cell therapy manufacturing. Labor costs are high for these products, so automation is one approach to reducing cost. In particular, he said, microfluidics might help with automating the small volume assays for the multiple parameters needed for product characterization. One challenge, he said, is to develop a system for distributed manufacturing, a key feature of autologous cell therapy, that can gain regulatory approval. One way to achieve regulatory approval, he added, is to demonstrate that the new manufacturing system generates a product with the same key attributes as the same product produced with the current state-of-the-art process.

Discussion

Singhvi said he wants to see a new approach to manufacturing viral vectors for gene delivery developed. This would entail the difficult task of establishing stable producer lines containing genes of the host cell and controlling the expression of those genes in a way that produces stable particles in which a large proportion of the particles contain the gene of interest.

Ostrov said she wants to see more innovation in biomanufacturing methods at an early stage, with new methods developed, for example, by International Genetically Engineered Machine competition teams. This would only be possible, she said, with better communication between people engaged in biomanufacturing and those conducting early R&D activities. Communicating failures would be particularly important in such an effort.

BIOMANUFACTURING INFRASTRUCTURE AND TOOLS FOR SCALING

The main challenges of autologous cell manufacturing, said Greg Russotti (Century Therapeutics), are that it is labor intensive, requires multiple manipulations that increase costs and manufacturing times, and has variable manufacturing reliability. In his view, the solution to these challenges is automation and robotics, which can also be applied to allogeneic therapies. While scaling out is important for autologous cell manufacturing as batches are patient-specific, scaling up is key for allogeneic cell therapy manufacturing, he said. This requires bioreactors, a mechanism for harvesting the cells, and equipment to fill vials with cryopreserved cells in a manner that does not damage the cells.

Producing cultivated meat products, said Brett Schreyer (SciFi Foods), requires scaling mammalian cell production methods developed by the biopharmaceutical industry by many fold and at a much lower cost. He added that being green will only take a product so far—it must be cost advantaged. Techno-economic modeling, he said, is key when planning to operate at large scale. For food production, Schreyer highlighted that supply chain issues are critical, particularly the need to obtain millions of kilograms of food-grade amino acids at a low cost. Two lessons he learned are the importance of having sufficient cash on hand and while proof of concept is hard, scale up and eventual commercialization of bio-based products are even harder and more expensive; it is in scale up and commercialization where many startup companies fail.

Kris Tyner (Culture Biosciences) said it is possible to derisk some scaling issues, especially during the process development stage, by leveraging small, cloud-based bioreactors in a cloud-connected, high-throughput bioprocessing laboratory and by using large datasets. She said the number of replicates needed to generate sufficient data for statistically driven decision making is a function of process variability, effect size, and tolerance for false positive and negative results. Failure to run enough replicates, said Tyner, can waste resources and delay programs. She noted that a complete understanding of process metrics requires fully tracking inputs and outputs, such as evaporation. In addition to its cloud-based bioreactors, her company developed recipe-based programming that enables easily and accurately varying process step points across conditions.

Charles Isaac (Fermic) explained that his company, a contract manufacturing organization, has about 2.3 million liters of fermentation capacity across different scales, including the pilot scale (1,000-5,000 liter fermenters) through the commercial scale (190,000 liter fermenters). He said his company’s most successful clients have built their own downstream processing facilities on Fermic’s toll production site. In his view, strong management support and coordination between the client company and toll manufacturer are necessary for success when going forward with technology transfer and scale-up processes. Isaac said transferring analytical methods to toll manufacturers reduces the time to get meaningful data, as does having a robust pilot-scale model at the 3,000-4,000 liter scale before trying to go to larger scales in the tens or hundreds of thousands of liters.

Discussion

Emily Greenhagen (Ginkgo Bioworks) noted that infrastructure innovation needs to go beyond fermentation and include downstream processing, particularly given the diversity of downstream processes required for biomanufacturing. Tyner added that innovation is needed around bioreactor design, including introducing different types of equipment on the market in order to reduce the cost and time of pilot-scale demonstrations. She added that her company has innovated at different scales, including with small-scale, 5-liter bioreactors. Schreyer said automation at smaller scales and high-throughput technologies could reduce the failure rate when transitioning to the pilot scale. For cell-based therapies, Russotti said innovation is important for in-process monitoring, automation, process standardization, and cryopreservation. Each speaker noted the importance of workforce development and expressed their support for establishing shared pilot facilities, for example.

DISCLAIMER This Proceedings of a Workshop—in Brief was prepared by JOE ALPER, STEVEN MOSS, and ANDREW BREMER as a factual summary of what occurred at the workshop. The statements made are those of the rapporteurs or individual workshop participants and do not necessarily represent the views of all workshop participants; the planning committee; or the National Academies of Sciences, Engineering, and Medicine.

WORKSHOP PLANNING COMMITTEE KRISTALA L. J. PRATHER (Chair), MIT; EMILY GREENHAGEN, Ginkgo Bioworks; BRIAN D. KELLEY, VIR Biotechnology; JAMES PHILP, OECD; SARAH RICHARDSON, MicroByre; KRISHNENDU ROY, Georgia Institute of Technology; and DEEPTI TANJORE, ABPDU.

REVIEWERS To ensure that it meets institutional standards for quality and objectivity, this Proceedings of a Workshop—in Brief was reviewed by JEFFREY BAKER, NIIMBL; PATRICK M. BOYLE, Ginkgo Bioworks; and KRISTALA L. J. PRATHER, MIT. LAUREN EVERETT, National Academies of Sciences, Engineering, and Medicine, served as the review coordinator.

SPONSORS This workshop was supported by Schmidt Futures. Dr. Keith Yamamoto (University of California, San Francisco) provided in-kind support through a grant from the National Science Foundation.

For additional information regarding the workshop, visit https://www.nationalacademies.org/our-work/successes-and-challenges-in-biomanufacturing-a-workshop.

Suggested citation: National Academies of Sciences, Engineering, and Medicine. 2023. Successes and Challenges in Biomanufacturing: Proceedings of a Workshop—in Brief. Washington, DC: The National Academies Press. http://doi.org/10.17226/26846.

Division on Earth and Life Studies Copyright 2023 by the National Academy of Sciences. All rights reserved.