Biotransformation of p-xylene into terephthalic acid by engineered Escherichia coli

Terephthalic acid (TPA) is an important industrial chemical currently produced by energy intensive and potentially hazardous p-xylene (pX) oxidation process. Here we report the development of metabolically engineered Escherichia coli system for biological transformation of pX into TPA. The engineered E. coli strain harbours a synthetic TPA pathway optimized through manipulation of expression levels of upstream and downstream modules. The upstream pathway converts pX to p-toluic acid (pTA) and the downstream pathway transforms pTA to TPA. In a two-phase partitioning fermentation, the engineered strain converts 8.8 g pX into 13.3 g TPA, which corresponds to a conversion yield of 96.7 mol%. These results suggest that the E. coli system presented here might be a promising alternative for the large-scale biotechnological production of TPA and lays the foundations for the future development of sustainable approaches for TPA production.

products formed is much higher than that of the educt consumed, there must be a mistake. 6. L.147-L.180: though I appreciate that the authors tried several gene/promotor arrangements in their expression plasmids, this part needs to be reduced to at least 50% (better even less). The text is not much more as what can be seen in Fig. 2b/Fig. legends. 7. L.182-187: it is trivial that TA is present as dianion at the pH of culture conditions: reduce this statement to a minimum 8. L.196-L.243: Again this part reporting on the optimization of the fermentation should be shortened to 50%, the authors should rather report on the best conditions and then summarize what parameters wereessential.

Reviewer #2 (Remarks to the Author):
Our society is totally dependent in a multitude of ways upon chemicals provided by the chemical industry. However, efforts to achieve sustainable development necessitate departure from polluting and high level energy-consuming processes based on non-renewable feedstocks, through exploitation of environmentally friendly chemistry options. A main plank in sustainable chemistry procedures is biocatalysis. Exploitation of biocatalytic procedures requires that they be economically competitive. Significant hurdles for biocatalytic processes for bulk chemicals are modest yields and susceptibility of the biocatalyst to substrate-interemediate-product toxicity.
In this paper, the authors propose a (ultimately) biocatalytic route from the renewable substrate glucose to terephthalic acid (TPA), the monomer of the commodity polymer PET, though they describe here an intermediate process of conversion of p-xylene to TPA, as a proof-of-principle.
The results presented constitute an elegant biocatalysis process based on a rationally designed metabolic pathway constructed in E. coli from the upper xylene degradation pathway of Pseudomonas putida F1 and the lower p-toluene sulfonate pathway of Comamonas testosteroni T-2. The biocatalytic activity of the construct was assessed in a 2-phase bioconversion system in which the xylene was supplied in the apolar phase. Initial experiments revealed bottlenecks within the designed pathway and several inducible constructs were made to relieve them and optimise product yield. The potential toxicity of TPA caused by high substrate feeding rates was analysed and shown to be unproblematic. In one case, a yield of 97% was recorded, which is extraordinary and, if reproducible when upscaled, will represent a breakthrough in hydrocarbon bioconversions.
All in all, this is an excellent and important paper documenting the feasibility of bioconversions for bulk chemistry.
While not yet documenting a process of conversion of a renewable substrate to TPA, it shows proof-of-principle, since other publications show the potential feasibility of conversion of renewables to p-xylene. In any case, the exploitation of petroleum constituents as feedstocks for chemicals production, despite being unsustainable over long time spans, is, unlike the use of fossil fuels for energy generation, not a sustainability priority, so a scale-up of the process described in this submission, and its implementation as a viable process for large scale production of TPA, may be envisaged.
I congratulate the authors on this important advance and only have one, rather trivial question: they examined the toxicity of TPA applied externally but, in this process, TPA is produced internally. Perhaps they may wish to comment on this, and on TPA release after formation?

Reviewer #3 (Remarks to the Author):
In this manuscript, the authors describe the development of an E. coli strain that can convert pxylene (pX) to terephthalic acid (TPA). The engineered pathway relies on known degradative routes for p-xylene and p-toluene sulfonate. The authors first demonstrate independent validation of the up-and downstream pathways, then demonstrate a two-phase partitioning bioreactor to achieve conversion of pX to TPA. This is an interesting, largely well-written manuscript. The suitability for this journal arises from the nature of the product being targeted -a very highvolume chemical compound -and the potential for achieving completely bio-based production of TPA. The manuscript would be considerably strengthened by addressing the comments below to improve clarity: 1. p. 4, lines 73 and 74 -I don't think it's quite correct to start that there were "detailed studies" on enzymes in this paper. The pathway relies on known pathways and largely confirms previous observations rather than providing any new biochemical insights. The authors are better served by limiting this statement to "pathway reactions" since that is consistent with both the context and the results.
2. p. 5, lines 102-104 -As written, this sentence suggests that BADH does not use NADH and thus would result in a cofactor imbalance. According to Metacyc, the cofactor requirements for BADH are the same as for the second oxidative step of XMO. Thus, I don't see what using XMO provides an advantage with respect to redox balance.
3. p. 6, lines 121-122 and Supp Fig 3a -Is the production of pTA in this figure actually significant? It seems that a small amount may have been detected at 90 min, but without error bars, it's impossible to know if this is "real"? Were there biological replicates that consistently showed pTA production? Any showing accumulation, i.e., detection at more than one time point? 4. p. 9, line 183 -suggest changing "has not been known" to "has not been evaluated" or something similar 5. p. 9, lines 187-194 and Supp Fig 6 -The data do not support the conclusion that there is no effect on growth at TPA-salt concentrations up to 10 g/L. There is a clear difference in max OD between the 0 g/L control and the 5 and 10 g/L samples. It may be true that the growth rates during exponential phase are not different, but there are not enough data points to accurately calculate a growth rate given that there is only one data point between inoculation at time t=0 and stationary phase. If the authors want to make claims about growth rate, more data points are needed to ensure that measurements are being made during the exponential phase.
6. p. 15, line 324 -The amount of substrates added need to be provided with more detail than "an appropriate amount." For each experiment, the specific substrate and amount added should be provide. A similar comment is provided below in reference to Supp  show productivity originating from pX, but pX titer is not shown anywhere. I understand that the solubility is quite low and thus it may have been unmeasurable, but if this is the case, it should be noted in the figure legend and/or the methods. Panel 3a also shows spikes in metabolites, which suggests some sort of feeding regimen. It would also be very helpful to use the same abbreviations (in parentheses) in the figure legend as are used in the main text. Table 5 is incorrectly listed as Supp Figure 5. Also, the units for pX added (superscript note 'b') need to be provided.

Responses to the Comments Editor's Comments
Dear Dr Lee, Your manuscript entitled "Biotransformation of p-xylene into terephthalic acid by engineered Escherichia coli" has now been seen by 3 referees. You will see from their comments below that while they find your work of interest, some important points are raised. We are interested in the possibility of publishing your study in Nature Communications, but would like to consider your response to these concerns in the form of a revised manuscript before we make a final decision on publication. We therefore invite you to revise and resubmit your manuscript, taking into account the points raised. Please highlight all changes in the manuscript text file.
We look forward to seeing the revised manuscript and thank you for the opportunity to review your work.
Best regards, Chuanfu An, Ph.D. Associate Editor, Nature Communications [Response]: Thank you very much. Although the reviewers' comments were rather minor, they were invaluable in making our paper much clearer and improved. We would like to thank you and the reviewers for the comments.

Reviewers' comments:
Reviewer #1 (Remarks to the Author): The manuscript by Luo and Lee reports on the biotransformation of terephthalic acid (TA) from p-xylene by heterologous gene expression in E. coli. TA is a globally important, largevolume commodity chemical that is known for its use in PET polymers. Usually, it is chemically produced from fossil-oil derived p-xylene at high temperatures and pressures. For this reason there will be a demand for biotechnological alternatives for PA formation in the future. While recent progress has been achieved in the sustainable synthesis of p-xylene from biomass-derived sources, attempts for the biotransformation of TA from p-xylene have not resulted in satisfying yields, yet.
The authors impressively achieved the production of TA from p-xylene with a yield of 96.9 mol% (13.3 g TA from 8.8 g p-xylene) by expressing seven genes involved in the oxidation of methyl groups to carboxylic acid functionalities. The yield is without doubt a milestone in the field and of great interest for researchers studying sustainable alternatives for petroleum derived chemicals. The authors designed a smart redox-neutral pathway comprising the combined action of oxygenases and dehydrogenases, and provide a number of optimization steps including different arrangement of promotors on expression plasmids or the two-phase fermentation system. In summary, the work provided is not only a proof-of-principle but also provides a promising approach for future large-scale biotechnological TA production.
The manuscript requires some editing, and especially the optimization approaches have been described in too much detail. I recommend that the authors should more focus on the successful approaches and reduce the less successfully designed experiments to a minimum.
They should rather summarize the important parameters for optimized TA production.
[Response] Thank you so very much for appreciating the importance of our work. Regarding your suggestion to reduce the optimization approaches, the other two reviewers are fine with the current level of details. Thus, we think that this level of detail is in fact needed for the other researchers to follow what we did for strain development and bioconversion process development. If editor thinks that we should reduce the details, we will do so. Thank you very much again.
Other points: 1. L.68: in Ref. 10 the heterologously expressed gene was of course known, so this comment is not really right.
[Response]: Thank you. We originally wanted to emphasize the limitation of using wild-type bacterial isolates from nature for TPA production from pX. To make this point clear, the sentence is now changed to "Also, in the case of wild-type bacterial isolates, the underlying catalytic mechanisms…".
2. L.70: detailed reaction mechanisms have not been studied in this work, and they are also not important for the purpose of this study.
[Response]: Thank you. As the reviewer suggested, we rephrased the sentence to "Thus, it is necessary to first elucidate and validate each reaction steps of the bioconversion process…".
[Response]: Thank you. We removed "activities" as suggested. [Response]: As we responded earlier, we think that the current level of details is needed for the readers to understand the strategies employed for strain development. If editor thinks that we should reduce the description, we will do so. 7. L.182-187: it is trivial that TA is present as dianion at the pH of culture conditions: reduce this statement to a minimum.
[Response]: Thank you for this comment. Yes we agree that it is obvious to the experts like you. However, to some readers who are not familiar, TPA and TPA disalt might be confusing.
The concentrations presented in grams per liter (in toxicity test) can be quite misleading if we do not distinguish TPA vs. TPA disalt clearly. 8. L.196-L.243: Again this part reporting on the optimization of the fermentation should be shortened to 50%, the authors should rather report on the best conditions and then summarize what parameters were essential.
[Response]: As we responded earlier, we think that the current level of details is needed for the readers to understand the strategies employed for bioconversion process. If editor thinks that we should reduce the description, we will do so.

Reviewer #2 (Remarks to the Author):
Our society is totally dependent in a multitude of ways upon chemicals provided by the chemical industry. However, efforts to achieve sustainable development necessitate departure from polluting and high level energy-consuming processes based on non-renewable feedstocks, through exploitation of environmentally friendly chemistry options. A main plank in sustainable chemistry procedures is biocatalysis. Exploitation of biocatalytic procedures requires that they be economically competitive. Significant hurdles for biocatalytic processes for bulk chemicals are modest yields and susceptibility of the biocatalyst to substrateinteremediate-product toxicity.
In this paper, the authors propose a (ultimately) biocatalytic route from the renewable substrate glucose to terephthalic acid (TPA), the monomer of the commodity polymer PET, though they describe here an intermediate process of conversion of p-xylene to TPA, as a proof-of-principle.
The results presented constitute an elegant biocatalysis process based on a rationally designed metabolic pathway constructed in E. coli from the upper xylene degradation pathway of Pseudomonas putida F1 and the lower p-toluene sulfonate pathway of Comamonas testosteroni T-2. The biocatalytic activity of the construct was assessed in a 2-phase bioconversion system in which the xylene was supplied in the apolar phase. Initial experiments revealed bottlenecks within the designed pathway and several inducible constructs were made to relieve them and optimise product yield. The potential toxicity of TPA caused by high substrate feeding rates was analysed and shown to be unproblematic. In one case, a yield of 97% was recorded, which is extraordinary and, if reproducible when upscaled, will represent a breakthrough in hydrocarbon bioconversions.
All in all, this is an excellent and important paper documenting the feasibility of bioconversions for bulk chemistry.
While not yet documenting a process of conversion of a renewable substrate to TPA, it shows proof-of-principle, since other publications show the potential feasibility of conversion of renewables to p-xylene. In any case, the exploitation of petroleum constituents as feedstocks for chemicals production, despite being unsustainable over long time spans, is, unlike the use of fossil fuels for energy generation, not a sustainability priority, so a scale-up of the process described in this submission, and its implementation as a viable process for large scale production of TPA, may be envisaged.
[Response]: Thank you so very much for appreciating the importance of our work. I congratulate the authors on this important advance and only have one, rather trivial question: they examined the toxicity of TPA applied externally but, in this process, TPA is produced internally. Perhaps they may wish to comment on this, and on TPA release after formation?
[Response]: Thank you for the great comment. Indeed, we examined the TPA toxicity to E. coli cells by applying TPA desalt externally. As the reviewer knows, it is rather difficult to measure in vivo chemical concentration that starts inhibiting cell growth (in vivo toxicity).
However, in vitro toxicity tests, as done for the toxicity measurement to many chemicals by our group and others, at least guide us the approximate tolerance level, which is what we wanted to determine before strain engineering.

Reviewer #3 (Remarks to the Author):
In this manuscript, the authors describe the development of an E. coli strain that can convert p-xylene (pX) to terephthalic acid (TPA). The engineered pathway relies on known degradative routes for p-xylene and p-toluene sulfonate. The authors first demonstrate independent validation of the up-and downstream pathways, then demonstrate a two-phase partitioning bioreactor to achieve conversion of pX to TPA. This is an interesting, largely well-written manuscript. The suitability for this journal arises from the nature of the product being targeted -a very high-volume chemical compound -and the potential for achieving completely bio-based production of TPA. The manuscript would be considerably strengthened by addressing the comments below to improve clarity: [Response] Thank you very much for your positive comments.
1. p. 4, lines 73 and 74 -I don't think it's quite correct to start that there were "detailed studies" on enzymes in this paper. The pathway relies on known pathways and largely confirms previous observations rather than providing any new biochemical insights. The authors are better served by limiting this statement to "pathway reactions" since that is consistent with both the context and the results.
[Response]: Thank you for the comment. We agree. Thus, the sentence is now changed to "Here we report development of a metabolically engineered E. coli strain capable of oxidizing pX into TPA, based on detailed studies on pathway reactions …".
2. p. 5, lines 102-104 -As written, this sentence suggests that BADH does not use NADH and thus would result in a cofactor imbalance. According to Metacyc, the cofactor requirements for BADH are the same as for the second oxidative step of XMO. Thus, I don't see what using XMO provides an advantage with respect to redox balance.
[Response]: As you can see from several databases including MetaCyc as well as in the literature, XMO converts NADH to NAD + as a cofactor whereas BADH converts NAD + to NADH during the corresponding reactions they catalyze. Thus there is a difference in the cofactor usage between these two enzymes. Thus, the use of XMO in our study is advantageous with respect to redox balance as described in the manuscript.
3. p. 6, lines 121-122 and Supp Fig 3a -Is the production of pTA in this figure actually significant? It seems that a small amount may have been detected at 90 min, but without error bars, it's impossible to know if this is "real"? Were there biological replicates that consistently showed pTA production? Any showing accumulation, i.e., detection at more than one time point?
[Response]: Thank you for the great comment. The purpose of the experiments in Supplementary Fig. 3 was to characterize the successful heterologous expression of enzymes of interest and validation of the expected reactions through detecting the products. We thus focused mainly on detecting the formation of target products during the whole-cell activity assay. In the case of Supplementary Fig. 3a, the products of interest are pTALC and pTALD as can be seen in our pathway design. It was found that these two metabolites were successfully produced at relatively high yields. So, our purpose was achieved. It was also interesting to find that pTA seemed to be accumulated in tiny amount. Thus, we revised our manuscript accordingly. Thank you again. 4. p. 9, line 183 -suggest changing "has not been known" to "has not been evaluated" or something similar.
[Response]: Thank you. Corrected as suggested. 5. p. 9, lines 187-194 and Supp Fig 6 -The data do not support the conclusion that there is no effect on growth at TPA-salt concentrations up to 10 g/L. There is a clear difference in max OD between the 0 g/L control and the 5 and 10 g/L samples. It may be true that the growth rates during exponential phase are not different, but there are not enough data points to accurately calculate a growth rate given that there is only one data point between inoculation at time t=0 and stationary phase. If the authors want to make claims about growth rate, more data points are needed to ensure that measurements are being made during the exponential phase.
[Response]: Thank you for the great comment. Yes, indeed we agree that we should not use the term "growth rate" with a few sampling points. Thus, as you thankfully pointed out, we changed the description as follows: "Exposing cells to various concentrations of TPA disodium salts up to 10 g/L (equivalent to TPA concentration of 7.9 g/L) showed only slight reduction in the final optical density. E.
coli cells could tolerate up to 40 g/L of TPA disodium salts (equivalent to TPA concentration of 31.6 g/L). At this TPA disodium salts concentration, the final optical density was a half that of the control without TPA disodium salts ( Supplementary Fig. 6)." 6. p. 15, line 324 -The amount of substrates added need to be provided with more detail than "an appropriate amount." For each experiment, the specific substrate and amount added should be provide. A similar comment is provided below in reference to Supp Fig 3. [Response]: Thank you for the comment. As you suggested, we provided the concentrations of substrates in the revised manuscript. 7. p. 17, lines 371-373 -For these latter runs, was there also just a single bolus addition of 50-ml nutrient solution? It is noteworthy that Fig 3 shows no accumulation of glucose after the onset of feeding while Supp Fig 7 clears shows a spike. This suggests a different in either feeding strategy or physiology that should be explained.
[Response]: The latter runs were based on the DO-stat feeding strategy, and thus there was no single bolus addition of nutrient solution. In the case of Fig. 3, the DO-stat nutrient feeding operated perfectly so that the residual glucose was maintained at relatively low concentration.
In the case of Supplementary Fig. 7, it belongs to the preliminary runs where the strategy of manual feeding by single bolus addition was applied for nutrient supplementation. Therefore, they are two different modes of nutrient feeding. These were clearly described in our original manuscript.
8. Supp Fig 2 -Additional descriptions are required for the arrows, i.e., specify the proteins that are being pointed to, along with the expected molecular weights.
[Response]: Thank you. As the reviewer suggested, we added the information in the revised manuscript. show productivity originating from pX, but pX titer is not shown anywhere. I understand that the solubility is quite low and thus it may have been unmeasurable, but if this is the case, it should be noted in the figure legend and/or the methods. Panel 3a also shows spikes in metabolites, which suggests some sort of feeding regimen. It would also be very helpful to use the same abbreviations (in parentheses) in the figure legend as are used in the main text.
[Response]: Thank you for the comment. Indeed, the reason why pX consumption profile was not shown was that the whole-cell activity assay was conducted in aqueous solution in which pX could not be dissolved due to its insolubility in water. Thus, it is rather difficult to accurately monitor pX consumption as the reviewer noted already. We thus described this in the revised manuscript as the reviewer suggested.
Regarding the spikes of metabolites in Supplementary Fig. 3a, we already responded to the Reviewer 1's comment.
Regarding abbreviations, we corrected our mistakes. Thank you. Table 5 is incorrectly listed as Supp Figure 5. Also, the units for pX added (superscript note 'b') need to be provided.

Supp
[Response]: Thank you for pointing out the mistake. We corrected it. Regarding the units for pX, we already provided it ("g" for gram) in our original manuscript.