Intrinsic responsible innovation in a synthetic biology research project

This paper presents, from the perspectives of both social scientists and microbiologists, a case study of the implementation and practice of Responsible Innovation (RI) in a UK-based synthetic biology project. We highlight the impact of interdisciplinary working and examine the benefits that arise from creating the time and space for shared reflection on research. Our discussions over the course of the project included concerns about the potential escape to the environment of laboratory-constructed genetic material and alternatives to the role that antibiotic resistance genes play in synthetic biology. As a result, the design of a key product of the project was altered in ways that go beyond normal institutional requirements. We highlight our view of the importance of continued interdisciplinary collaboration and the utility of the framework of Responsible Innovation in achieving this. We conclude with observations on the difficulties of sustaining such collaborations.


Introduction
The following paper describes how we, the authors, came together from different disciplinary backgrounds in a large synthetic biology project and researched the implementation of a framework for Responsible Innovation (RI).We detail how a key part of the project presented itself as a case study through which we could explore RI and the practical outcomes achieved.We also suggest that this collaborative paper is itself an example of RI in practice; evidence of how social scientists working alongside scientists in a project together create spaces for reflection on the nature of the scientific research, its goals and the means undertaken to achieve them.We will use our case study to explore some of the challenges each discipline faced and the implications these may have for Responsible Innovation in other projects.
The work we describe reflects the drive by UK funding bodies to find ways in which RI can be implemented in the natural and engineering sciences.In our case this context is working within a program grant "Synthetic Portabolomics: Leading the way at the crossroads of the Digital and the Bio Economies" (hereafter Portabolomics) funded by the UK's Engineering and Physical Sciences Research Council (EPSRC).Portabolomics incorporates strands of work in mathematics, computer sciences and microbiology, underpinned by research on RI and efforts to implement it across a research team of around 25 people.
The EPSRC publishes on their website (EPSRC 2018) a framework for RI given the acronym AREA.Funded researchers are encouraged to (A)nticipate, (R)eflect, (E)ngage and (A)ct as they conduct their research.In the course of our discussion, we will touch on all the underlying concepts of RI but focus in this publication on Reflection.
Before turning to the case study and our analysis, we provide some contextualizing discussion of synthetic biology and the development of the RI agenda.
Background: synthetic biology and socio-ethical issues in biotechnology Synthetic biology After two decades of development, synthetic biology remains a rapidly evolving field with several distinct identities and no settled single definition.There has been intense scrutiny of its development by social scientists, succinctly described by McLeod and Nerlich (2017).Synthetic biology can encompass a range of approaches, the most common three being, the building of "artificial" life from non-living components, the reduction of existing living "natural" cells to their minimal constituent parts, and the re-engineering of "natural" systems to perform tasks deemed by its proponents to be useful to humanity.Each of these approaches may be used to generate knowledge or to create biological products for industry or medicine.In their insightful analysis of the field, Schyfter and Calvert (2015) discuss how the latter approach above has come to dominate policy and funding, through the prominence of one group of synthetic biologists, whom they term "committed engineers".This group's view, that biology can be transformed into a discipline based on engineering principles, has influenced UK policymakers and funders, leading to synthetic biology being promoted as a significant potential contributor to the UK economy (Osborne [2012]; Willets [2012]) and, as Schyfter and Calvert note, focused on the production of economically valuable commodities.Portabolomics, as we will describe, occupies an interesting position in this landscape, being publicly funded yet ultimately concerned with aiding the bioeconomy by developing an enabling technology rather than specific commodities.between, these two approaches was described in depth by Owen and Pansera (2019), and so we need not reprise that here.As noted, Portabolomics is funded by the EPSRC, who developed a framework for the implementation of RI (Owen [2014]; Stilgoe et al. [2013]).This framework was given the acronym AREA (Anticipate, Reflect, Engage and Act) and laid out a means by which RI might be implemented in research projects.The EPSRC describe RI, . as a process that seeks to promote creativity and opportunities for science and innovation that are socially desirable and undertaken in the public interest, .which acknowledges that innovation can raise questions and dilemmas, .yet is often ambiguous in terms of purposes and motivations and unpredictable in terms of impacts, beneficial or otherwise, .and creates spaces and processes to explore these aspects of innovation in an open, inclusive and timely way (Adapted from EPSRC [2018] website).
Two of us (KT & SW) have written on the knowledge and views of experienced scientists about RI, drawn from our engagement and research in Portabolomics and synthetic biology more widely in the UK.An important point we discussed was that, in contrast to the ELSI approach of delegating socio-ethical analysis to others, the RI approach requires scientists themselves to take on the task of anticipating, reflecting on and engaging with the issues that their research raises for wider society (Stilgoe et al. [2013]; Macnaghten et al. [2016]).We found that our interviewees had limited knowledge or understanding of RI and that the ELSI paradigm was dominant in their thinking.This in turn led to a conflation of "responsibility" with "risk reduction".Considering how to change this situation, we noted that the EPSRC view was, and remains, that scientists may not yet be able to implement the AREA framework without assistance, For Responsible Innovation to take place in a meaningful way, it will be important that we and our researchers nurture and promote partnerships with other disciplines and spheres of expertise and facilitate training to enable these skills [required for RI] to be developed and taken forward.(EPSRC 2018) Our findings confirmed this view.The call by EPSRC to integrate social science and ethics in science projects has a long history, and in the case of synthetic biology has been a feature of projects since the field emerged.Balmer et al. (2015) describe how they, as social scientists, sought to engage with synthetic biologists and reflect on the typology of roles that were ascribed to them in different projects.Some of these are familiar from the ELSI framework, where social scientists or humanities scholars taking up such roles were seen as being "a representative of society" and hence "experts in the views of publics" (p9), and others that are more novel, such as "the wife" (p10), which involved taking on a role "managing the emotional labor of a collaborative project" or being seen as a "trophy [and a] symbol of ethical conduct in the synthetic biology research enterprise" (p10-11).Such role types reflect both the "continued legacy of ELSI logics and practices" (p8) and the misunderstandings arising from this form of interdisciplinarity.Glerup, Davies, and Horst (2017) reported from research on RRI with synthetic biologists and nanotechnologists that responsibility was a misunderstood topic.They noted that in one early conversation they were told "nothing really responsible goes on here" and continue, Our interpretation of this remark is that he in fact meant the oppositethat their scientific projects were very responsible.He considered his own and his colleagues' work of such quality that they simply did not need to talk about 'responsibility' and how to enhance it.(Glerup, Davies, and Horst 2017, 324) They describe "a set of 'bottom-up' practices of responsibility" and consider that this "is centred around responsibility for ensuring a robust scientific process" (p324-5).In other words, scientists felt a responsibility to conduct their science in a professional and technically excellent manner.The authors describe further enactments of responsibility, finding that the scientists they observed demonstrated through their actions both reflection and a willingness to engage with wider publics that go unrecognized as practices of RRI.Our work finds similarities several years on from their research but adds to their observations (i) the impact of having social scientists embedded in research projects and (ii) the value of adding time for RI to be considered within the regular project meetings.
With this background in mind, we turn now to our approach and the detail of the case study we chose to explore.

Methods and approaches to research
KT & SW write: The inclusion of social scientists in Portabolomics follows the established pattern of interdisciplinary working highlighted by Balmer et al. (2015).Over the course of our engagement with Portabolomics, we have sought to research how RI was understood by project members and use our findings to identify ways in which the principles of RI might be implemented across the various laboratories involved.As part of our research, we chose as a useful instrumental case study a piece of work by AJ & HM that was fundamental to Portabolomics.Crowe et al. (2011) describe instrumental case studies as those that use one particular example to gain insight and understanding of an issue or phenomenon and, … can be used to explain, describe or explore events or phenomena in the everyday contexts in which they occur.… In contrast to experimental designs, which seek to test a specific hypothesis through deliberately manipulating the environment … the case study approach lends itself well to capturing information on more explanatory 'how ', 'what' and 'why' questions … (Crowe et al. 2011, 4, emphasis in the original.) New Genetics and Society 5 These authors identify that, "each case should have a pre-defined boundary which clarifies the nature and time period covered by the case study (i.e. its scope, beginning and end)" (p5) and so AJ and HM's work appeared ideal as a clearly defined task within the larger project and, although their product was unique, the techniques and laboratory practices used were representative of the work of "wet lab" 2 synthetic biology.In this way the case could have something to say about wider issues.
The research was qualitative in nature and ethical approval for it was obtained from the Humanities and Social Sciences Faculty Ethics committee of Newcastle University. 3A total of 19 semi-structured interviews were conducted with experienced colleagues (computer scientists, synthetic biologists and microbiologists) working on Portabolomics, and synthetic biologists and geneticists working in other institutions around the UK.Interviews with scientists close to Portabolomics were conducted between October 2016 and February 2017; interviews with others were conducted in August 2018.Interviewees from outside the project team were identified in institutions that are engaged in synthetic biology research or apply its techniques to developments in agriculture.The interviews were aimed at gaining insights into the views and experiences of experienced scientists from different disciplines of (i) conducting synthetic biology research and (ii) the potential impacts that synthetic biology might have in the near term.The findings were used in both another publication and in reports to the Portabolomics management team and advisory board.While the interviews were wide-ranging in scope, some of the discussion was of relevance to aspects of the case study reported here and pertinent extracts and discussion will be presented.Interviews were audio recorded, were fully transcribed by a professional service and anonymized and deidentified (as far as possible) by KT.
Additional to the interviews, KT & SW participated in the regular scientific and managerial meetings of the Portabolomics team (including HM and AJ), 2016-20 and 2016-22 respectively, during which reviews of progress on the genetic engineering work were common.We were able to observe the interactions of the scientists involved and seek clarification on our understanding during the discussions.As noted, perhaps unusually, two of the scientists leading the genetic engineering work in Portabolomics are authors of this paper (AJ & HM) and as such add richness to the description and analysis of the case study.
KT and SW were able to initiate discussion of RI issues widely through having a standing item on the agenda of monthly team meetings.As part of the approach to RI, we used our presence at team meetings to introduce the Portabolomics scientists to the findings of social scientific research that they had not previously been aware of and discuss our reactions to the scientific work being reported.We were also able to make observations of the interactions of the research team in formal settings such as annual progress reviews with funders and the advisory board members, and in team workshops.
Analysis was conducted using a grounded theory approach (Silverman 2001) in which cycles of data collection and analysis enabled the construction of a more theoretical understanding of the phenomena observed.Such an approach is widely applied to the examination of complex social interactions, of which this type of laboratory science is one.By following the progress of the Portabolomics project KT & SW were able to develop an understanding of practices of responsibility (Taylor and Woods 2020) and the implications for RI.

Survey
In addition to the participatory and fieldwork elements of the case study, in 2019 KT & SW conducted a shortanonymoussurvey of all the Portabolomics team.This sought respondent's views on various aspects of our work to implement the EPSRC framework for RI.Both closed and free text questions were offered and while space considerations mean that detailed analysis is not presented here, though that analysis was reported to Portabolomics management, key findings are mentioned in our discussion below.

The case study
The commercial exploitation of synthetic biology research involves designing, making and testing "genetic constructs", pieces of DNA that will perform a particular task when inserted into an organism (most commonly microbes, i.e. bacteria and yeasts that geneticists term "hosts"), for example to produce a commercially useful organic molecule.Such microbes, e.g. the bacterium Escherichia coli, exist as many different "strains", each genetically distinct but closely related.Multiple problems can arise in the process of translating laboratorybased design and proof-of-concept work into a host organism suitable for commercial application.These problems are often related to the markedly different growth conditions required for large-scale synthesis of the desired product (see for example, Fillet et al. 2015) and can usually only be overcome by redesigning the genetic construct, requiring additional labor, and resulting in increased costs, delays in production and a lower-than-expected economic return.The Portabolomics concept addresses this "redesign problem" by proposing a standardized "Portabolomics plasmid", 4 capable of further modification to carry a desired genetic construct easily from bacterial strains used in the laboratory to different species of bacteria 5 that are used industrially, without the need for further engineering.The name Portabolomics is a neologism, conveying the desired goal of making a "portable" "genomic" solution to the redesign problem (see the website https://portabolomics.ico2s.orgfor details).
Portabolomics, if successful, will be a step in the development of an enabling technology for commercial synthetic biology.As such, there is no specific chemical product being generated that will impact on wider society as there might be in New Genetics and Society 7 other synthetic biology projects (e.g.menthol, Meckin and Balmer [2019]).That enabling technology will include the Portabolomics plasmid and our case study is its design and construction.

Some useful science
For the purposes of this paper and for the sake of clarity we will focus only on bacteria in the following description.The bulk of a bacterium's genetic code (i.e. the information that confers heritable traits) is contained in a long DNA molecule called a chromosome.A bacterium proliferates by growing and making a copy of its chromosome before dividing into two parts (known as daughter cells), each containing one copy of the original chromosome (Figure 1A).Chromosomes are large and fragile molecules that are challenging to isolate and manipulate in a test tube (in vitro).In the 1950s bacteria were discovered to contain additional, much smaller, circular, self-replicating DNA molecules termed "plasmids" (Zinder and Lederberg 1952) (Figure 1B).Multiple copies of a plasmid may exist in a bacterium at one time, and they are replicated along with the chromosome during cell division and growth.Plasmids are (relatively) simple to isolate and manipulate and thus are commonly used in molecular biology research as the means to carry efficacious genetic information, and for this reason are termed "vectors".Plasmids can be inserted into bacterial cells to carry out some function, for example to direct the expression of a gene encoding an enzyme that makes a useful biomolecule.In a laboratory context this is often a fluorescent protein due to its ability to be readily detected within a living cell (Figure 1B).
An essential element of a plasmid is a "replicon", a system required for initiating replication of the DNA molecule.Replicons often have two key components: an "origin" and an "initiator".The origin is a specific DNA sequence within the plasmid where DNA replication starts.The initiator is a gene which expresses a protein that acts at the origin, starting the process of DNA replication to make more copies of the plasmid (Figure 2A).Importantly, once the initiator protein commences the reaction, the host cell's DNA replication "machinery" is recruited to the plasmid to finish the synthesis of new copies.Because the plasmid replicon system interacts with the host's DNA replication machinery, as the host proteins evolve over time, so must the plasmid replicon.This co-evolution is highly specific for the partners involved and explains why plasmids are generally restricted to functioning in a narrow range of related microbial species and fail to survive outside this range.

New Genetics and Society 9
A problem of migration and containment AJ & HM write: The production of high value biomolecules such as enzymes and organic compounds is an important aim for biotechnology.In most cases, plasmids are engineered to carry genes encoding the proteins of a biosynthetic pathway that synthesizes a desired biomolecule.Through years of research, it has become clear that the efficiency of biosynthetic production pathways can be significantly influenced by the physiology of the host organism, and such "expression plasmids" generally only perform well in a narrow range of hosts.However, it is difficult to predict the optimal host for any given expression plasmid.Therefore, researchers often need to re-engineer plasmids each time they desire to test their biosynthetic pathway in a new host.The time and resource required for this reengineering exercise incurs significant costs.
The Portabolomics strategy to address this re-engineering challenge is to design a single plasmid replicon to work in a broad range of host microbes without requiring further modification.One solution to the problem of host specificity would be to generate a single plasmid containing multiple independent replicons, with the hope that at least one system would work in a desired host organism.Importantly, however, this approach poses a serious problem.If such a plasmid were to find its way into the environment, then there is an increased probability that it would eventually be taken up and replicated by another bacterial species, potentially with unknown or adverse consequences.Therefore, a different approach was required that would address the challenge of adapting a plasmid to "host specificity", while also ensuring that it could not be functionally maintained if released into the environment.
One solution that addresses this problem is to design a replicon system that separates the replication origin from the replication initiator.An early decision in the design of the Portabolomics plasmid was to remove the initiator gene from the plasmid, and instead express it from the chromosome of various industrially important host bacteria (Figure 2B).This approach had two key benefits.First, this allowed us to tailor expression of the initiator gene to each host by modifying its expression signals (e.g.transcription and translation, the activities required for a cell to synthesize a protein from a gene).Second, this physical separation creates an innate biocontainment mechanism for the system.Because the DNA to encode the initiator protein has been removed from the plasmid, should the plasmid ever move accidentally to a different host species, it would not contain the information to initiate its own replication and therefore will not be passed to daughter cells (Figure 2C).This approach is an example of the biological containment envisioned in the recommendations made by the Asilomar conference.
KT & SW write: A few months into our involvement in the project we reached a view that the Portabolomics plasmid could be a potential concern.Our developing understanding of the concept led us to think that the Portabolomics plasmid could quite easily be accidentally transmitted into organisms for which it was not intended.Designing a plasmid that would easily move between different bacterial species was, after all, the project's objective.However, when we raised this in a meeting, colleagues made it clear that this had been anticipated by the microbiologists (led by HM) early in the design process and steps taken from an early stage to mitigate this possibility of accidental transfer as described above.
Interestingly for those engaged in the study of responsible innovation, this approach was deemed necessary and put into practice by HM and other Portabolomics scientists without input from social scientists.We suggest that this can be seen as an example where elements of RI practice, Anticipation and Acting, were undertaken by the scientists, though at a time when, we have shown, there was a low level of understanding of the AREA framework (Taylor and Woods 2020).
The containment problem lies in a "comfort zone" for laboratory-based scientists, the area of risk reduction, which is often considered by scientists to be a major public concern.Our interview data is consistent with other findings that biologists' understanding of uncontrolled gene transfer as a public concern is often based on reactions to the genetically modified (GM) crops debates of the early 2000s.However, it is well known that geneticists misunderstand the nature of public opposition to GM crops (Marris [2001]; Marris and Calvert [2020]), and so continue to focus on risk reduction in the hope that this will lead to increased public trust in science.
This was one area in which our regular participation in research team meetings and presentations was helpful in raising awareness.We found that project colleagues were unaware of Claire Marris' early work on the "myths" created by some scientists of public concerns about GMOs and her subsequent writing on the persistence of such thinking among synthetic biologists (Marris 2001;Marris 2015;Marris and Calvert 2020).Presenting a summary of that, and other work in this area, led to a lively discussion.It is a matter of regret that funding did not permit the planned re-interviewing of research team members four years into the project to gauge the impact this had.
While the Portabolomics research team anticipated a potential containment problem and set out to address it, this was done despite a commonly held view that engineered bacteria derived from common laboratory strains would not survive in the wild.
Probably a micro-organism engineered in the lab and by mistake released out would be outcompeted, straightforwardly outcompeted (Interviewee 1 ).
Another interviewee noted, So it's really very difficult to engineer a new bacterium, to engineer to change the DNA and to make sure that DNA stays from one generation to another to another ).

New Genetics and Society 11
A further consideration was that the bacteria being used in the work would be, … crippled in so many ways because they're lab strains and they've been living in the lap of luxury on our agar plates … they probably wouldn't [survive] that long  Despite this, concerns were not fully alleviated as, "it's this concept of the mobile element, it has that little bit of risk to it" (Interviewee 6 ).
Thus, even though bacteria containing the Portabolomics plasmid would likely not survive in the wild, there remained a possibility that the plasmid could move to environmentally adapted viable bacteria and escape into the environment through that route.The expense in time, effort, and money to mitigate this possibility was considered worthwhile.
The strategy chosen to avoid the accidental release of the Portabolomics plasmid (and whatever genes it is engineered to carry) can be traced back to the 1975 Asilomar conference and recommendations for biological containment.Asilomar participants, … agreed that most of the work on construction of recombinant DNA molecules should proceed provided that appropriate safeguards, principally biological and physical barriers adequate to contain the newly created organisms, are employed (Berg et al. 1975(Berg et al. , 1981)).
Berg et al. also described biological barriers as being of two types: (i) fastidious bacterial hosts unable to survive in natural environments, and (ii) nontransmissible and equally fastidious vectors (such as plasmids) able to grow only in specified hosts.The approaches to these barriers became, and remain, standard laboratory practices.We will return to this point.While the details of approaches to biological containment may be complex to non-specialists, we hope that the simplified description provided above gives a flavor of the type of manipulations required to finalize the Portabolomics plasmid design, which extend beyond the standard that would be ordinarily expected.
A problem of antibiotic resistance KT & SW write: While an appropriate mechanism for containment of the Portabolomics plasmid was considered at the outset of the project, this was not the case with the issue of antibiotic resistance.We raised this subject in a Portabolomics team workshop in 2017 and this serves as an example of how the routine considerations of RI issues and open discussion amongst colleagues identified a societal concern and how this could be addressed.
The use of antibiotic resistance genes in microbial genetic research was considered unremarkable by the research team as it is a routine and long-established practice in laboratories.It was one, however, that to our "lay ears", against a background of ongoing public concern about antibiotic resistance, sounded, if not alarming, then at least likely to be concerning to non-scientists.
AJ & HM write: As we discussed above, plasmids are the vector of choice for a significant proportion of synthetic biology research.The DNA sequence of a plasmid can be manipulated to incorporate new genes that will perform a desired function.Such manipulations are done in vitro, outside the bacterial cell.Of course, to make the desired product the plasmid must be introduced back into a host cell.We need not discuss the details of this process ("genetic transformation") here; however, it is important to note that it is generally an inefficient process, and a method is required for selecting those rare bacteria which contain the engineered plasmid from the majority that have not taken it up (Figure 3).
During engineering of the Portabolomics plasmid, this critical selection process was achieved by the use of genes which encode enzymes that render the host cell resistant to a specific antibiotic (i.e.only cells containing the Portabolomics plasmid can grow in the presence of the antibiotic that normally kills the bacteria).Using antibiotic resistance genes is standard laboratory practice; it is simple and cheap to implement, and there are sets of antibiotics paired with resistance genes available for a broad range of microbes.
Whilst this is an appropriate method for selection in a laboratory environment, discussions in the Portabolomics team meetings highlighted that the proliferation of an antibiotic resistance gene within the environment posed a potential biomedical risk and added to societal concerns about increases in antibiotic (or antimicrobial) resistance generally.Therefore, it was agreed that the use of antibiotic selection in the final Portabolomics plasmid design would not be appropriate.
The growing problem of antimicrobial resistance in human pathogens has been a subject of much recent concern (Berendonk et al. 2015).While the biocontainment strategy restricts the Portabolomics plasmid from freely replicating and propagating in an unintended host, it may yet be possible that the genetic information contained on the plasmid could be passed to a foreign organism (i.e.fragments of the plasmid DNA are captured and integrated into host genome, thus ensuring their stable inheritance).If the foreign bacteria happen to be pathogenic, then the newly acquired antibiotic resistance gene would confer protection.It has emerged that many bacterial pathogens have multiple antibiotic resistance genes, making them nearly impossible to kill using current approaches in modern medicine.
Therefore, rather than using antibiotic selection in the final design for the Portabolomics plasmid, a different selection method is under examination, auxotrophy; when an organism cannot synthesize an essential biomolecule required for its growth.In this approach, bacterial hosts for the Portabolomics plasmid will be genetically modified to disrupt the ability to produce an essential metabolite (e.g. an amino acid, the building block of proteins).To survive, these "auxotrophic" bacterial strains need to be provided with the key metabolite in their diet.Therefore, a functional copy of the genes that synthesize the essential metabolite will be integrated into the Portabolomics plasmid.Thus, when seeking the bacteria that have taken up the Portabolomics plasmid we grow these microbes in conditions that lack the essential metabolite, and only those bacteria that New Genetics and Society 13 Figure 3. (A) Bacterial transformation results in a mixed culture of cells, with a minority of cells carrying the plasmid (purple cells) and many without (green cells).These are spread over appropriate media and allowed to grow.In the absence of selection, a "lawn" of mostly untransformed cells will appear.(B) When antibiotic selection is used, if untransformed cells are plated to media with antibiotic no bacteria will grow.(C) However, if transformed cells are plated media with antibiotic discrete colonies of plasmid bearing bacteria are observed.
have taken up the Portabolomics plasmid will be able to produce the required metabolite and survive.
KT & SW write: The routine laboratory practice of using antibiotic resistance as a selection mechanism in genetic manipulation of microbes is so embedded that only three of 19 participants in our interviews raised the subject.For example, relating to Portabolomics, KT: … on that subject, just thinking about antibiotics, there was talk last week about antibiotic resistance genes.Interviewee 6: It's a tricky topic, yeah.In the lab, yes we use a whole range of antibiotic markers in our pet strains of Bacillus subtilis.Now no one is going to care [or see] relevance to it becoming multi-drug resistant, we've got them in the lab because it's non-pathogenic.But you create something that can mobilise those antibiotic markers around, then it could jump into a pathogen ).
Added potential complexity around this subject were also raised by this interviewee who said, The one that's fun really is the resistance markers already exist in nature because, as I've said, most of the antibiotics we've got are made by bacteria  emphasis in speech).
While the Portabolomics plasmid is being developed antibiotic resistance genes will continue to be used, as these samples are intended to remain contained in a laboratory setting.However, two interviewees who work in other areas noted that should any deliberate environmental release be contemplated, regulations would enforce the removal of those resistance genes.
KT: … do you think these new techniques raise any new issues that weren't there with older GM … ?Interviewee 17: I think they remove a lot of issues so I mean we talk about, in the past we've always used antibiotic resistance or some kind of resistance to a herbicide or whatever to select for lines that are transgenic, so that are GM, and we still do that with CRISPR/Cas but then we get rid of that.You get rid of the thing that's causing the change so it's no longer got any genes which confer resistance to antibiotics or to herbicides or anything and that would be a very stringent step before anything would be released anyway.Interviewee 18 was rather outspoken when discussing regulations, It's pretty horrendous!It's extremely expensive and extremely longwinded.… these risk assessments have been carried out for a long, long time now, so-but there has to be an environmental risk assessment carried out … should some of these genes escape, for example, um, antibiotic resistance marker genes was a problem for a long, long, long, long time except nobody really uses them anymore [in plantbased work] (Interviewee 18 [67][68][69][70][71][72][73][74][75][76][77][78][79][80][81][82][83][84], emphasis in speech).
It is perhaps telling that these latter two interviewees work in areas that include the release of genetically modified organisms outside laboratory settings.Given that

New Genetics and Society 15
Portabolomics aims to provide an enabling technology for industry, we consider that their views are a useful insight into aspects of work that may arise in future, and they validate our suggestion to find an alternative to antibiotic resistance genes for the Portabolomics plasmid.While regulation may seek to prevent the spread of antibiotic resistance genes via deliberate release, it remains the case that accidental release cannot be legislated against and so the actions undertaken by AJ, HM and colleagues should be seen in this context.

Findings and discussion
Responsible Innovation (RI) was developed as a concept in which scientists are encouraged to consider the social consequences of their work.However, the required reflexivity is not a skill that scientists will likely have acquired.For example, in early discussion HM told KT, … I'm very much enjoying thinking about these new things.… I think those are things I would like to learn more about.So all I would say is because you're an expert, if you want to talk to somebody about DNA replication come to me and I can explain that to you in greater detail.If I want to learn something more about my ethical responsibilities, I guess I'd like somebody like you, an expert, to come and tell me you really need to consider this, this and this.(Interview October 2016).
For that reason, the Portabolomics program of work was designed (by a small group of synthetic biologists) to include an RI strand, run by social scientists.One result of the subsequent collaboration is the work described here.
The case study described above has offered an insight into some of the working practices and thought processes of the scientists involved.Although not recognized by colleagues as "Responsible Innovation", the practices of responsibility demonstrated nonetheless exemplify aspects of reflection and anticipation as defined in the funder's interpretation of RI.While the problem of containment is an example of scientists' own reflection on their workgoing beyond what is necessary for routine safetythe antibiotic resistance problem would not have been considered or addressed without the time and space for reflection being made available that resulted from embedding RI in the project meeting agendas.
Unlike Glerup et al. (p330) KT & SW did not find project colleagues to be hostile to the language or ideas of RI.However, it became clear that there was a difference in engagement with the ideas by those at different points in their career.In our survey, early career researchers (ECRs) reported a general feeling that RI was something that they would begin to consider once more established in their careers.Their present focus was on demonstrating their scientific capability and gaining the experience necessary to advance their careers, mirroring the findings of McLeod, de Saille, and Nerlich (2018).However, all respondents reported a desire to engage in continued discussion and recognized the value of having time away from normal laboratory work to do so.
Glerup et al. made several suggestions for the successful implementation of RRI in science, one being, If RRI is not to be seen as an additional burden or imposition, it has to be acknowledged that any kind of scientific reflection or wider engagement takes planning, time and skills.RRI thus needs to be embedded within systems through which the labour it entails can be acknowledged and valued.(p330) We report here an instance of this recommendation in practice in a UK (RI) context.Indeed, we meet the suggestions they also make for the inclusion of social scientists long-term in projects and bringing to the attention of scientists how the "official" language of RRI (or RI) might mirror the practises they already undertake, thus providing an opportunity to draw attention to and build upon these practices of responsibility.
Both Glerup, Davies, and Horst (2017) and this present work identify scientists conducting their research in a "responsible" manner and so we considered whether Randles (2017) concept of "de facto rri" might better describe this phenomenon.
De-facto responsible research and innovation (or rri) refers to what actors already do, in collective fora, in order to embed institutionalised interpretations of what it means to be responsible; into the practices, processes organisational structures and outcomes of research and innovation.(Randles 2017, 20) While de-facto rri may describe the standard laboratory practices around containment and disposal procedures, critically, what we describe for Portabolomics differs, in that concerns regarding accidental release were taken account of and acted upon in a manner that goes beyond what would be considered reasonable in other projects.This decision was taken precisely because the Portabolomics plasmid was being designed to be easily moved between different species of bacteria and the recognition of the potential problem resulted from the scientists' own reflection.Further, the widespread and unremarked use of antibiotic resistance genes in bacterial genetics research does have some institutional control, in relation to record keeping, but the research team's consideration of replacement with an alternative approach is not a general requirement or consideration.Nor would the approach described here be likely to become institutional practice, given the effort and expense involved.We also note that we, as biologists and social scientists, worked together to ultimately change the design of the endproduct plasmid regarding antibiotic resistance genes and therefore suggest that the practices of responsibility identified differ fundamentally in nature from defacto rri.While the institution embeds requirements for laboratory design and working practices based on previous generations of scientists' interpretations of responsibility (e.g. in regard to the use of autoclaves to sterilize materials before disposal), this case study has shown that laboratory practices can extend beyond those embedded requirements when the trajectory of the work is reflected upon.In other words, the individual scientists have a clear idea of what it means to be New Genetics and Society 17 responsible in ways that go beyond the routinized culture of health and safety practices and requirements.In undertaking practices of anticipation and reflection, identifying the potential for societal harm from the accidental release of the Portabolomics plasmid, the scientists have been performing the actions required by the RI framework, without identifying them as such.
We note that one key aspect of RI, engagement in public dialogue, is missing from this account.While public engagement activities were undertaken, and will be reported on elsewhere, any societal benefit of the direction of the research was taken for granted in the design and funding of the project.Also, the findings of the engagement activities had no subsequent effect on the progress of the Portabolomics project, though individuals involved did show a deeper subsequent engagement with the ideas of RI.
The issue of the use of antibiotic resistance genes is more complex than accidental release.As noted above, the Asilomar conference was a pivotal juncture in the early days of genetic science.One of the key aspects of recombinant DNA technology for which they sought the 1974 moratorium was the use of antibiotic resistance genes.In particular, they suggested avoiding, Construction of new, autonomously replicating bacterial plasmids that might result in the introduction of genetic determinants for antibiotic resistance or bacterial toxin formation into bacterial strains that do not at present carry such determinants; or construction of new bacterial plasmids containing combinations of resistance to clinically useful antibiotics unless [these] already exist in nature.(Berg et al. 1974, 303) However, the conference final report and recommendations only mentioned the use of antibiotic resistance genes in connection with experiments involving human pathogens (Berg et al. 1975(Berg et al. , 1983)).In the time between the Asilomar conference and now, the use of antibiotic resistance genes in molecular biology (including synthetic biology) has become ubiquitous, and unremarked upon by practitioners.These genes provide a cheap, easy and reliable way to carry out the routine work of moving plasmids into bacteria (Figure 3).While it is generally the case that antibiotics used in synthetic biology are no longer currently common in clinical use (for example kanamycin), it is also true that some, such as chloramphenicol, are still in use both as front-line medication in low-income countries and in specific applications around the world, e.g.chloramphenicol is an active ingredient in over-the-counter eye drops in the UK.Therefore, reducing the spread of genes which confer resistance to clinically efficacious antibiotics is an important consideration.
We suggest that one of the important lessons from our case study is that the implementation of Responsible Innovation requires ongoing reflection and action by scientists, even when a problem has been seemingly "fixed" by historical precedent, institutional practice, and legislation.Societal concerns about antibiotic resistance genes may not be in the forefront of the minds of scientists working at the bench and using a tried-and-tested laboratory technique acquired while they were undergraduates, however, as we have shown, scientists are willing to reflect upon, discuss and address problems such as these when given the time, space and skills to do so.Pansera et al. (2020) report on their experiences of the implementation of RI in synthetic biology, working in the context of a research center in Bristol with multiple projects and in an institutional setting that has a long tradition of public engagement activity.Our case differs in certain respects but iswe believecomplementary and adds to those authors' arguments and observations.The starting point for the inclusion of RI research and practical work in the Portabolomics program of work was a requirement in the conditions of the award.In common with the Bristol setting, our institution has expounded a desire to engage as a "Civic University" (Goddard et al. 2016).In addition, SW was Director of a Research Center which had over two decades of experience in research, public engagement activities, and collaborations across multiple interdisciplinary projects.The strategy adopted by KT & SW in Portabolomics did not, as in the Bristol case, begin with a focus on public engagement but rather with formal and informal engagement with the scientists on the technicalities of the research.Pansera et al. report that they adopted "formal and informal encouragement rather than coercion" (p399; emphasis in original) in their approach to embedding the concepts of RI in the center.This is also the approach taken within Portabolomics; avoiding top-down or bureaucratic imposition of activities to be undertaken or reports to be presented.
A formal approach to the discussion of social and ethical issues raised by the work on Portabolomics was attempted, originally designed by the scientists in the project's development phase; the Responsible Innovation Directorate (RID).The RID was a group that provided a forum for the discussion of issues brought to its attention by team colleagues.It consisted of KT and SW with four laboratory leaders from across the project together with the project manager and an early career researcher representative.Two members external to the project were also appointed.The working method adopted was that all Portabolomics laboratory leaders were asked to set aside 10 min of their team meeting every month or two as time to discuss RI.Further, any colleague with an issue to raise could contact any RID member at any time.A small number of individual laboratory meetings were observed by KT and this, together with later discussion with colleagues, revealed that the approach did not appear to be working.In the three years of its operation (2017-19) only one person (HM) suggested issues for the RID to discuss, and one RID member even confessed before the final meeting to having forgotten about their membership of the body.While the authors' participation in, and observations of, the project do not immediately lead us to believe that there are additional serious biological, safety or societal concerns to be raised, we had hoped for better engagement with the RID process.

New Genetics and Society 19
In relation to antibiotic resistance genes, our case study provides some evidence of the impact of RI on laboratory practice and the direction of the research; something that Pansera et al. found difficult to identify in their situation.We agree that the tacit, cognitive and intangible nature (p400, italics in original) of the results of engagement with RI are difficult to capture.But we have reported above on the impacts that the discussions amongst the authors and other Portabolomics scientists have had on the direction of the research.We recognize that the practices of responsibility carried out in the day-to-day work of the laboratory, examples of the tacit knowledge transmitted from experienced scientist to less experienced, are difficult to measure but open discussion and curiosity on the part of all colleagues can bring such practices to light.
An important observation we have made is that when time and space are given to discuss issues of responsibility, almost all of the scientists we have worked with have proved more than willing to engage in productive conversations.While experienced scientists mostly contributed to discussions, there was a reticence among the ECRs to contribute during meetings when senior staff were present.We found this reticence was lessened when amongst their peer groups, perhaps in a similar way to the experiences of McLeod, de Saille, and Nerlich (2018) in their research.We gave the example above of how the Portabolomics team were exposed to the work undertaken by Marris.On other occasions, it became clear that colleagues were unaware of the concept of a "deficit model of public engagement" and were also unaware of the literature on dual use of biotechnology (e.g.Marris, Jefferson Lentzos [2014]).KT and SW were able to lead some discussion of these issues and KT worked with some early career researchers on some successful public engagement activities that proved to be educational for all involved and will be reported elsewhere.
The inclusion of RI as a standing item on Portabolomics meeting agendas gave team members the time and space not normally available for reflection on the issues raised by the direction of the research.We found in practice that little encouragement was needed as project colleagues seemed to relish the opportunity to discuss these aspects of their work.It remains a disappointment that funding for KT was not sufficient to enable a second round of interviews with the experienced scientists engaged on the project, that would have explored their views, and especially changes in opinion, of synthetic biology resulting from three or more years' engagement in the project.
The stories of this case study detailed above would likely have remained confined to the group of scientists involved in the Portabolomics work program, without the conversations that arose from the routine inclusion of RI as an agenda item in team meetings.Furthermore, while the details of the problems and the proposed solutions might have been made available publicly in technical descriptions of the finished Portabolomics plasmid, the more likely case is that the reasons for identifying these problems and their root causes would not have been considered relevant in scientific papers.By encouraging the discussion among colleagues, and indeed in the writing of this paper, we believe this has produced an important output from this interdisciplinary work.
We believe the "hidden" work of anticipation, reflection and action could be usefully brought to a wider audience in considerations of public trust in science.The responsibility of individual scientists is rendered invisible by the everyday, pragmatic, considerations of how to go about their work in a professional manner; something recognizable as phronesis, the demonstration and practice of practical wisdom.The recognition by both society in general and the scientists themselves that RI is something that is not an "add-on", "outsourced" to specialist ethicists but is an intrinsic part of what it is to do science well could go some way towards moving forward on public debates that are provoked by science.
We all continue to believe that it is important for social science and humanities researchers to work with natural and computer scientists.There is certainly a role to be played in capacity-building amongst early career researchers and postgraduate students.To that end, RI has been added to the synthetic biology teaching agenda (Hallinan et al. 2019), and KT & SW have provided teaching to MSc students and Newcastle University's iGEM team.
For other projects in biosciences and other disciplines there appears to be a clear benefit in establishing RI discussion in routine project meetings where space and time can be made in otherwise busy schedules to permit the open exchange of ideas on social, ethical and other issues among colleagues who may otherwise not make the time to do so.While further research and engagement with RI is clearly needed, we hope that we have been able to add to the growing literature on real-world examples of its implementation.

Postscript 2023
Finally, we should note that this paper has been in preparation for several years, reflecting the difficulty of the biological and genetic engineering work and the need to publish the technical material in a scientific journal before discussing it here.In that time, we have all ended our participation in the Portabolomics project.HM's involvement ended with completion of the construction of the basic Portabolomics plasmid containing the genetic changes necessary for containment.He now focusses on fundamental research on DNA replication and reports little current need to consider RI issues.AJ has moved to a non-research position but reports lessons learnt and maintaining an interest in RI should his work require it.These responses perhaps reflect the realities of the lack of ongoing interdisciplinary work.SW has retired but maintains an interest in RI through an Emeritus position with the University.Funding decisions meant that KT's work ended, though the scientific work deemed necessary for the Portabolomics concept continued.
At the time of writing (mid 2023) the Portabolomics plasmid is not yet working in different strains of bacteria as expected.Although auxotrophy remains part of the plan, it may be some time before its implementation in the laboratory as the New Genetics and Society 21 use of antibiotic resistance is such an accepted practice in contained conditions.In this regard, while we would claim that Portabolomics colleagues acted in the case of constructing the plasmid with the separation of the replication mechanism as described above, we perhaps cannot claim that action has resulted from the consideration of auxotrophy.While a design is ready to replace antibiotic resistance genes in the Portabolomics plasmid when it is trialed in a commercial setting, it remains to be seen whether commercial interests will bear the necessary costs of auxotrophy's implementation, even under contained conditions.It may be that the institutionalized practice of using antibiotic resistance genes is too culturally embedded, convenient and cost-effective to be overcome.
The realities of the implementation of RI remain that it is subject to the vagaries of funding.Present models do not support long-term interdisciplinary engagements, should social scientists wish to participate (see Marris and Calvert 2019 for a deeper discussion on that point).However, we all agree that the effort would be worthwhile, as would providing teaching on RI at undergraduate level be, as an essential step in developing a culture of RI in a future cohort of researchers.Notes 1.Recombinant DNA technology is the process of joining together DNA from different sources, including different species of organisms, to make a new DNA molecule that can then be put into a new "host" organism, commonly a microbe.Such DNA molecules may be termed "genetic constructs" or often in synthetic biology, using the language of engineering, "circuits".2. Typically, synthetic biology is divided into "wet lab" and "dry lab" where wet refers to the work of microbiology or other biochemical practices and dry refers to the computer-based aspects of synthetic biology such as design and modeling.3. Approval reference 8635, 2016; 12/10/2016.4. The Portabolomics plasmid is based on a commonly used bacterial plasmid (see main text) that project colleagues have experience of working with successfully.5. Research laboratories typically use well known strains of E. coli or Bacillus subtilis for their work, however industrial processes are often carried out using different strains of these bacteria or entirely different species such as Corynebacterium or Clostridium.

Figure 1 .
Figure 1.(A) Overview of bacterial cell cycle.Figure 1(B) Summary of some of the key biological processes occurring within a bacterial cell required for propagation and expression of genetic information.

Figure 2 .
Figure 2. (A) In replication of a "wild type" plasmid a replication initiation protein (Rep) is produced from the plasmid (light blue) and acts at a site known as the origin of replication (green) to recruit host factors and drive replication of the plasmid.(B) In the engineered "Portabolomics plasmid" system the rep gene is removed from the plasmid and expressed from the host chromosome.The origin is maintained on the plasmid and plasmid replication can commence in the same manner as the parent plasmid.(C) Were the Portabolomics plasmid allowed to escape and be taken up by environmental bacteria, the plasmid would be unable to replicate due to lack of Rep.