A ribose-functionalized NAD+ with unexpected high activity and selectivity for protein poly-ADP-ribosylation

Nicotinamide adenine dinucleotide (NAD+)-dependent ADP-ribosylation plays important roles in physiology and pathophysiology. It has been challenging to study this key type of enzymatic post-translational modification in particular for protein poly-ADP-ribosylation (PARylation). Here we explore chemical and chemoenzymatic synthesis of NAD+ analogues with ribose functionalized by terminal alkyne and azido groups. Our results demonstrate that azido substitution at 3′-OH of nicotinamide riboside enables enzymatic synthesis of an NAD+ analogue with high efficiency and yields. Notably, the generated 3′-azido NAD+ exhibits unexpected high activity and specificity for protein PARylation catalyzed by human poly-ADP-ribose polymerase 1 (PARP1) and PARP2. And its derived poly-ADP-ribose polymers show increased resistance to human poly(ADP-ribose) glycohydrolase-mediated degradation. These unique properties lead to enhanced labeling of protein PARylation by 3′-azido NAD+ in the cellular contexts and facilitate direct visualization and labeling of mitochondrial protein PARylation. The 3′-azido NAD+ provides an important tool for studying cellular PARylation.

I previously reviewed the manuscript for another journal and I recommended for publication after one round of revision that addressed my initial comments. So my comments will be brief here and I again support the publication at Nature Communication.
The manuscript describes the development of new NAD+ analogs that can be used to label the substrate proteins of PARPs. The authors have demonstrated that the new probes have certain improved properties compared to existing ones. Given the important functions of PARPs in various biological pathways, I believe such probes will be useful for the elucidating the functions of PARPs. Furthermore, the chemo-enzymatic methods developed here will also be useful for the synthesis of other NAD+ analogs. For these two reasons, I support the publication of the manuscript.

NCOMMS-19-06435
A Ribose-Functionalized NAD+ with Unexpected High Activity and Selectivity for Protein Poly-ADP-Ribosylation Zhang et al generate NAD+ derivatives with functional groups added to the nicotinamide ribose. They accomplish this by using the enzyme NRK1 to phosphorylate modified variants of nicotinamide ribose as starting materials, and then the enzyme NMNAT to join the modified nicotinamide mononucleotide (NMN) variants with ATP to produce the NAD+ derivatives. They develop an efficient method for producing and purifying these NAD+ derivatives, and then test whether and how efficiently they can serve as substrates for the enzyme PARP-1, relative to the native substrate NAD+. One of the derivatives, compound 6, is utilized by the enzyme PARP1 with an efficiency similar to that of NAD+. An interesting feature of the poly(ADP-ribose) made from this compound is that it appears to be resistant to degradation by PARG, poly(ADP-ribose) glycohydrolase, and could thus potentially serve as an interesting tool for investigating the cellular consequences persistent poly(ADP-ribose) levels. Another interesting feature is that compound 6 appears to be poorly utilized by PARP5a and PARP10. There are limitations to the utility of compound 6 in that it appears to require permeabilization of cells in order to be delivered. The approach is interesting and thoroughly done, and the derivatives could potentially answer some interesting biological questions. In this regard, the study is mostly tool development and does not really push the boundaries of our knowledge of PARP enzymology or cell biology. An example of how this new technology can be used to improve our knowledge would add an important element to the study and broaden the importance of the work.
conclude that high molecular branched polymers of similar structure to that formed using NAD as a substrate are the result when using the 3-N3-NAD. Clarification will require additional study, and in this communication the authors should simply and clearly point out that more work will be required to characterize the products of the reaction with the new substrate 3-N3-NAD.
There are several minor points that I believe that authors should be asked to address prior to publication: 1. On page 5-6, yields of 68% are reported for the enzymatic conversion of NR6 to the NAD analog 6. It will be of interest to a reader to know if this refers to an isolated yield of 6 or to a yield calculated on the basis of HPLC analysis at the conclusion of the reaction. The authors should clarify the reporting of this result. It will also be of interest to know the maximum quantities of 6 that have been produced using this enzymatic synthesis.
2. Page 7 line 162-163. Lack of substrate activity for compound 5 could alternately be due to electronic properties of 5. The electron withdrawing 2-azide adjacent to the pyridinium-leaving group is predicted to decrease the rate of pyridine-ribose bond cleavage significantly. In a previous study [A.L.Hanlon et. al., JACS (1994) 116, 12087-12088}, 2'-azido-NAD was shown to be resistant to both NAD glycohydrolase catalyzed and to chemical hydrolysis. The 2'-azide was hydrolyzed by the NAD glycohydrolase approximately 10,000-fold more slowly than was NAD. Compound 5 could therefore be a slowly cleaved or a non-cleavable substrate analog. Incidentally, 2-N3-NAD was first reported in this study.
3. Page 14, lines 297-298. The experiment that the authors report demonstrates that 6 is not a substrate for SIRT2, not that there is an essential role for the nicotinamide ribotide 3'-OH in NAD analog 6. An azide group is bigger and electronically different from an -OH, and the azide simply might not fit or might constrain that analog into a different confirmation. We can only conclude that compound 6 is not a substrate and that substitution of an -OH for an -N3 is not tolerated by this enzyme. 4. It is important that the authors define the purity of the compounds that are tested in this study. It is expected that tested compounds will be highly pure. Therefore I would recommend the addition of a table or a paragraph to Methods that describes the purity of NAD analogs 1-6 and the method or methods that were used to assess the tested compounds purity. I expect (from the Supplementary Information)that the method will be analytical HPLC, but the purity of the compounds tested is never explicitly defined in the description of their synthesis. 5. The synthesis of the six NAD analogs described in this study begins with the construction of substituted ribose derivatives, the formation of a 1-bromo-ribose derivative, and the displacement of the 1-bromide with nicotinamide to form the nicotinamide-ribose derivative. The stereochemistry of the reaction product must be defined, and it's additionally important to insure that the NAD derivatives that are eventually synthesized contain the expected nicotinamide betariboside. Normally the beta-linkage is formed when the bromo-ribose has a 2-benzoyl substituent, but this is not the case for compound 5-7 and NR-5 where a 2-azide is present. In this case and for 1-7 and 3-7 (containing 2-ether substituents) how was the stereochemistry of the products assigned, and how was it proved that the products are the beta-anomers as drawn?

Reviewer 1
The manuscript describes the development of new NAD+ analogs that can be used to label the substrate proteins of PARPs. The authors have demonstrated that the new probes have certain improved properties compared to existing ones. Given the important functions of PARPs in various biological pathways, I believe such probes will be useful for the elucidating the functions of PARPs. Furthermore, the chemo-enzymatic methods developed here will also be useful for the synthesis of other NAD+ analogs. For these two reasons, I support the publication of the manuscript.
We do appreciate the reviewer's comments.

Reviewer 2
…. There are limitations to the utility of compound 6 in that it appears to require permeabilization of cells in order to be delivered. The approach is interesting and thoroughly done, and the derivatives could potentially answer some interesting biological questions. In this regard, the study is mostly tool development and does not really push the boundaries of our knowledge of PARP enzymology or cell biology. An example of how this new technology can be used to improve our knowledge would add an important element to the study and broaden the importance of the work.
We do appreciate the reviewer's comment. To demonstrate the utility of compound 6, we applied it to study mitochondrial poly-ADP-ribosylation (PARylation). While the nucleus is known as the predominant subcellular location for protein PARylation, multiple studies indicated that PARylation may exist in mitochonria [1][2][3][4][5][6] . And in vitro biochemical studies confirmed several mitochondrial proteins as substrates of PARP1 7 . Despite emerging but debating roles of PARylation in regulating mitochondrial DNA metabolism, there are limited approaches for studying mitochondrial PARylation. No activity-based probes are available to analyze mitochondrial PARylation. Given its high activity and selectivity for protein PARylation, compound 6 was attempted to label mitochondrial PARylation in cells permeabilized with 0.025% Triton X-100. Confocal microscopic analysis of cells treated with 6 showed that in addition to the predominant PARylation in nucleus, considerable PARylation signals were colocalized with mitofilin, a mitochondrial inner membrane protein ( Figure 8A). The observed PARylation signals were suppressed by veliparib inhibitor. These results suggest the presence of PARylation in mitochondria.
We then isolated mitochondria from cells and confirmed the purity of the mitochondria fractions using a monoclonal anti-COX IV antibody for COX IV as mitochondrial protein loading controls, a monolconal anti-GAPDH antibody for GAPDH as a cytoplasmic protein marker, and a polyclonal anti-Histone 2A.Z antibody for Histone 2A.Z as a nuclear protein marker. Consistent with confocal imaging results, immunoblot analysis of the isolated mitochondria fractions clearly revealed significant protein labeling in the presence of 6. Treatment of cells with H2O2 resulted in increased protein labeling by 6 in the mitochondria fractions, which were suppressed by veliparib inhibitor ( Figure 8B). To the best of our knowledge, this is the first time for visualization and labeling of PARylation in mitochondria using an activity-based probe. These results provide new and direct evidence for mitochondrial protein PARylation and demonstrate 6 as a valuable tool for studying cellular PARylation. The results and experimental methods for confocal microscopic and immunoblot analysis of mitochondrial PARylation by 6 are described on pages 15-16 and 26-27.

This study does not isolate and characterize the polymers that are produced by PARP using 3-N3-NAD as a substrate. Neither does is distinguish between the analogs relative rate of reaction for the initiation, the subsequent polymerization and the branching steps.
It is therefore premature to conclude that high molecular branched polymers of similar structure to that formed using NAD as a substrate are the result when using the 3-N3-NAD. Clarification will require additional study, and in this communication the authors should simply and clearly point out that more work will be required to characterize the products of the reaction with the new substrate 3-N3-NAD.
We do appreciate the reviewer's comment. Further work is required to characterize the poly-ADPribose (PAR) polymers formed by compound 6, including the individual steps of initiation, elongation, and branching for PARP1-catalyzed poly-ADP-ribosylation (PARylation) with 6 as the substrate. This is now described in the Discussion section in the second paragraph on page 17.

On page 5-6, yields of 68% are reported for the enzymatic conversion of NR6 to the NAD analog 6. It will be of interest to a reader to know if this refers to an isolated yield of 6 or to a yield calculated on the basis of HPLC analysis at the conclusion of the reaction. The authors should clarify the reporting of this result. It will also be of interest to know the maximum quantities of 6 that have been produced using this enzymatic synthesis.
We do appreciate the reviewer's comment. The yield of 68% for enzymatic synthesis of 6 from NR6 is an isolated yield and calculated based on the weight of purified 6 by HPLC. At the end of the enzymatic reactions, 6 was purified by HPLC and fractions containing the desired product were combined, concentrated, and lyopholized to offer pure 6 as a colorless solid, which was used for the following experiments. We have clarified in the first paragraph on page 6 that the reported yields are isolated yields.
Upon confirming the facile production of 6 from NR6 through the two-step enzymatic approach, the enzymatic reactions were scaled up under the same conditions and 12.2 mg of 6 was produced and purified for the later experiments, which is the maximum quantity of 6 we have produced through enzymatic synthesis. This information is now provided in the first paragraph on page 6.
3. Page 7 line 162-163. Lack of substrate activity for compound 5 could alternately be due to electronic properties of 5. The electron withdrawing 2-azide adjacent to the pyridinium-leaving group is predicted to decrease the rate of pyridine-ribose bond cleavage significantly. In a previous study [A.L.Hanlon et. al., JACS (1994) 116, 12087-12088}, 2'-azido-NAD was shown to be resistant to both NAD glycohydrolase catalyzed and to chemical hydrolysis. The 2'-azide was hydrolyzed by the NAD glycohydrolase approximately 10,000-fold more slowly than was NAD. Compound 5 could therefore be a slowly cleaved or a non-cleavable substrate analog.
We do appreciate the reviewer's comment. The 2′-azido substitution is likely to increase the resistance of the adjacent N-glycosidic bond to chemical and enzymatic cleavage. Thus, the lack of substrate activity for compound 5 could be possibly caused by the blocked formation of branched PAR and/or the significantly decreased cleavage rate of the N-glycosidic bond resulting from electron-withdrawing 2′-azido group. This is now discussed in the first paragraph on page 17.

Page 14, lines 297-298. The experiment that the authors report demonstrates that 6 is not a substrate for SIRT2, not that there is an essential role for the nicotinamide ribotide 3'-OH in NAD analog 6. An azide group is bigger and electronically different from an -OH, and the azide simply might not fit or might constrain that analog into a different confirmation. We can only conclude that compound 6 is not a substrate and that substitution of an -OH for an -N3 is not tolerated by this enzyme.
We do appreciate the reviewer's comment and agree that we can only conclude that compound 6 is not a subtrate of SIRT2 and azido substitution at 3′-OH of NR moiety is not tolerated by SIRT2. This has been revised in the third paragraph on page 17.

5.
It is important that the authors define the purity of the compounds that are tested in this study. It is expected that tested compounds will be highly pure. Therefore, I would recommend the addition of a table or a paragraph to Methods that describes the purity of NAD analogs 1-6 and the method or methods that were used to assess the tested compounds purity. I expect (from the Supplementary Information) that the method will be analytical HPLC, but the purity of the compounds tested is never explicitly defined in the description of their synthesis.
We do appreciate the reviewer's comment. The purity of the reported compounds was determined by HPLC and all of them are highly pure with more than 97.9% HPLC purity. Each NAD + analogue (500 μM) was analyzed separately by reverse-phase HPLC. The purity of 1-6 were then calculated based on the proportion of their corresponding integrated peak areas in total integrated peak areas. The HPLC chromatograms and determined purity of 1-6 are now shown in Figure S8 and Table S2, respectively, on page S20 in Supplementary Information. The method for the purity analysis of NAD + analogues 1-6 is now described in the Methods section on pages 20-21. And we clarify that highly pure 1-6 were used for the study at the end of page 7.
6. The synthesis of the six NAD analogs described in this study begins with the construction of substituted ribose derivatives, the formation of a 1-bromo-ribose derivative, and the displacement of the 1-bromide with nicotinamide to form the nicotinamide-ribose derivative. The stereochemistry of the reaction product must be defined, and it's additionally important to insure that the NAD derivatives that are eventually synthesized contain the expected nicotinamide beta-riboside. Normally the beta-linkage is formed when the bromo-ribose has a 2-benzoyl substituent, but this is not the case for compound 5-7 and NR-5 where a 2-azide is present. In this case and for 1-7 and 3-7 (containing 2-ether substituents) how was the stereochemistry of the products assigned, and how was it proved that the products are the beta-anomers as drawn?
We do appreciate the reviewer's comment. The stereochemistries of generated intermediates Obenzoyl protected NR1-6 were determined as β-isomers on the basis of 1 H-1 H COSY experiments to confirm proton assignments and subsequent NOESY experiments (Figures S2 and S2-a-l). Since the configurations at C4 of compounds 1-7, 2-7, 3-7, 4-6, 5-7, and 6-5 are the same as in the corresponding starting materials, the configurations at C1 of these compounds could be determined using NOESY spectroscopy. As shown in the Figures S2, S2-b, S2-d, S2-f, S2-h, S2-j and S2-l, the proton H 1 of each compound has no correlation with H 3 , H 5 , and H 6 , but correlates with the H 4 proton. These results support the cis relationship between the H 1 and H 4 and that the synthesized products contain desired nicotinamide beta-riboside. The COSY and NOESY spectra for the generated intermediates O-benzoyl protected NR1-6 are shown on pages S8-14 in Supplementary Information.
3. Also, the variations in active site loops that are proposed to underlie the differences in compound 6 utilization ( Figure 5F), this really only represents one potential explanation. There very well could be more subtle differences in active site composition that could give rise to different efficiencies of using compound 6. Since the contribution of the active site loop is not directly tested, it is worth pointing out that other possibilities exist, in order to avoid giving the impression that the active site loop is the confirmed source of specificity compound 6 specificity.
We do appreciate the reviewer's comment and agree that in addition to the highlighted active site loop, more subtle differences at the catalytic sites of those PARP enzymes could result in different activities with compound 6. These possibilities are now discussed in the second paragraph on page 17.

Reviewer 3
This manuscript successfully addresses a major issue in PARP research, and also significantly describes new methodology for the synthesis of NAD analogues. The revised manuscript successfully addresses all of the issues that have been raised during initial review, and I recommend that that paper now be published in Nature Communications.
We do appreciate the reviewer's comments.