Advances in stem cell science and bioengineering have given rise to many types of synthetic living models of human biology. These lab entities do not neatly fit into any of the existing approaches to bioethics guidelines and institutional review, or into the areas covered by engineering ethics. Traditionally, bioethics and engineering ethics have stood as separate spheres of scholarship and practice. These fields should be combined to form a new hybrid approach – “bioengineering ethics” – to address research that unites the biological capacity of human cells with engineered platforms.
Historically, research ethics arose out of a concern for the treatment and research uses of “natural kinds”  (entities found in the natural world) that are believed to have some degree of moral status or level of moral considerability, be they human beings, fetuses in utero, embryos, animals, genes, gametes, etc. Ethical standards for review committees have developed over the past forty plus years and are now considered to provide robust guidance and appropriate oversight for studies involving typical kinds of research entities. However, “non-natural kinds” of entities have emerged rapidly over the past few years – particularly, synthetic human biological constructs – and existing ethical canons and oversight infrastructure are not well suited to address these new creations. As discussed below, there exists much uncertainty over even basic research ethics issues, such as which institutional committee should be responsible for reviewing this form of research and what ethical standards ought to be employed for determining approval.
Traditional research ethics is not sufficient to capture all important ethical aspects of this cutting-edge research, for these traditional standards tend to focus on informed consent requirements for original cell line donors and their genetic privacy interests. But this typical bioethics approach says little about where the ethical limits lie for the myriad ways in which stem cells can be radically bioengineered in the lab after they have been ethically procured. A new form ofbioengineering ethics should be developed by way of a novel combination of bioethics and engineering ethics to promote the responsible development and use of synthetic entities built from human cells.
New Research Entities
Engineered entities are now being generated in laboratories from human stem cells to form biologically dynamic, living models of human biology. These models can be used to study various aspects of human development and to test new drugs and therapeutics. Prominent among these models are organoids (small stem cell-derived 3D structures that self-organize into functional cell types and recapitulate basic organ functions) and embryo models (stem cell-derived simulations of post-implantation embryos). Organoids and embryos models are just the tip of the iceberg of what is possible, however. George Church’s lab at Harvard Medical School is actively interested in bioengineering various aspects of human biology to confer even more advanced specific capabilities and traits to lab entities for research. In a 2017 article, Church and his colleagues, John Aach, Jeantine Lunshof, and Eswar Iyer coined the term “SHEEFs” to refer to these and other new lab creations, which stands for “synthetic human entities with embryo-like features.”  “Around the same time, a broader term “M-CELS” was coined by a research consortium based at the Massachusetts Institute of Technology (MIT) called EBICS (Emergent Behaviors of Integrated Cellular Systems) – a National Science Foundation Science and Technology Center. The term M-CELS stands for “multi-cellular engineered living systems,” and in comparison to SHEEFs, M-CELS encompass an even broader array of synthetic entities that can be made to model human biology.  Essentially, M-CELS are lab entities built from human stem cells (or their direct derivatives) paired with engineered non-biological components. Some M-CELS research is aimed at forming functional living machines that might become capable of sensing and information processing. An early example of this was described in 2014 in the Proceedings of the National Academy of Sciences of the United States of America (PNAS).  Here EBICS researchers developed a 3D printed hydrogel “bio-bot” made of mammalian skeletal muscle that could be controlled with electrical stimulation. Technological capabilities have evolved in the past few years to the point that it is now conceivable that research might expand upon this 2014 experiment. For example, by attempting to link human brain organoids to 3D printed artificial bone-like scaffolds seeded with muscle and nerve cells, EBICS researchers might investigate whether these types of M-CELS can be programmed to model the neural-muscular interface of real human beings.
This growing trend toward creating complex living models of human biology (i.e. organoids, embryo models, SHEEFs, and M-CELS) is a natural progression for stem cell research and bioengineering. (For convenience, the umbrella term M-CELS will be used throughout this essay to refer to all of these living models of human biology.) Many scientists have come to the realization that working with stem cells in flat, two-dimensional culture systems is limited for understanding how real tissues are formed in the body. Real tissue systems, organs, and embryos arise through self-organizing cell behaviors in three-dimensional environments with necessary mechanical and chemical stimuli. M-CELS of various types offer significant biomedical research benefits because they enable scientists to study accurate representations of human biology at the benchside without having to utilize human subjects or animals to study tissue and organ formation, the developmental effects of genetic diseases, or new drug targets.
Insufficiency of Current Oversight to Address M-CELS
Alongside these great scientific promises, however, the quasi-human/artificial ontological ambiguity of these synthetic models complicates how one might think about their moral status as research objects and the ethical limits surrounding their creation and use. Researchers working with M-CELS have intimated that they are uncertain about how far ethically they can push their experiments in the lab.  Researchers do not want to hear after spending significant amounts time and energy on a project that they have ventured into an ethical red zone. And institutional regulators do not want to see the publication of contentious research that did not first go through rigorous ethical review. Yet, despite these wishes, no one really knows at this point exactly what kinds of ethical issues will be raised that will be important to consider and manage as this new form of research develops.
Attempts thus far to identify and address ethical uncertainties at this early stage of research have been inadequate. As researchers active in this nascent area, the Church lab leadership has maintained that they have a social responsibility to think proactively about the ethics of this research and its future directions. A few years ago, they sought guidance from their institutional oversight body, the Harvard Embryonic Stem Cell Research Oversight (ESCRO) Committee. The Harvard ESCRO decided not to define bright-line limits at that time in the case of SHEEFs and asked to be updated as the science progresses toward synthetic entities that may have “morally troubling” features.  The Church lab also sought direction from guidelines issued by the International Society for Stem Cell Research (ISSCR), which I helped develop. However, the 2016 ISSCR guidelines relating to the creation and use of synthetic biological entities were too vague to be of much help on the applied and specific questions raised by the Church lab. (I should note that the ISSCR guidelines were left vague on this issue because the ISSCR realizes this is a novel subfield of stem cell science and thus should be closely followed before more specific guidelines can be offered). 
Perhaps one chief reason why we are currently in this state of regulatory and ethical uncertainty arises from an underappreciated historical quirk. Ever since guidelines for stem cell research were first formulated over a decade ago, all studies using well-established stem cell lines confined only to in vitro studies (i.e. no human or animal subject involvement) have been routinely categorized as the “least controversial” form of stem cell research. As such, they normally did not require close ethical monitoring by research institutions, much less full review by stem cell oversight boards. But now, with the rise of complex biological models of human biology – all of which are confined to the sphere of benchside research – the traditional fast-track approvals process for many forms of in vitro stem cell research may no longer be warranted. It is also unclear whether stem cell oversight committees, as they are currently constituted, are the best institutional bodies to review this type of research, since the human cells are usually differentiated by the time they are paired with heavily engineered non-biological components. In short, this is not the typical picture of “stem cell research” to which stem cell oversight committees have grown to become accustomed.
A Productive Path Forward: Bioengineering Ethics
At roughly the same time that the Church lab and ESCRO Committee conversations at Harvard were occurring, I independently conceived of and published a bioethics article in Cell Stem Cell calling for the need to expand the scope of bioethics and its approach to dynamic models of human development by bringing components of contemporary engineering ethics into the mix.8 My article was prompted by advances in organoid technology as well as the embryo modeling work of Dr. Jianping Fu’s lab at the University of Michigan, which raised intriguing ethical and legal questions about the status of embryo models and the limits of their use. I argued in this article that a productive path forward would be to bring bioethicists and scientists together at the benchside to discuss collaboratively the ethical choices and value tradeoffs informing early research decisions during the design phase of their experiments, and to remain deeply engaged through the development and implementation of experiments, helping to navigate ethics throughout. One important benefit of this new approach is that it would avoid presenting bioethics as always a reaction to radical developments after the fact, which largely focuses people’s bioethical attention on a new technology’s ethical, legal, and social implications. Instead, my proactive collaborative approach calls for bioethicists to be the co-designers of research trajectories and choices made by scientists at the benchside, thus helping to infuse ethical reflection far more upstream during the development phases of new biotechnologies.
Often-used methods for defining new ethical practices and standards in cutting-edge science, such as stem cell research and genome editing, do not seem optimal for addressing the extraordinary variety of synthetic biological constructs possible in human modeling research, which can change configurations very quickly based on relatively unconstrained decisions at the bench. For example, multidisciplinary working groups and expert workshops may be too slow, infrequent, top-down, and removed from the action at the benchside to adequately identify and address emerging ethical issues in this area. (Recall that the ISSCR guidelines are too vague to guide labs like the Church lab.) We must take an alternative, much more nimble path, directly through the trenches.
It should be acknowledged that some institutions have research ethics consultants who can provide advice to those institutions’ own research teams. However, the bioengineering ethics approach I outline here is distinct from these types of consult services in two important ways.
First, research ethics consultation services almost always focus on improving informed consent processes at some pre-Institutional Review Board (IRB) stage of the investigators’ research. Research ethics consultants are also often asked to help find ways to reduce risk to study subjects in the design of the research protocol. In contrast, bioengineering ethics should explore ethical areas that lie far outside human subjects protections considerations. Furthermore, the type of cutting edge research referred to above does not qualify as human subjects research as such. Thus, there may be important emerging ethical issues for this field that would not be captured appropriately, or at all, by existing research ethics consultation services.
Second, institutional research ethics consultation services are summoned at the request of the research team onlyafter they have identified a concern regarding their research project (again, usually around human subjects protections). Bioengineering ethics, on the other hand, should be structured in a proactive way so that ethicists are well positioned to identify potential issues at the earliest stages of the research that the teams otherwise might have overlooked, and that can then be considered in the protocol. The ethicists and research teams can also help inform each other by collaboratively thinking through any ethical issues that may appear during the lifecycle of a research project.
Bioengineering ethics can offer a fresh new way to approach research ethics aimed specifically at the creation and use of complex bioengineered constructs. Traditionally, bioethics and engineering ethics have represented different fields of scholarship and practice. These two fields must be combined in order to address adequately M-CELS research, which unites the biological capacity of human cells with engineered platforms and artificial support systems. The collaborative nature of contemporary engineering ethics provides a valuable reference point for scientists in the lab to understand the need to explicitly discuss the trade off decisions that inevitably drive design choices. Given that new biological models are being generated through a dynamic blend of engineered components and autonomous cell behaviors, what would it mean to bring engineering ethics into the mix? To answer this question, one must consider important advances in contemporary engineering ethics.
New Conceptual Ground
Given the strong engineering aspects involved in creating dynamic models of human biology, it is tempting to assume that their associated ethical issues will be limited to their uses and societal impacts, as is the case with many other artifacts or tools created for biomedical research. But this view merely resurrects an old-fashioned version of engineering ethics. Engineering ethics has traditionally focused on issues that are already familiar to many scientists, such as anticipating the negative social implications of engineered products, or assigning blame for adverse outcomes, or promoting the professional virtues of whistleblowing or public education. But contemporary engineering ethics goes far beyond these considerations and is rooted in the belief that engineering itself is a value-driven activity and that, as such, there exists a range of possible values, including ethical values, that can inform the choices engineers make during the design process.  There will of course always exist some restrictions on the range of engineering choices available based, for example, on regulatory requirements, intended uses and goals, safety, cost, and – with respect to the present issue – biological relevance. Nevertheless, there will also typically exist more design choices than can be simultaneously fulfilled, and an engineer’s decision of which trade-offs are acceptable will not be a value neutral one. Doing engineering ethics in this contemporary sense involves actively contemplating which values ought to guide the engineering task at hand and why, thereby creating awareness of how trade-offs between design choices are being framed. Contemporary engineering ethics is therefore collaborative; values from varying perspectives need to be weighed explicitly during the design phases of engineering projects.
For M-CELS research to proceed responsibly, the field of bioethics can help by incorporating the principle (borrowed from engineering ethics) that stem cell scientists and their collaborating bioengineers must actively deliberate about their guiding values during the design stages of their experiments, along with bioethicists who can assist in these bioengineering deliberations. Although many different biomedical researchers could benefit from taking a multi-perspectival approach to experimental design, one chief advantage of bringing contemporary engineering ethics into M-CELS research is that researchers at the bench will be encouraged to reflect openly on the implicit ethical choices they are making in designing their models while still satisfying their research aims. This reorientation would require a shift in emphasis for both engineers and bioethicists to arrive at a new form of bioengineering ethics. As I have argued elsewhere, what is needed is a fresh approach that utilizes contemporary engineering ethics, which accepts that engineering itself is a value-laden activity and that the values that drive design decisions are often themselves ethical in nature.  Regulators, scientific organizations like the ISSCR, and other scientists and trainees may benefit from a fresh approach by viewing the ethics of M-CELS research as arising out of a proactive collaborative endeavor between scientists and bioethicists.
It is my hope that bioengineering ethics will normalize regular lab interactions between bioethicists and scientists. It may also revive fundamental debates within engineering ethics itself that have pitted competing definitions of engineers’ professional norms of conduct against one another. For example, some scholars have argued that the primary orientation of engineers’ responsibilities is to the public good (i.e. engineers are important conduits for promoting social well being), while others have questioned this claim, arguing that engineers are not trained to make ethical and policy decisions themselves.  Part of the task of defining what bioengineering ethics is will require careful reflection about what the bioengineer’s socially responsible roles may be, if there are any.
In debates about emerging biotechnologies, it is well known that tensions exist between technological pessimists and technological optimists. The former believe that, although not all technological advancements ought to be opposed, developers should internalize a critical attitude toward technology and its promise of producing social good. Some may argue, for instance, that M-CELS developers should be technological pessimists in this sense. If this is the case, then one would first need a clear idea of what the real scientific possibilities are. Otherwise all one is left with are vague admonitions about “drawing lines” in order to avoid going “too far.” Thus there is a need for ethicists to work closely with scientists at all stages of the research process.
Technological optimists on the other hand may risk not being reflective enough of the development of new biotechnologies. M-CELS researchers (who tend to side with technological optimists) should be prompted to acknowledge that technology can have undesirable aspects. Getting to the point where bioengineering ethics can be carved out as a productive space for both technological pessimists and optimists will first require communicative capacity-building on the workfloor of the lab, which will be necessary for bioengineering ethics to take shape as a new approach to research ethics.
Above I have sketched the bare outlines of what bioengineering ethics would look like. There still remains much that needs to be worked out. I conclude this essay by highlighting four areas that warrant further development.
First, complex multi-cellular constructs like M-CELS pose a unique problem for bioengineers that other engineered constructs made from non-living matter do not entail – namely the potentially unpredictable nature of biologically autonomous, self-organizing human cells. While M-CELS designers might intend for their constructs to behave in certain desired ways, biology may offer surprises that upset their best-laid plans. However, the possibility that biology might surprise us is no reason to dismiss the bioengineering ethics approach I have outlined here. As knowledge grows of how to harness the autonomous capabilities of human cells, I am confident that bioengineers will become better positioned to design model systems that are more controllable and reproducible. Progress will depend on the fact that – since early attempts at incorporating bioengineering ethics into design choices will themselves be experimental – bioengineers and their collaborators will have to watch iterations of these early attempts closely and learn from them.
Second, in thinking about the role that values must play in making design trade off decisions during the course of M-CELS development, it is easy to gloss over a fundamental philosophical difficulty. I call this challenge the “incommensurability problem.” Saying that decision makers ought to make explicit trade off decisions at the benchside implies that the design goals that must be balanced against one another are, in principle, commensurable. That is, it might be presumed that there exists a common unit of measurement upon which to justify how this balancing act can be decided. But is there in fact such a common unit of measurement available to bioengineers?
Incommensurability results when the values of different aims or goods cannot be reduced to a common measure for comparison and choice. In cases of true incommensurability (as opposed to the practical incompatibility of not being able to have one’s cake and eat it, too), the value of each option cannot be placed on the same scale of measurement without grossly distorting our understanding of what it means for each option to be valuable in its distinct way. Many, perhaps even most, design goals that cannot be simultaneously met or maximized could be constrained by incommensurability problem, not just practical or budgetary constraints.
Maybe one way to avoid the incommensurability problem is to argue that it is not really incommensurability we should worry about but rather incomparability. Incommensurability occurs when there is no common unit of measurement against which we can rationally base our comparisons and decisions. But this lack need not preclude our ability to make comparisons among options. One leading philosophical explanation is that comparisons between two incommensurable goods can be achieved if these comparisons are made in terms of some “covering value” that holds between them.  Such comparisons will take the form ‘X is better than Y with respect to covering value V.’ For example, one might argue that a legal career is better than an artistic career with respect to income, even though the value of law and art are incommensurable, or that doing philosophy is better than bowling with respect to the engagement of one’s higher-order mental faculties. In the case of M-CELS, one might imagine that the relevant covering values used to make comparisons across conflicting design goals could boil down to fundamental design principles that bioengineers would agree upon from the outset, such as cost effectiveness, scalability, controllability, and reproducibility that are best served by various design options for the M-CELS in question. Alternatively, the appropriate covering values might be determined through consultation with other stakeholders, such as the various publics (e.g. patients) who will likely benefit from a particular M-CELS technology.
Of course, determining these basic bioengineering design principles or other possible covering values is yet another area in need of development for bioengineering ethics. This does not appear to be a straightforward task to me. Nevertheless, the articulation of determinative covering values could aid – indeed, might even constitute – the standards of evaluation for M-CELS protocols by research review committees. Before we consider what M-CELS institutional review and oversight would look like, however, and who would be responsible for conducting it, we need to spell out the covering values that are necessary to make trade off decisions about design choices at the benchside.
Finally, fostering a new bioengineering ethics as I have sketched above will require high-level attention and dedicated resources from departments, universities, and funding agencies.  It is not enough to say that ethicists and bioengineers ought to work together during the life cycle of an M-CELS research project. There also needs to be multi-level institutional mechanisms and incentives to make these collaborations possible. For example, universities and funding agencies should allocate time and resources to enable ethicists, researchers, and trainees to collaborate in the manner that bioengineering ethics demands, and to reward their efforts by establishing norms and financial structures that recognize such work.
In closing, I believe these challenges, both philosophical and organizational, are tractable. Attention to these issues will aid the responsible development of M-CELS research. And the further development of bioengineering ethics could provide an appropriate model for the ethical conduct of other areas of science and engineering that depend on technologically and ontologically novel research entities for their advancement.
Insoo Hyun, PhD, Director of Research Ethics at the Center for Bioethics, can be reached at insoo_hyun (at) hms.harvard.edu
 Beebee, Helen and Nigel Sabbarton-Leary (eds.). The Semantics and Metaphysics of Natural Kinds. (Abingdon: Routledge, 2010).
 Aach, John, Jeantine Lunshof, Eswar Iyer, and George M. Church. "Addressing the ethical issues raised by synthetic human entities with embryo-like features." eLife 6 (2017). https://doi.org/10.7554/eLife.20674.
 Kamm, Roger D., Rashid Bashir, Natasha Arora, Roy D. Dar, Martha U. Gillette, Linda G. Griffith, Melissa L. Kemp, et al. "The promise of multi-cellular engineered living systems." APL Bioengineering 2 (2018). https://doi.org/10.1063/1.5038337.
 Cvetkovic, Caroline, Ritu Raman, Vincent Chan, Brian J. Williams, Madeline Tolish, Piyush Bajaj, Mahmut Selman Sakar, et al. "Three-dimensionally printed biological machines powered by skeletal muscle." PNAS 111, no. 28 (2014). https://doi.org/10.1073/pnas.1401577111.
 Personal communications, Insoo Hyun and the scientific participants of the EBICS Meeting, "Workshop on Multi-Cellular Engineered Living Systems" M-CELS Workshop 2021, Q Center, St. Charles, IL, August 2-4, 2018.
 Harvard University Embryonic Stem Cell Research Oversight (“ESCRO”) Committee. Ethical issues related to the creation of synthetic human embryos (Cambridge MA: Petrie-Flom Center, 2018). http://petrieflom.law.harvard.edu/resources/article/ethical-issues-relat....
 Kimmelman, Jonathan,. Insoo Hyun, Nissim Benvenisty, Timothy Caulfield, Helen E. Heslop, Charles E. Murray, Douglas Sipp, et al. "Global standards for stem-cell research." Nature. 533, no. 7603 (2016): 311-313. https://doi.org/10.1016/j.stemcr.2016.05.001.
 Hyun, Insoo. "Engineering ethics and self-organizing models of human development: opportunities and challenges." Cell Stem Cell 21, no. 2 (2017): 718-720. https://doi.org/10.1016/j.stem.2017.09.002.
 van de Poel, Ibo and A.C. van Gorp. "The need for ethical reflection in engineering design." Science, Technology, & Human Values 31, no. 3 (2006): 333–360. https://doi.org/10.1177/0162243905285846.
 Harris, Jr. Charles, Michael S. Pritchard, Ray James, Elaine Englehardt, Michael J. Rabins (eds.). Engineering Ethics: Concepts and Cases. 6th ed. (Boston: Cengage, 2014).
 Chang, Ruth. Introduction to Incommensurability, Incomparability, and Practical Reason. Edited by Ruth Chang. (Cambridge: Harvard University Press, 1997).
 Sample, Matthew, Marion Boulicault, Caley Allen, Rashid Bashir, Insoo Hyun, Megan Levis, Caroline Lowenthal, et al. "Multi-cellular engineered living systems: building a community around responsible research on emergence." Biofabrication 11, no. 4 (2019). https://doi.org/10.1088/1758-5090/ab268c.