In vivo cell type-specific CRISPR knockdown of dopamine beta hydroxylase reduces locus coeruleus evoked wakefulness

Locus coeruleus (LC) neurons in the brainstem have long been associated with attention and arousal. Optogenetic stimulation of LC-NE neurons induces immediate sleep-to-wake transitions. However, LC neurons also secrete other neurotransmitters in addition to NE. To interrogate the role of NE derived from the LC in regulating wakefulness, we applied in vivo cell type-specific CRISPR/Cas9 technology to disrupt the dopamine beta hydroxylase (dbh) gene selectively in adult LC-NE neurons. Unilateral dbh gene disruption abolished immediate arousal following optogenetic stimulation of LC. Bilateral LC-specific dbh disruption significantly reduced NE concentration in LC projection areas and reduced wake length even in the presence of salient stimuli. These results suggest that NE may be crucial for the awakening effect of LC stimulation and serve as proof-of-principle that CRISPR gene editing in adult neurons can be used to interrogate gene function within genetically-defined neuronal circuitry associated with complex behaviors.

the actual co-release of both DA and NE are instructive for awakening. Therefore, please remove the phrase "and not other neurotransmitters co-released in the LC" in all concluding sentences.
• The results with dbh CRISPR could be interpreted in three alternative ways to the authors' interpretation. a) Locus coeruleus function may be altered by dbh knockout beyond loss of NE. The authors will need to show intact anatomy, projection, and release properties of LC neurons (release of something other than NE). b) Presumably, dopamine accumulates as a result of dbh knockout. Could the phenotype due to increased DA release? The authors could address this with a TH knockout which abolishes the synthesis of both DA and NE. c) The results could be due to off-target effect of the CRISPR approach. The best ways to address this are to have different guide RNAs in different vectors and show the same phenotype, or ideally to perform a rescue experiment with dbh. If a rescue experiment is performed, it would allay some of the concerns in (b) but not (a). The authors should either address the above concerns experimentally, or discuss them explicitly in the text and change the title to "effect of locus coeruleus dbh", not norepinephrine. Minor comments • We have learnt in many scenarios that CRISPR is not efficient in postmitotic cells. It is surprising that the efficiency is so high (85% reduction in DBH immunoreactivity) in this case. Why do the authors think this is the case? Please discuss to highlight the impact of the results. • Fig 2d, Fig 2c and 2d, please display the EMG with different scales -it was almost impossible to see any change with this scale of plotting.

Reviewer #3 (Remarks to the Author):
The MS titled, "In vivo cell type-specific gene editing reveals the awakening effect of locus coeruleus norepinephrine" is about a study showing that a selective disruption of Dbh in LC neurons reduces the arousing effect normally induced by LC activation. This is a straightforward study of broad interest that, on the whole appears to be well carried out. There are two important controls whose absence may be of concern. First, it is presumed that DA release by these LC TH-expressing cells is unaffected by the manipulation (Dbh mutation) but this has not been tested. If this is beyond the scope of this study then the conclusion, "NE release and not other neurotransmitters co-released in the LC" needs amendment limited to consideration of the necessary role of Dbh. Second, comparable activation of LC SgDbh to LC control neurons is also presumed. This too should be verified. It is all the more critical in light of figure2b "typical" labelling of ChR2-eYFP in sgDbh vs sgControl which looks to be different.
Minor points: 1. It might be nice to see what TH/Cas9 looks like in the VTA that should also express TH:Cre. 2. A reference for the DJ serotype would be helpful 3. An earlier study might be referenced for the use of floxed alleles and cre-expressing AAV's (Scammell, et al, JoNS 2003) 4. On line 88 was the "sdDbh or sgControl" meant or should this be an "and" (2 injections in the same mouse). 5. In figure 2g,h what does the probability of firing refer to? There are multiple transitions/mouse and multiple mice (n=?) Are these avg latency probabilities/mouse? If so what is the range of the number of transitions measured/ mouse? 6. In the fig2 legend, rather than "light-evoked" maybe "light-activated ChR-evoked" should be used. 7. It looks like fig2e,f involves matched data-at least for no-light vs sgDbh. This should be illustrated by showing each matched pair and the avg +/-SEM. If all three conditions are matched, please show this as well.
In this original study, Yamaguchi & de Lecea used CRISPR/Cas9 technology to knockout dopamine beta hydroxylase (dbh) gene selectively in post-mitotic norepinephrine neurons of the locus neurons (LC-NE) in mice. The authors provide convincing evidence for in vivo targeting of dbh gene and demonstrate the necessity of norepinephrine as a wake-and attention-promoting neuromodulators using in vivo optogenetic sitmulation of recombinant dbh KO LC-NE neurons using EEG/EMG sleep recording and behavioural response to rat odors.
The authors provide careful control of CRISPR/Cas9 recombination, control conditions and NE level quantification after LC-NE targeting.
These data are very interesting and provides new tools to manipulate (genetically) neural circuits locally, without the limitations, mainly the time to generate double/triple transgenic line and the temporal resolution of the CRISPR/Cas9 approach.
The manuscript and figures are well presented, experimental procedures and results are convincing, and, supportive of the conclusions. We do appreciate the time and effort you have dedicated to providing insightful feedback on ways to strengthen our paper.
Although the authors provide convincing evidences that NE alone is supporting the arousal effect of LC-NE neurons, the question remains as to whether LC-NE are expressing/releasing GABA or Glutamate?
While we did not directly examine the release of GABA and Glutamate from the dbh-/-LC, we confirmed that firing rate and intrinsic electrophysiological properties were not different between sgControl and sgDbh-expressing LC cells (Figure2a, b). Also, we confirmed the number of LC cells was comparable between dbh-/and control mice (Figure1e). Taken together, these results indicate that expression of sgDbh and Cas9 does not have detrimental effect on LC cells.
Similarly, these data suggest that most of the NE released by LC neurons results from de novo synthesis of NE (through dbh action onto dopamine molecule). Then, what is the role of monoamine vesicular transporter on NE vesicular content. Could the authors discuss this?
Norepinephrine molecules released into synapse are absorbed back into cytoplasm of LC cells by the norepinephrine transporter. Those norepinephrine molecules would be repackaged into synaptic vesicles. This pathway can regulate the amount of norepinephrine in vesicles in addition to de novo synthesis. Since we disrupted the dbh gene specifically in the LC in this study, the NE amount of LC synaptic vesicles are totally dependent on absorption of NE derived from other norepinephrine neurons. Here, we show the wake lengths of LC-specific dbh-mutated mice were significantly reduced. Also, we found the amounts of norepinephrine in the LC target areas are reduced. These results indicate that the amount of norepinephrine, derived from other areas and then absorbed by the LC, was not sufficient to rescue NE content in the LC. We have expanded the discussion to address this issue in main text.
Minor comments: -L42-46: The link of psychostimulants, wake-promoting effect and Dopamine/Norepinephrine is clear, but the rationale of the parapgraph is unclear. Please reword the sentence.
-few typos We have edited this paragraph to make the rationale clear.
Reviewer #2 (Remarks to the Author): Yamaguchi and de Lecea applies CRISPR/Cas9 technology in vivo to disrupt the dopamine beta hydroxylase gene in TH+ neurons in the locus coeruleus, and shows a critical role of norepinephrine in the transition to and maintenance of awakening. The results are straightforward, and interesting both technically and biologically. The manuscript can be accepted conditioned upon adequate addressing of the following questions. We wish to express our appreciation to you for their insightful comments on our paper. We feel the comments have helped us significantly improve the paper.
Major comments • Although the manuscript shows a crucial role of NE, it does not address other neurotransmitters co-released in the LC. One could imagine, in principle, a scenario where the actual co-release of both DA and NE are instructive for awakening. Therefore, please remove the phrase "and not other neurotransmitters co-released in the LC" in all concluding sentences.
As you mentioned, the result of dbh knockout alone cannot exclude the possibility that the other neurotransmitters than NE also have the awakening effect. Thus, we deleted the sentences of "and not other neurotransmitters co-released in the LC" in all concluding sentences.
• The results with dbh CRISPR could be interpreted in three alternative ways to the authors' interpretation. a) Locus coeruleus function may be altered by dbh knockout beyond loss of NE. The authors will need to show intact anatomy, projection, and release properties of LC neurons (release of something other than NE). To address this question, we examined the electrophysiological properties of LC neurons with cell-attached recordings. The LC cells were held at -70 mV and their firing rate was determined. Firing rate was not different between sgControl and sgDbh expressing cells (Fig2a). Also, the current responses to hyperpolarizing voltage were not different between sgControl and sgDbh-expressing LC cells (Fig2b). Importantly, these results were consistent with those by Paladini and colleagues (2006), who found no intrinsic differences in LC cells from control and dbh KO mice. Also, we confirmed that the numbers of LC cells were comparable between dbh-/and control LC (Fig1e). Taken together, these results indicated that dbh disruption with sgRNA and Cas9 does not alter the integrity of LC cells. Thus, we presumed that the phenotype of LC-specific dbh knockout mice was not due to the altered function of LC cells beyond loss of NE. These results and discussions were incorporated in main text. b) Presumably, dopamine accumulates as a result of dbh knockout. Could the phenotype due to increased DA release? The authors could address this with a TH knockout which abolishes the synthesis of both DA and NE. We have added a new Figure.2d to show that the level of dopamine is not altered between dbh-/and control LC neurons. It is probably because dopamine is quickly metabolized by other enzymes than Dbh in the LC cells. We have added this discussion in L161-164.
c) The results could be due to off-target effect of the CRISPR approach. The best ways to address this are to have different guide RNAs in different vectors and show the same phenotype, or ideally to perform a rescue experiment with dbh. If a rescue experiment is performed, it would allay some of the concerns in (b) but not (a).
During the past six months we have tried to rescue the expression of dbh in sgDbh-infected LC by injecting AAV carrying sgDbh-resistant dbh cDNA to address this question. However, unfortunately, the sgDbh-resistant dbh did not express well in the LC. Instead, we showed the number of cells was comparable between dbh-/and control (Fig1e) and also electrophysiological properties of dbh-/-LC were intact (Fig2a, b). Therefore, we presumed that the result was not due to the detrimental off-target effect of dbh CRISPR.
The authors should either address the above concerns experimentally, or discuss them explicitly in the text and change the title to "effect of locus coeruleus dbh", not norepinephrine.
We have changed the title to "effect of locus coeruleus dopamine beta-hydroxylase".
Minor comments • We have learnt in many scenarios that CRISPR is not efficient in postmitotic cells. It is surprising that the efficiency is so high (85% reduction in DBH immunoreactivity) in this case. Why do the authors think this is the case? Please discuss to highlight the impact of the results. In this study, we adopted the dual sgRNA strategy to improve the efficiency of gene disruption. Also, we used AAV-DJ, a new efficient AAV serotype which can infect almost 100% of LC neurons. Because sgDbh and Cas9 are constitutively expressed in the AAV sgDbh infected-LC neurons and then Cas9 cleaves the target sequence until it gets mutated, we can expect all target exons in LC neurons infected with AAV are mutated in either case. In Fig1g and h, we showed the percentage of frameshift mutations in all mutated dbh Exon2 and Exon3 after the injection the AAV-sgDbh were 82.6% and 85.4%, respectively. Therefore, focusing on a single allele, the percentage that at least one exon of exon2 or exon3 has a frameshift mutation is 100(1-(17.4/100 x 14.6/100))=97.4%. Thus, the percentage that both alleles have frameshift mutations resulting in the dbh disruption would be 100(0.974 x 0.974) = 94.9%. This is the reason why we obtained such a high efficiency of gene disruption in vivo.
This discussion was added in main text.
• Fig 2d, For Fig 2c and 2d, please display the EMG with different scales -it was almost impossible to see any change with this scale of plotting. We have replaced the figure for the better understanding of readers.
Reviewer #3 (Remarks to the Author): The MS titled, "In vivo cell type-specific gene editing reveals the awakening effect of locus coeruleus norepinephrine" is about a study showing that a selective disruption of Dbh in LC neurons reduces the arousing effect normally induced by LC activation. This is a straightforward study of broad interest that, on the whole appears to be well carried out. There are two important controls whose absence may be of concern. First, it is presumed that DA release by these LC TH-expressing cells is unaffected by the manipulation (Dbh mutation) but this has not been tested. If this is beyond the scope of this study then the conclusion, "NE release and not other neurotransmitters co-released in the LC" needs amendment limited to consideration of the necessary role of Dbh. We would like to thank the reviewer for the careful reading and insightful comments on our manuscript. We have added a new Fig2d to show that the level of dopamine is not altered between dbh-/and control LC neurons. Also, we have deleted the phrase "and not other neurotransmitters co-released in the LC" in all concluding sentences to establish a clear focus.
Second, comparable activation of LC SgDbh to LC control neurons is also presumed. This too should be verified. It is all the more critical in light of figure2b "typical" labelling of ChR2-eYFP in sgDbh vs sgControl which looks to be different.
We have added a new Fig2a, b to show the electrophysiological properties of sgDbh-infected LC neurons are comparable to the control cells. Also, we have replaced the image of ChR2-eYFP to show more "typical" expressions of ChR2-eYFP in the LC.
Minor points: 1. It might be nice to see what TH/Cas9 looks like in the VTA that should also express TH:Cre. We agree with the reviewer and have added a new Supplemental Fig1 to show that VTA Th-positive neurons also express Cas9.

A reference for the DJ serotype would be helpful
We have added a reference of AAV-DJ serotype in L69.
3. An earlier study might be referenced for the use of floxed alleles and cre-expressing AAV's (Scammell, et al, JoNS 2003) We have added the reference in L74.
4. On line 88 was the "sdDbh or sgControl" meant or should this be an "and" (2 injections in the same mouse). We have corrected the word. 6. In the fig2 legend, rather than "light-evoked" maybe "light-activated ChR-evoked" should be used. We have corrected the word.