Thermal annihilation of photo-induced radicals following dynamic nuclear polarization to produce transportable frozen hyperpolarized 13C-substrates

Hyperpolarization via dynamic nuclear polarization (DNP) is pivotal for boosting magnetic resonance imaging (MRI) sensitivity and dissolution DNP can be used to perform in vivo real-time 13C MRI. The type of applications is however limited by the relatively fast decay time of the hyperpolarized spin state together with the constraint of having to polarize the 13C spins in a dedicated apparatus nearby but separated from the MRI magnet. We herein demonstrate that by polarizing 13C with photo-induced radicals, which can be subsequently annihilated using a thermalization process that maintains the sample temperature below its melting point, hyperpolarized 13C-substrates can be extracted from the DNP apparatus in the solid form, while maintaining the enhanced 13C polarization. The melting procedure necessary to transform the frozen solid into an injectable solution containing the hyperpolarized 13C-substrates can therefore be performed ex situ, up to several hours after extraction and storage of the polarized solid.

1. I strongly recommend to use level of polarization (%) instead of enhancement factor because it is absolute unity. It should be done in line 67 and in figureS3. You may speci fy enhancement in addition if you like. 2. . It is unclear from the very beginning that you are describing your own observations but not continuing the introduction. Also write this sentence in clearer way. 3. (Lines 129-132) The purpose is unclear at the very beginning. It is better to simplify the sentence and add one more sentence to clarify why you had to do this experiment and how it will be used in future. 4. Why you obtain polarization level less than unity? At the intro it was pointed out that in 70th polarization level around unity was reached! C ould you comment it or suggest an improvement? 5. You made your beads from H2O. But finally you dissolve the beads in hot D2O. I think it will be very valuable for the contribution to compare results for beads made from H2O and D2O. You have excellent methodology and it should be easy to implement. 6. (line 200) Power of MW-source is 55 mW. Please specify the amplitude in "Hz" that is very important to see which part of spectrum could be covered. You should add ESR spectrum of radical (maybe even calculated one) on the top of the figure S3, when it will be easy to see the part of spectrum that you hit during preparation of DNP. 7. Figure S3. You should add specifications of MW-irradiation: time, power, amplitude (in GHz) and carrier frequency. 8. (line 204 and 206) It is unclear why do you use 30-degree RF-pulse. 9. Figure S4. you did not fully specified MW-irradiation, and there is no information about angle of RF-pulse that you determined here. Section "Solid-state 13C NMR coil calibration and rf correction" should be checked: (1) "(see Fig. S3)" should be replaced by "(see Fig. S4)".
(2). Text should be before the figure. 10. In the SM H2O in several places is H2"Zero". In addition to that Figures should be revised with care: 1. Font size in figure 2 and 4 seems to be tiny. 2. In all figures with spectra or dependences the temperature should be specified on the figure explicitly. 3. The color of you beads should be the same as the color of corresponding dependences. i.e. 3.1. Figure 2 (d) cyan as you called it "blueish" color should be replaced by the same blue color like in part (e) and (f), like you do it for "yellowish" beads and corresponding plots. However changes should be consistent with Figure 2-4 which should be modified respectfully. 3.2. Figure 4 (b) Probably here color should be yellow? Because it was measured before thermalization. 3.3. Figure 4 (f) color should be blue because it is after thermalization. etc for all Figures. Now in the text you mix up "cold" and "worm" colors and "before" and "after" thermalization. It is better to use one color coding. Despite the above mentioned criticism it was pleasure to review the manuscript.

Reviewer #4 (Remarks to the Author):
The authors have cleverly and decisively demonstrated for the first time that it is possible to extract a frozen sample of pyruvic acid (PA) from the polarizer before performing a solid -to-liquid phase transition. The authors have produced transient photo-induced radicals for the polarization of PA which can be annihilated by raising the sample temperature while it is in a solid state. This process-hyperpolarizing a PA sample at 1K with transient radicals produced through UVirradiation-is a dramatic advancement in the dynamic nuclear polarization (DNP) of PA for the real-time study of metabolic pathways in vivo, and has many obvious advantages. A method for hyperpolarizing carbon-13 compounds through DNP based on the use of nonpersistent radicals is in many ways an ideal one. The current method for polarizing carbon -13 compounds using DNP is to mix the target molecule with a radical and then radiate the frozen sample with microwave frequency at a temperature of ~1K. After achieving the desired level of polarization, the sample is then extracted for in vitro or in vivo studies after in situ dissolution. While this method has been used effectively for studying metabolic pathways in vivo in real -time, it has several disadvantages. For one, in situ dissolution is a very complex process that complicates the in-vivo applications of DNP studies. For example, a minor mistake in following the complex dissolution protocol leads to frozen lines that cause system downtime, failed dissolution and unsuccessful studies. This is an extremely important factor for human studies, where subject preparation is highly complex and the availability of a robust dissolution method is critical. Second, in situ dissolution leads to the extraction of the polarized sam ple in liquid form, limiting its usage due to the very short T1 relaxation time of hyperpolarized PA in this state. Finally, any clinical applications require that the radicals be completely separated from the sample prior its administration, a process which could lead to a loss of polarization. This paper effectively addresses all of these limitations. The paper displays a sound research design, from which the authors have derived very exciting results. Using their established method of polarizing PA with non-persistent radicals produced by UV radiation of the sample, they have shown that this method can be modified to annihilate the radicals by changing the sample temperature. Figure 1 powerfully demonstrates the presence and absence of radicals at temperatures below and above 192K, respectively. This is an ideal method for polarizing PA for in-vivo studies. Due to the significant difference in T1 relaxation time between the liquid and solid phases of hyperpolarized PA (tens of seconds in the liquid pha se, as opposed to many tens of minutes in the solid state), this method could dramatically expand the use of this technology for medical and nonmedical applications; taking advantage of the much longer T1 relaxation time of solid hyperpolarized PA, it allows a frozen sample to be extracted from the polarizer and transferred to another location without a significant loss of polarization. This is a major advantage, since the polarizer and the MRI scanner no longer need to be in close physical proximity to on e another. Indeed, this method means that it is now possible to use one polarizer to produce polarized samples for use in multiple scanners in multiple locations. While the paper effectively demonstrates the capabilities of the described method, it does not do as effective a job of discussing its limitations, which should be briefly discussed in the manuscript. These limitations include: 1. The low polarization percentage currently achievable with this method. A look at the X -band ESR spectrum of the PA sample collected at 77 K indicates that this method will never be able to produce the same polarization level that can be achieved when using a trityl radical. This represents a fundamental limitation of the described method; despite the fact that they have used a 7 T DNP system, the somewhat unfavorable level of polarization they report (around 12%) could dramatically limits this method's clinical applications. It might be useful to discuss the attributes of the ESR spectra provided here compared to the ESR spectra of a trityl radical. Specifically, how would these changes in ESR spectra affect the polarization? 2. The method demonstrated here is currently only applicable to a finite number of molecules, so the authors should discuss why it cannot be used for other interesting biological molecules. In fact, one could argue that molecules with low solubility may benefit from this method more significantly, as it does not require a uniform mixture of radical and carbon-13 compounds.
In summary, this is an outstanding piece of work. Though the above suggestions would make it more accessible to general readers, this is an excellent manuscript even in its present state. We would first like to thank the reviewer for his valuable comments. We agree that the lower polarization level and longer polarization time are currently two important limitations of the proposed method. However, because the samples can be irradiated days in advance and stored in liquid nitrogen, we do not think that the time required to perform the UV irradiation should be included in the calculation of the effective time for the DNP process. We would also like to emphasize that the irradiation time is directly linked to the radical quantum yield, which will depend on the light source used for the irradiation. It is therefore not yet clear what the minimum irradiation time would be for an optimal UV source, but a complete optical study is beyond the scope of this manuscript.

Reviewers' comments
We have added the following text before the concluding paragraph to address these limitations: "The liquid-state 13 C polarization obtained in the present work is severalfold lower than the maximum values obtained by dissolution DNP using the same polarizer and trityl radicals (Yoshihara, Can et al. 2016). The narrower ESR line width of trityl radicals as compared to UV-induced radicals is expected to lead to higher 13 C polarization (Eichhorn, Takado et al. 2013), but the main reason for the lower polarization is the sub-optimal radical concentration, which also explains that the build-up time constant is larger than in an optimized PA sample prepared with trityl radical. Preliminary results show that this concentration could be improved by using a broadband UV source, but a complete optical study is beyond the scope of this manuscript.
Modulating the microwave frequency or performing 1 H-13 C cross-polarization could also improve the polarization level and reduce the build-up time (Bornet, Melzi et al. 2013, Bornet, Milani et al. 2014), but these methods require additional hardware."

This is a very good contribution in the field of Dynamic Nuclear Polarization (DNP)
demonstrating that the lifetime of 13C hyperpolarized agent can be significantly extended to over many hours, and can exceed more than 10 hours. This would be interesting to many working in this field. I am very much in favor of publishing this contribution. If this manuscript had submitted to JPC, I would recommend its acceptance minus the "insults" to the field of molecular imaging dominated by the PET community rather than 13C DNP and expanding the references section. Given that this is Nature Communications, I am recommending a minor revision to additionally voice my criticism that the manuscript's introduction should be re-written in the broader context of hyperpolarization rather than dissolution DNP alone. The findings and the work conducted though is of the highest caliber and the manuscript / this contribution will be of broad interest to scientists around the world. I note that more than 100 DNP systems have been sold around the globe (and many home built setups were developed), so this work would be of immediate relevant to a wide range of scientists: from physicists to medical doctors.

Key Points:
1) First sentence of the Main text "Because hyperpolarized 13C MRI is currently the only modality allowing real-time metabolic imaging in vivo": has no merit. Hyperpolarized 13C MRI is the recent metabolic imaging modality, but certainly not "the only". This is a huge insult to the entire field of molecular imaging, and no credible scientists would be able to accept this sentence. I can name a few techniques: detection of redox and other reactions with optical imaging, EPR imaging of oxygenation states, FMISO PET, hyperpolarized 129Xe sensing of receptor imaging… There is a long list.
We agree with the reviewer that the introduction should be expended to cover all molecular imaging modalities, even those that have not yet been used clinically. We have revised the manuscript accordingly. We have also incorporated the references mentioned in the reviewer's point #4, included alternative hyperpolarization techniques (to take into account reviewer's point #5) and emphasized the reasons why 13 C is the most versatile nucleus for biomedical applications (reviewer's point #2). The first part of the introduction has therefore been replaced by the following: "Hyperpolarized 13 C MRI is one among several molecular imaging techniques proposed in the recent years to detect biochemical changes in vivo (Golman, in't Zandt et al. 2006, Nelson, Kurhanewicz et al. 2013. Most imaging modalities, including MRI, computed tomography (CT), positron emission tomography (PET), single photon emission computed tomography (SPECT), ultrasound, and a variety of optical imaging methods can be adapted to reveal insights into cellular function (James and Gambhir 2012). MRI can provide direct biochemical information via the spectroscopic dimension of nuclear magnetic resonance (NMR), allowing simultaneous acquisition of signals from a substrate and its metabolic products, hence yielding true metabolic imaging. The relatively low sensitivity of NMR can be circumvented using hyperpolarization techniques such as spin-exchange optical pumping (SEOP) for gases, and dissolution DNP, parahydrogen-induced polarization (PHIP) as well as the so-called "brute force" method for liquids (Comment 2013, Witte and Schroder 2013, Nikolaou, Goodson et al. 2015. In the context of metabolic imaging, 13 C is the most adapted nucleus because of its ubiquitous presence in the vast majority of biomolecules, its large chemical shift dispersion allowing to easily differentiate the various species, its low natural abundance, and its relatively long longitudinal relaxation time at specific molecular positions such as in a carboxyl group (Golman, Olsson et al. 2003, Golman, in't Zandt et al. 2006, Mansson, Johansson et al. 2006, Kurhanewicz, Vigneron et al. 2011, Comment and Merritt 2014, Brindle 2015. PHIP was the first hyperpolarization technique proposed for in vivo 13 C MRI (Golman, Axelsson et al. 2001), but dissolution DNP became more popular because of its versatility for applications in biomedical imaging (Ardenkjaer-Larsen, Fridlund et al. 2003, Ardenkjaer-Larsen 2016." 2) There is a second issue with this first sentence. 13C is not the only hyperpolarized nucleus that can be used to track metabolism in vivo. I would rather see the authors to articulate, why 13C is so ubiquitous for biomedical uses: i.e. huge chemical shift dispersion, ubiquitous presence in the vast majority of biomolecules and metabolites, and low natural abundance background. This is Nature Communications, and the authors owe it to the reader to guide it to the importance why this work is a big deal (and it is).
See answer to point #1. We added the following note after the sentence mentioned by the reviewer:

3) The literature and the
"…(this time can be extended in some specific molecules through the formation of a so-called longlived state (Vasos, Comment et al. 2009, Warren, Jenista et al. 2009 We agree with the reviewer and have included these references (see our answer to point #1). 169-174. One should also remember that SNR is a true value in NMR/MRI rather than EMF itself. Since this paper is focused on DNP rather than sensitivity of detection I would suggest re-wording the sentence to: "The strength of the emf can be increased through the increase of the amplitude of the nuclear spin magnetization." or something similar. Reword the following sentence accordingly; Note that polarization also increases with the field strength.
We replaced the sentence by one similar to the sentence suggested by the reviewer, i.e.: "The strength of the emf can be enhanced by increasing the amplitude of the nuclear spin polarization,…"

7) There are certainly other biomolecules that can be hyperpolarized with DNP. This study is focused on pyruvate. The authors need to comment on the applicability of the demonstrated technique to other biomolecules. This is not done.
We are currently working on extending this method to other compounds beyond pyruvic acid and we already know that most alpha-keto acids should have similar properties after UV irradiation, f.i. with pyruvic acid. These studies are still ongoing and will be presented in future publications. We have however added the following sentence in the conclusion: "The herein presented technique can be used to polarize other 13 C-compounds as well as other nuclei following the method proposed by Capozzi et al. (Capozzi, Hyacinthe et al. 2015). Since other alpha-keto acids exhibit similar photochemical behavior as PA (Leermakers and Vesley 1963), we expect to be able to extend this technique to other biomolecules in the near future." 8) In part, the reason why authors demonstrate such long T1 in the solid is due to the use of low temperature. This has been studied extensively in the context of 129Xe, and this needs to be discussed: e.g. more recently by Saam and co-workers https://arxiv.org/pdf/1607.01072.pdf We have added a sentence to highlight the similarity with hyperpolarized 129Xe: "Note that it has also been demonstrated that hyperpolarized 129 Xe can be stored for an extended period of time and transported in the solid form at cryogenic temperature (Hersman, Ruset et al. 2008, Limes, Ma et al. 2016."

pravdivtsev.a.n@gmail.com
Review It is well written paper and a great contribution to the DNP field. It will be of interest to wider audience because it has interesting physical and chemical background and has direct medical application. The paper has well written separate experimental section that will able one to reproduce the experiments. I strongly recommend it for publication in Nature Communication after minor revision.

I strongly recommend to use level of polarization (%) instead of enhancement factor because it is absolute unity. It should be done in line 67 and in figureS3. You may specify enhancement in addition if you like.
We fully agree with the reviewer that polarization level should be used because it is an absolute value. Supplementary Figure 3 has been modified accordingly. All other results presented in our manuscript are expressed in % polarization. In line 67, where we refer to work published by Ardenkjaer-Larsen et al. (PNAS 2003), we think that it would be incorrect to specify a polarization level instead of the enhancement factor the authors chose to report in their original manuscript. Lines 75-77). It is unclear from the very beginning that you are describing your own observations but not continuing the introduction. Also write this sentence in clearer way.

(
We have replaced the sentence by the following two sentences: "These radicals, formed at 77 K by UV-irradiation of frozen aliquots containing PA, do not persist when brought to room temperature. Using electron spin resonance (ESR), we determined that the radicals are annihilated if the sample temperature increases above 190±2 K (Fig. 1)." 3. (Lines 129-132) The purpose is unclear at the very beginning. It is better to simplify the sentence and add one more sentence to clarify why you had to do this experiment and how it will be used in future.
We have replaced the ambiguous sentence by the following two sentences: "The elimination of the radicals in the solid state alleviates the crucial requirement for dissolution to take place inside the high-field and low-temperature environment of the polarizer. To demonstrate that it is indeed possible to perform the solid-to-liquid transformation outside of the polarizer after extraction of the frozen beads without considerable reduction of the 13C polarization, an additional set of experiments was performed."

Why you obtain polarization level less than unity? At the intro it was pointed out that in 70th polarization level around unity was reached! Could you comment it or suggest an improvement?
We have added the following text before the concluding paragraph to address these limitations and discuss potential improvements: "The liquid-state 13 C polarization obtained in the present work is severalfold lower than the maximum values obtained by dissolution DNP using the same polarizer and trityl radicals (Yoshihara, Can et al. 2016). The narrower ESR line width of trityl radicals as compared to UV-induced radicals is expected to lead to higher 13 C polarization (Eichhorn, Takado et al. 2013), but the main reason for the lower polarization is the sub-optimal radical concentration, which also explains that the build-up time constant is larger than in an optimized PA sample prepared with trityl radical. Preliminary results show that this concentration could be improved by using a broadband UV source, but a complete optical study is beyond the scope of this manuscript.
Modulating the microwave frequency or performing 1 H-13 C cross-polarization could also improve the polarization level and reduce the build-up time (Bornet, Melzi et al. 2013, Bornet, Milani et al. 2014), but these methods require additional hardware."

You made your beads from H2O. But finally you dissolve the beads in hot D2O. I think
it will be very valuable for the contribution to compare results for beads made from H2O and D2O. You have excellent methodology and it should be easy to implement.
We agree with the reviewer that replacing H2O by D2O is indeed an interesting experiment. In fact, we had already performed such experiment but did not include the results in the manuscript.
The reason is that there was no statistically significant difference between the two types of samples. Perhaps the results would be more conclusive if PA was also replaced by deuterated PA, but we think that this would not be of great interest for future in vivo studies and it would be beyond the scope of this manuscript. We nevertheless added the following sentence in the Methods section: "Note that the observed enhancement and time evolution were similar for samples composed of The microwave B1 field is ill-defined in this type of DNP experiments since the sample is larger than the wavelength and there is no resonant cavity to perform ESR experiments. We however agree with the reviewer that adding the calculated ESR spectrum in Supplementary Figure 3 makes it easy to compare the microwave spectrum with the width of the ESR line width. The figure has been updated accordingly.

(line 204 and 206) It is unclear why do you use 30-degree RF-pulse.
A larger flip angle is necessary to increase the SNR of the thermal equilibrium signal in order to accurately determine the solid-state polarization. We added the following sentence to clarify this point: "The DNP enhancement was obtained by computing the ratio between the two NMR signal integrals. A larger flip angle (30°) was necessary to increase the signal-to-noise ratio of the thermal equilibrium signal in order to accurately determine the solid-state polarization." 9. Figure S4. you did not fully specified MW-irradiation, and there is no information about angle of RF-pulse that you determined here. Section "Solid-state 13C NMR coil calibration and rf correction" should be checked: (1) "(see Fig. S3)" should be replaced by "(see Fig. S4)". (2). Text should be before the figure.
We thank the reviewer for noticing these errors and missing information. "see Fig. S3" was replaced by "see Fig. S4" and the text moved before Fig. S4. The legend was replaced by the following: "13C NMR signal integral as a function of the acquisition number. The consecutive single-scan acquisitions spaced by 1 s. rf excitation was done with a single 5 us and 6.3 W square pulse, corresponding to a flip angle of 6.0±0.5°). The measurements were performed at 4.2 K and 7 T after having partially polarized the sample for 10 min with 55 mW microwave power (as measured at the source output) at 196.633 GHz (optimal frequency according to Supplementary Fig. 3)."

In the SM H2O in several places is H2"Zero".
In addition to that Figures should be revised with care: 1. Font size in figure 2 and 4 seems to be tiny.

Figure 4 (b)
Probably here color should be yellow? Because it was measured before thermalization.

Figure 4 (f) color should be blue because it is after thermalization. etc for all Figures.
Now in the text you mix up "cold" and "worm" colors and "before" and "after" thermalization. It is better to use one color coding.
We updated the figures and manuscript according to the reviewer's suggestions.

Despite the above mentioned criticism
it was pleasure to review the manuscript.
We thank the reviewer for his positive comment.

Reviewer #4 (Remarks to the Author):
The authors have cleverly and decisively demonstrated for the first time that it is possible to extract a frozen sample of pyruvic acid (PA) from the polarizer before performing a solid-to-liquid phase transition. The authors have produced transient photo-induced radicals for the polarization of PA which can be annihilated by raising the sample temperature while it is in a solid state. This process-hyperpolarizing a PA sample at 1K with transient radicals produced through UV-irradiation-is a dramatic advancement in the dynamic nuclear polarization (DNP) of PA for the real-time study of metabolic pathways in vivo, and has many obvious advantages.
A method for hyperpolarizing carbon-13 compounds through DNP based on the use of non-persistent radicals is in many ways an ideal one. The current method for polarizing carbon-13 compounds using DNP is to mix the target molecule with a radical and then radiate the frozen sample with microwave frequency at a temperature of ~1K. After achieving the desired level of polarization, the sample is then extracted for in vitro or in vivo studies after in situ dissolution. While this method has been used effectively for studying metabolic pathways in vivo in real-time, it has several disadvantages. For one, in situ dissolution is a very complex process that complicates the in-vivo applications of DNP studies. For example, a minor mistake in following the complex dissolution protocol leads to frozen lines that cause system downtime, failed dissolution and unsuccessful studies. This is an extremely important factor for human studies, where subject preparation is highly complex and the availability of a robust dissolution method is critical. Second, in situ dissolution leads to the extraction of the polarized sample in liquid form, limiting its usage due to the very short T1 relaxation time of hyperpolarized PA in this state. Finally, any clinical applications require that the radicals be completely separated from the sample prior its administration, a process which could lead to a loss of polarization. This paper effectively addresses all of these limitations.
The paper displays a sound research design, from which the authors have derived very exciting results. Using their established method of polarizing PA with non-persistent radicals produced by UV radiation of the sample, they have shown that this method can be modified to annihilate the radicals by changing the sample temperature. Figure 1 powerfully demonstrates the presence and absence of radicals at temperatures below and above 192K, respectively. This is an ideal method for polarizing PA for in-vivo studies.
Due to the significant difference in T1 relaxation time between the liquid and solid phases of hyperpolarized PA (tens of seconds in the liquid phase, as opposed to many tens of minutes in the solid state), this method could dramatically expand the use of this technology for medical and non-medical applications; taking advantage of the much longer T1 relaxation time of solid hyperpolarized PA, it allows a frozen sample to be extracted from the polarizer and transferred to another location without a significant loss of polarization. This is a major advantage, since the polarizer and the MRI scanner no longer need to be in close physical proximity to one another. Indeed, this method means that it is now possible to use one polarizer to produce polarized samples for use in multiple scanners in multiple locations.
While the paper effectively demonstrates the capabilities of the described method, it does not do as effective a job of discussing its limitations, which should be briefly discussed in the manuscript. These limitations include: 1.The low polarization percentage currently achievable with this method. A look at the X-band ESR spectrum of the PA sample collected at 77 K indicates that this method will never be able to produce the same polarization level that can be achieved when using a trityl radical. This represents a fundamental limitation of the described method; despite the fact that they have used a 7 T DNP system, the somewhat unfavorable level of polarization they report (around 12%) could dramatically limits this method's clinical applications. It might be useful to discuss the attributes of the ESR spectra provided here compared to the ESR spectra of a trityl radical. Specifically, how would these changes in ESR spectra affect the polarization?
We first would like to thank the reviewer for his positive feedback.
Regarding his first point, we agree that the discussion should be extended. We have added the following text before the concluding paragraph to discuss the ESR spectra and how the current method could be possibly improved: "The liquid-state 13 C polarization obtained in the present work is severalfold lower than the maximum values obtained by dissolution DNP using the same polarizer and trityl radicals (Yoshihara, Can et al. 2016). The narrower ESR line width of trityl radicals as compared to UV-induced radicals is expected to lead to higher 13 C polarization (Eichhorn, Takado et al. 2013), but the main reason for the lower polarization is the sub-optimal radical concentration, which also explains that the build-up time constant is larger than in an optimized PA sample prepared with trityl radical. Preliminary results show that this concentration could be improved by using a broadband UV source, but a complete optical study is beyond the scope of this manuscript.
Modulating the microwave frequency or performing 1 H-13 C cross-polarization could also improve the polarization level and reduce the build-up time (Bornet, Melzi et al. 2013, Bornet, Milani et al. 2014), but these methods require additional hardware." 2. The method demonstrated here is currently only applicable to a finite number of molecules, so the authors should discuss why it cannot be used for other interesting biological molecules. In fact, one could argue that molecules with low solubility may benefit from this method more significantly, as it does not require a uniform mixture of radical and carbon-13 compounds.
We are currently working on extending this method to other compounds beyond pyruvic acid and we already know that most alpha-keto acids should have similar properties after UV irradiation, f.i. with pyruvic acid. These studies are still ongoing and will be presented in future publications. However, as suggested by the reviewer, we have added the following sentence in the conclusion to discuss the possible extension of the method: "The herein presented technique can be used to polarize other 13 C-compounds as well as other nuclei following the method proposed by Capozzi et al. (Capozzi, Hyacinthe et al. 2015). Since other alpha-keto acids exhibit similar photochemical behavior as PA (Leermakers and Vesley 1963), we expect to be able to extend this technique to other biomolecules in the near future." We also added a sentence to highlight the reviewer's point, which we fully agree with, concerning the potential of this method for molecules with low solubility: "In this study, we take advantage of the non-persistent nature of specific photo-induced radicals to produce hyperpolarized 13C-substrates that can be extracted from the DNP apparatus without the need for a dissolution process. The frozen solid containing hyperpolarized 13C-substrates can consequently be melted at a later time, in a remote location, and this method also extends the potential applications of hyperpolarized 13C MRI to biomolecules with low solubility." In summary, this is an outstanding piece of work. Though the above suggestions would make it more accessible to general readers, this is an excellent manuscript even in its present state.
We thank the reviewer for his comment.
List of added references (alphabetical order)