Generation of a CRF1-Cre transgenic rat and the role of central amygdala CRF1 cells in nociception and anxiety-like behavior

  1. Marcus M Weera
  2. Abigail E Agoglia
  3. Eliza Douglass
  4. Zhiying Jiang
  5. Shivakumar Rajamanickam
  6. Rosetta S Shackett
  7. Melissa A Herman
  8. Nicholas J Justice
  9. Nicholas W Gilpin

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This manuscript describes, characterizes, and validates a novel transgenic tool that will be useful for the study of stress neurobiology and the function of the stress-related neuropeptide corticotropin releasing factor. This tool will be especially relevant to the study of stress and related fields, such as addiction, in which rats are a critical model system. There is comprehensive histochemical and functional validation of the novel rat that broadly supports its proposed utility for visualizing and providing access to central amygdala CRF1 receptor-expressing neurons.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #2 and Reviewer #3 agreed to share their name with the authors.)

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Corticotropin-releasing factor type-1 (CRF 1 ) receptors are critical to stress responses because they allow neurons to respond to CRF released in response to stress. Our understanding of the role of CRF 1 -expressing neurons in CRF-mediated behaviors has been largely limited to mouse experiments due to the lack of genetic tools available to selectively visualize and manipulate CRF 1 + cells in rats. Here, we describe the generation and validation of a transgenic CRF 1 -Cre- td Tomato rat. We report that Crhr1 and Cre mRNA expression are highly colocalized in both the central amygdala (CeA), composed of mostly GABAergic neurons, and in the basolateral amygdala (BLA), composed of mostly glutamatergic neurons. In the CeA, membrane properties, inhibitory synaptic transmission, and responses to CRF bath application in td Tomato + neurons are similar to those previously reported in GFP + cells in CRFR1-GFP mice. We show that stimulatory DREADD receptors can be targeted to CeA CRF 1 + cells via virally delivered Cre-dependent transgenes, that transfected Cre/ td Tomato + cells are activated by clozapine-n-oxide in vitro and in vivo, and that activation of these cells in vivo increases anxiety-like and nocifensive behaviors. Outside the amygdala, we show that Cre- td Tomato is expressed in several brain areas across the brain, and that the expression pattern of Cre- td Tomato cells is similar to the known expression pattern of CRF 1 cells. Given the accuracy of expression in the CRF 1 -Cre rat, modern genetic techniques used to investigate the anatomy, physiology, and behavioral function of CRF 1 + neurons can now be performed in assays that require the use of rats as the model organism.

Article activity feed

  1. Version published to 10.7554/elife.67822 on eLife
  2. eLife

    Evaluation Summary:

    This manuscript describes, characterizes, and validates a novel transgenic tool that will be useful for the study of stress neurobiology and the function of the stress-related neuropeptide corticotropin releasing factor. This tool will be especially relevant to the study of stress and related fields, such as addiction, in which rats are a critical model system. There is comprehensive histochemical and functional validation of the novel rat that broadly supports its proposed utility for visualizing and providing access to central amygdala CRF1 receptor-expressing neurons.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #2 and Reviewer #3 agreed to share their name with the authors.)

  3. eLife

    Reviewer #1 (Public Review):

    In the manuscript by Weera et al., the authors describe a novel transgenic rat in which Cre/tdTomato is expressed under the Crhr1 promoter effectively fluorescently tagging CRF1 containing neurons in the rat brain and allowing for cre-dependent manipulations; they use cre-dependent chemogenetic manipulations in this paper. They validate this rat line by focusing on CRF1 expression in the CeA. Through use of ISH, IHC and electrophysiology they validate and characterize CRF1-expressing neurons (i.e. tdTomato+/Cre+) neurons. The authors go on to use chemogenetics to demonstrate that increased activity of CRF1 containing neurons is anxiogenic and potentiates mechanical nociception. Generally, they do not report sex differences, although there does seem to be a sex difference in the spontaneous firing rate of CRF1+ CeA neurons.

    In my view, this paper appears to be two incomplete studies put into one manuscript. The manuscript seeks to characterize a novel tool, generated by this laboratory, that could be of great value to the CRF field. As is the case with other transgenic mouse and rat tools that have come before it, this requires a brain wide characterization that describes the specificity and penetrance of cre-recombinase under Crhr1 promoter in several brain regions, not just one. Take the case of the Pomerenze Crh-Cre rat. In that case, cre is only expressed in CRF+ neurons that are GABAergic, not glutamatergic, excluding cre expression in the PVN. If this new tool is going to be widely utilized, it is important we understand what exactly it is and it is not. It cannot be assumed that the specificity and penetrance are uniform across the brain. If you look at the commonly used Crh-cre mouse line, there are significant differences in penetrance across regions (see Walker et al., 2019, Neuropharmacology). Without this brain-wide information, the tool has reduced utility to CRF researchers.

    In addition, the authors make the case for the necessity of this tool "given" that there are behavioral studies that can be done only in rats, but not mice. However, they do not really put this supposition to task. Can this tool give us new insight into the role of CRF1 in the CeA or any brain region that cannot be done in mice. This would be the kind of experiment that would be of broad interest to the stress neurobiology field.

    The second study within this manuscript is utilizing this tool to gain insight into the role of CRF1 regulation of CeA neurons in males and females and how it impacts anxiety-like behavior and nociception. First, they uncover some interesting sex differences in the spontaneous firing of CRF1+ CeA neurons, yet they do not follow up that finding. The anxiety-like behavior replicates the consensus of findings in the literature. However, the role of CRF in the CeA in nociception is a bit controversial and I think delving into that controversy using this new tool would have made the manuscript more impactful and of interest to a broad readership. To summarize a collection of studies, there are some that indicate CRF/CRF1 in the CeA mediates stress-induced analgesia, others show that intra-CeA CRF increases paw withdrawal latency to both mechanical/thermal stimulation and others that show that stress causes hyperalgesia via CRF1 function in the CeA. It seems like a missed opportunity to offer new insights into this question.

  4. eLife

    Reviewer #2 (Public Review):

    Major Comments:

    1. The important contribution of this work is the introduction of a model organism that can improve on available mouse lines, in terms of what can be learned about CRFR1 neurons and behavior. The advantages of the rat line are many, including behavioral repertoire, anatomical targeting precision and body size/blood volume. The authors could do a better job of emphasizing this advantage throughout the manuscript, and provide concrete examples of how the rat line improves on murine models. Moreover, it will be important to note specific examples of how the 'behavioral repertoire' of the mouse is limited (page 4) relative to rat.

    2. Given this is a new model, some additional mapping of iCRE/TDtomato is needed to assure that the fidelity of transgene expression viz. known CRFR1-expressing regions extends beyond the amygdala.

    3. Some behavioral data in Figure 7 analyzed by t-test instead of ANOVA. The rationale for this analysis scheme is unclear. Authors explain that control and DREADD virus readouts were analyzed separately because control virus group was treated as a replication of a previous experiment. If the study was performed in the same cohort of rats, it would still seem appropriate to perform an ANOVA. If performed at another point in time or a separate experiment, breaking it out would be most appropriate.

    4. Male and female data have been combined for behavioral readout of Figure 7. Given the low 'n's for each sex here (as low as 3/sex in some groups), the argument for the conclusion of 'no sex difference' in anxiety-like behavior is extremely weak.

  5. eLife

    Reviewer #3 (Public Review):

    The present manuscript reports efforts to develop and validate a genetically-modified rat to allow visualization of and recombinatorial genetic access to cells that express the CRF1 receptor. Such a model is highly significant because of the inadequacy of current analogous genetic models in mice or other species for studying complex behaviors, especially those that involve small brain nuclei.

    In addition to the highly significant and much needed research tool sought and provided evidence of, there are many other major strengths to this paper, including:

    1. a well-reasoned bacterial artificial chromosome (BAC) design that drives both iCre recombinase and Tomato reporter via an optimized IRES element, with adequate pre-injection sequencing verification.

    2. the high experimental rigor with consideration of batch, blind and counter-balanced analyses at all levels of analysis.

    3. RNAScope analysis that clearly demonstrates strong (90-95%) cellular co-expression of Cre mRNA within CRF1 receptor mRNA-expressing amygdala cells and with the distinction seen between CeL vs. CeM populations recapitulating previously seen physiologic distributions.

    4. Successful utilization of Tomato-fluorescence to perform a comprehensive electrophysiologic assessment of identified cells that included intrinsic membrane properties, I-V relations, burst classification, and pharmacologically-isolated GABAA-R-mediated sIPSCs.

    5. Successful demonstration that fluorescently-identified cells are still predictably responsive to CRF stimulation effects on firing rate.

    6. Successful ex vivo evidence of recombination, inferred from functionally-relevant expression of an excitatory DIO-DREADD, observed as CNO actuator-driven increases in c-Fos expression and firing rate of fluorescently-identified neurons.

    7. Successful in vivo evidence of recombination, inferred from CNO actuator-driven increases in expected anxiogenic-like (plus-maze, open field) and mechanical nociceptive (von Frey) behavioral phenotypes.

    8. Demonstration that these CNO phenotypes differed from those seen under vehicle administration as well as from CNO-treated rats that received non-DREADD control AAVs, supporting the proposed genetic recombination with the DIO-DREADD construct.

    9. Clearly described, appropriate statistical analysis with clear results.

    10. Consideration of sex differences in both the genetic model and the central amygdala CRF1 system.

    Lesser, moderate weaknesses of the present manuscript include:

    1. The model's potential utility for studying brain regions other than the central amygdala is unknown because histochemical and functional validation were only performed for the central amygdala, In the model's favor, inspection of figures shows cortical and stria terminalis Tomato expression where qualitatively expected based on known physiologic CRF1 expression in rat. Estimates of CRF1-Cre concordance in these other brain regions was not provided however. I appreciate that functional validation of all brain regions was beyond the scope of the present work.

    2. The location, copy #, fidelity or potential mosaicism of BAC transgene insertions after germline transmission was not mentioned.

    3. Some aspects of histochemical concordance were not provided or explored. For example, the proportion of Cre-expressing neurons that expressed CRF1 mRNA was not reported to confirm specificity (it appears to be even higher which would be good]? Cellular co-expression of Tomato with CRF1 mRNA also was not explicitly quantified.

    Further, the measures of co-expression were based on a binary distinction between those with 3 or more (positive) vs fewer than 3 puncta (negative), and the basis for this threshold was not justified (e.g., PMID: 31451604). Because only a single binary threshold was used, it also is not known whether there is a continuous (vs. binary) relationship between CRFR1 and Cre expression on the other. The former could mean that different levels of expression may be associated with different degrees of labeling or genetic access to CRF1-positive cells.

    Finally, there might be slight sex differences in co-expression in the model. Using the binary distinction, there appeared to be greater co-expression of Cre mRNA within CRF1-positive cells in 1 sex vs. the other (95.8%+1.3 vs. 90.5%+2.6). It would have been useful to rule out statistically significant differences in semi-quantitative expression or semi-quantitative (rather than only binary) co-expression of CRF1 and iCre puncta.

    4. It is not clear whether the iCre expression obtained is sufficient to silence these populations via Cre-dependent constructs (e.g., for studies to reverse physiologic effect).

    5. It is unknown whether the model recapitulates physiologic patterns of CRF1 expression across cell types (e.g., neuron and glia in brain, corticotrophs in pituitary, etc.).

    6. The manuscript does not discuss possible genetic strain of origin issues. The rat was generated using a BAC that involves DNA from a different strain (F344) than that bred into (Wistar), so potential strain of origin differences in regulatory or enhancer elements for the Crhr1 gene might co-segregate in early generations with the transgene. Similar considerations are true with respect to linked alleles if the strain of the gamete donors, which was not specified, differed from Wistar. Such issues would become of lesser concern with continued Wistar breeding, but are not mentioned.

    7. The nomenclature does not follow IUPHAR and Rat Genome Database guidelines. While the the protein is appropriately referred to as the CRF1(subscript) receptor, the preferred IUPHAR and RGD nomenclature for the rat gene is Crhr1(italicized) (see https://www.guidetopharmacology.org/GRAC/FamilyDisplayForward?familyId=19 and https://rgd.mcw.edu/rgdweb/report/gene/main.html?id=61276).

    Overall, the authors appear to have developed and provided initial validation of a BAC transgenic Cre-Tomato reporter rat that is anticipated to be invaluable in dissecting the functional significance of amygdala CRF1 receptor systems in complex behaviors.

  6. eLife

    Reviewer #4 (Public Review):

    In this study, Weera and colleagues generated a new transgenic rat that allows for the visualization and manipulation of neurons expressing the corticotropin-releasing factor type-1 reeptors (CRFR1) in a cre-dependent manner. The description and validation of this transgenic rat line was rigorous, the paper is well written and the presentation of results is clear. More importantly, this new tool is important and useful for the neuroscience community as it will facilitate future studies aimed at studying the function of CRFR1-expressing neurons in different behavioral contexts.

    Strengths:

    1. The authors included male and female rats throughout the study, with most of the experiments with sufficient statistical power to allow conclusions related to sex as a biological variable. When pooled samples were shown, the sex of the rat used was clearly labeled. This level of transparency is critical for reproducibility and future use of this transgenic rat.

    2. The authors go above and beyond in the characterization of the transgenic line by showing not only fidelity of expression in the CeA but also the electrophysiological, synaptic and behavioral characterization of CRFR1-expressing neurons in the CeA of transgenic rat.

    Weaknesses:

    1. The characterization of the transgenic rat is thorough but entirely focused on the amygdala. Learning about the fidelity of expression in other brain regions is important to assess the potential use of these transgenic rats to study CRF-dependent behaviors and physiology in other brain structures, broadening its applicability to the neuroscience community.
  12. Version published to 10.1101/2021.02.23.432551 on bioRxiv