This article includes detailed protocols for genetic labeling of mouse skin, surgical denervation, skin biopsy and visualizing labeled epithelia by whole-mount β-galactosidase staining. These methods can be used to test the requirement for nerves in mouse models of normal and pathological skin.
Cutaneous somatosensory nerves function to detect diverse stimuli that act upon the skin. In addition to their established sensory roles, recent studies have suggested that nerves may also modulate skin disorders including atopic dermatitis, psoriasis and cancer. Here, we describe protocols for testing the requirement for nerves in maintaining a cutaneous mechanosensory organ, the touch dome (TD). Specifically, we discuss methods for genetically labeling, harvesting and visualizing TDs by whole-mount staining, and for performing unilateral surgical denervation on mouse dorsal back skin. Together, these approaches can be used to directly compare TD morphology and gene expression in denervated as well as sham-operated skin from the same animal. These methods can also be readily adapted to examine the requirement for nerves in mouse models of skin pathology. Finally, the ability to repeatedly sample the skin provides an opportunity to monitor disease progression at different stages and times after initiation.
Over the past few years, there has been a widening appreciation for the influence of nerves on diseases not typically regarded as classical neuropathies1-4. In the skin, recent experimental evidence has suggested that sensory nerves can modulate diverse pathologies ranging from psoriasis to cancer5-9. This has been demonstrated using techniques such as surgical denervation and pharmacological inhibition of neural function in rodents. In the case of psoriasis, these studies have provided a mechanistic framework for understanding why human psoriatic plaques regress following loss of neural function7,10-12.
Cutaneous nerves can also affect gene expression13,14 and are critical for mechanosensing in normal skin15. In particular, touch dome (TD) epithelia are comprised of a patch of columnar epidermal cells in juxtaposition with neuroendocrine Merkel cells innervated by slowly adapting type 1 (SA1) nerve fibers16-18. TDs mediate light touch sensation and have been shown to display Hedgehog pathway activity5,19. TD maintenance depends on innervation20,21, as nerves secrete Hedgehog ligands to sustain normal TDs and their associated Merkel cells19. In addition, innervation promotes Hedgehog-dependent tumor formation from TD epithelia5. Together, these studies reinforce the notion that intricate molecular interactions occurring between nerves and the surrounding cells in their niche are crucial for normal TD physiology as well as disease.
To interrogate the nature of these interactions, we describe here a series of in vivo techniques for manipulating gene expression in the TD, as well as for harvesting skin biopsies for TD visualization after lineage tracing. Finally, we describe procedures for performing unilateral surgical denervation, wherein nerves are removed from one side of the mouse dorsal skin, while leaving the contralateral side intact as a sham internal control. Several weeks after surgery, denervated and sham control skin are compared to assess changes that occur when nerves are ablated. Although these techniques are described in the context of normal TDs, the denervation procedure has been used to examine the requirement for nerves in mouse models of psoriasis6, wound healing13 and tumorigenesis5. Finally, since the skin is amenable to repeated biopsies, this provides an opportunity to monitor the long-term fates of labeled cells or to assess disease progression over multiple time points.
All procedures described in this protocol were performed in accordance with regulations established by the University of Michigan Unit for Laboratory Animal Medicine.
1. Induce Genetic Recombination in Mice
Note: The Gli1tm3(cre/ERT2)Alj/J mouse strain (Gli1-CreERT2)13 enables targeting of tamoxifen-induced genetic recombination to TD epithelia. Cross this strain with B6.129S4-Gt(ROSA)26Sortm1Sor/J reporter mice (LacZ)22 to generate Gli1;LacZ animals to visualize TD cells by whole-mount staining below.
2. Harvest Skin Biopsies
Note: Depending on the experiment, harvest skin biopsies several days to weeks after tamoxifen induction. For all surgeries, follow standard protocols for rodent surgery, including using sterile gloves, wearing a surgical mask or hair net, and covering the animal with a sterile surgical drape during the procedure.
3. Process Samples for Histology
Note: To fix and process the excised tissue, use either method below depending on application.
4. Visualize Samples by Whole-mount LacZ Staining
5. Surgical Denervation
By generating mice expressing tamoxifen-inducible Gli1-CreERT2 and a LacZ reporter allele, it is possible to visualize TD epithelia and track the fates of these cells over time. The entire denervation procedure typically can be completed within 1 hr per mouse and should cause minimal distress to the animal.
Our previous studies have indicated that nerves are crucial for maintaining both normal TDs as well as their associated Keratin 8+ Merkel cells (Figures 3A–C)5,19. Nerves are also critical for promoting Gli1 expression in the TD (Figure 3D). Given the relatively infrequent appearance of TD clusters throughout the skin (Figure 1B), it is imperative to sample multiple frozen sections to accurately quantitate TD frequency. Typically, we assess 15 non-consecutive sections (each 10 µm thick and ~1 cm long) from both sham and denervated skin from each animal. Following denervation, stable loss of nerves, both at the TD and throughout the skin, can be confirmed by the absence of immunohistochemical staining for standard pan-neural markers such as Neurofilament in either frozen or paraffin sections (Figures 3A and 3C, and as previously reported5). Alternatively, nerves can also be identified by expression of β3-tubulin or PGP9.56,9.
By using Gli1;LacZ mice, it is also possible to confirm both the requirement for nerves in activating Hedgehog signaling in the TD and in maintaining TD cell fate by varying the sequence of tamoxifen-induced recombination and denervation. If denervation is performed prior to recombination, for instance, this would test the requirement for nerves in activating the Hedgehog pathway, as monitored by Cre recombinase activity and levels, which are correlated with Gli1 expression in these animals. On the other hand, if denervation is performed after recombination, this would assess the requirement for nerves in maintaining already-labeled cells in the TD.
Figure 1. Skin biopsy and whole-mount LacZ staining of TD epithelia. (A) (Top) Photograph of mouse after biopsy and prior to suturing. (Lower left) Enlarged image of biopsy site. (Lower right) Skin sample obtained from biopsy with its dermis side spread flat on a dry paper towel. (B) Whole-mount LacZ staining of skin from a Gli1;LacZ mouse, 7 days after tamoxifen induction, depilated just prior to biopsy to improve skin visualization. Gli1+/LacZ+ TDs are labeled as intense blue clusters. Scale bar = 1 cm. Please click here to view a larger version of this figure.
Figure 2. Two approaches for denervating dorsal skin. (A) Cartoon diagram of innervated mouse skin. Red dotted lines indicate a single excision made along the dorsal midline to expose the underlying musculature on the trunk wall (purple) as well as the dermis (grey, asterisk) beneath the reflected skin. Dorsal cutaneous nerves traveling caudally appear to "bend" as they leave the trunk wall (the black arrow indicates one such bend). Nerve segments to be excised are demarcated by black dotted lines. (B) Photograph of intact nerves with sites of bending indicated (arrows). Blue arrows point to 2 nerve segments that will be removed. (C–D) Photographs showing denervation technique. Using ultra-fine forceps, grip the nerves 0.5 cm below their sites of bending and pull outwards. (E) Photograph showing the body cavity after removal of 2 nerve segments (blue arrows). The remaining nerves also need to be removed. (F–I) An alternative approach for nerve removal is depicted, where nerves are snipped at their proximal ends just below where they bend (arrowheads) (G), and also distally, close to their site of entry into the skin (arrowheads) (H). Note that the midline incision in these images is longer than typical for the purpose of better visualization. (J) Photograph of dermal side of the reflected skin flap to one side of the midline. (K) Nerves are outlined (black lines), with larger blood vessels indicated in red. The nerves located on the dermis side of the skin flap also need to be excised. Asterisk, underlying dermis from the reflected skin flap. Scale bar = 1 mm. Please click here to view a larger version of this figure.
Figure 3. Stable loss of nerves and deterioration of TDs after denervation. (A) Immunohistochemistry showing TD epithelia with Keratin 8+ Merkel cells (K8, green) and Neurofilament+ nerves (NF, red) in sham-operated skin. (B) Cartoon depiction, with TD epithelia highlighted in purple, Merkel cells in green, and sensory nerves in blue. (C) Immunohistochemistry showing denervated skin (den) lacking Merkel cell-neurite complexes within the TD area. Dashed yellow lines, hair follicle epithelium. Asterisk, background staining. (D) Whole-mount LacZ staining of dorsal back skin from Gli1LacZ/+ mouse 2 weeks after unilateral skin denervation. In the box to the left of the healed midline incision (arrow), abundant labeled TD epithelia are observed in sham-operated skin. To the right of the midline, TDs are not visible in denervated skin. Scale bar = 10 µm for (A), (C); and 1 mm for (D). Please click here to view a larger version of this figure.
Nerves serve crucial functions not only in sensation, but also in mammalian organ development, maintenance and regeneration13,24-27. As nerves have recently been implicated in diverse skin disorders, the techniques described here can be used to study the requirement for innervation in a variety of animal disease models. Indeed, the unilateral denervation technique allows for the direct comparison of skin with either intact or disrupted nerves from the same mouse. This provides an ideal internal control to compensate for animal-to-animal differences, with subsequent data analyses making use of a paired t-test.
While the procedures described here largely utilize the LacZ reporter gene, these experiments can be adapted such that the Gli1-CreERT2 allele is combined with other fluorescent reporter or conditional alleles to modify gene expression in the TD. For instance, Gli1-CreERT2 mice can be crossed with animals harboring conditional alleles of Patched1 (B6N.129-Ptch1tm1Hahn/J)28 to generate mice that form TD-derived tumors after tamoxifen induction5. It is important to note that the Gli1-CreERT2 strain also induces recombination in a subset of Gli1+ hair follicle stem cells that are physically separated from those in the TD13.
Following denervation, nerves in the skin remain stably ablated for several months (Figure 3C)5,19. In other studies, however, some re-innervation has been reported to occur over time6. The perdurance of the denervated phenotype may depend on the thoroughness of nerve removal, as it is absolutely critical to excise nerve segments between their exit from the chest wall to a point close to the sites of insertion into the dermis of the skin.
Completely removing the hair from the skin prior to biopsy can enhance the ability to subsequently visualize TDs by whole-mount staining (Figure 1B). This is accomplished by applying depilatory cream to clipped skin for 2 min, and then wiping the hair away in an anterior-to-posterior direction using cotton balls. Please note that depilation can affect hair cycle kinetics by promoting entry into the anagen growth phase. Alternatively, hair can be completely removed using a razor blade. In addition, whole mount immunohistochemistry can be performed to visualize TDs and Merkel cells on epithelial sheets separated from the dermis, as has been previously described23.
The possibility remains that surgical denervation may cause inflammation at the surgical site, potentially confounding any observed phenotypes. In our experience, we have not observed significant inflammation after denervation, likely because the collateral tissue damage incurred in the skin is slight if the procedure is done properly. To minimize the possibility that inflammation may affect results, additional controls can be incorporated into the experiment. For instance, we observed that denervation specifically inhibited TD-derived tumors, but not adjacent hair follicle-associated lesions in the same skin samples, arguing that denervation–and not a general wound-induced inflammatory response–likely inhibited tumorigenesis at the TD5.
It is important to note that surgical denervation ablates all cutaneous nerves, including sensory and sympathetic fibers5, and thus provides a general overall assessment of the influence of these nerves on either normal or diseased skin. Other experimental approaches, for instance using pharmacologics such as Botulinum neurotoxin to block neurotransmission, may yield more detailed mechanistic insights7, although it is unclear whether these agents inhibit retrograde secretion of cytokines such as Hedgehog ligands. Alternatively, compounds such as 6-hydroxydopamine have been used to ablate sympathetic nerves in the skin9. In addition, targeting the receptors for nerve-derived factors such as Calcitonin gene-related peptide and Substance P may be useful for interrogating specific interactions between nerves and the surrounding cells within their niche6. Ultimately, multiple strategies may be utilized in combination to identify, or at least rule out, potential signaling mechanisms.
Finally, targeted genetic deletion of nerve-derived factors in the neural lineage using either Wnt1-Cre or Advillin-Cre may represent the gold standard for elucidating the signals that are exchanged between nerves and their niche19. As neither of these strains are tamoxifen-inducible, however, some caution needs to be taken to ensure that disruption of these signals does not impair nerve development or proper targeting of neural afferents. Use of a tamoxifen-inducible Cre such as Advillin-CreERT2 may help circumvent these issues29.
Overall, the techniques described here–a combination of lineage tracing, cell visualization and surgical denervation–offer powerful approaches for studying the influence of nerves on normal and diseased skin. With experience, these procedures can be performed routinely and reliably, while causing minimal distress to the animal–or the investigator.
The authors have nothing to disclose.
The authors would like to thank Autumn Peterson for assistance with mouse photography, Daniel Thoresen for assistance with mice, and Drs. Nicole Ward and Abdelmadjid Belkadi for assistance with surgical denervation. These studies were supported by funding from the National Institute of Arthritis and Musculoskeletal and Skin Diseases (grants R00AR059796 and R01AR065409); the University of Michigan Department of Dermatology; the Biological Sciences Scholars Program; the Center for Organogenesis; the University of Michigan Comprehensive Cancer Center; and the John S. and Suzanne C. Munn Cancer Fund. S.C.P. was supported by funding from the National Institute of General Medical Sciences (grant T32 GM007315). This work was also supported by the NIH Intramural Research Program, Center for Cancer Research, National Cancer Institute.
Alcohol prep pads | PDI | B339 | |
AnaSed (Xylazine) | Lloyd | NADA 139-236 | |
Antibody, anti-Keratin 8 | Developmental Studies Hybridoma Bank | TROMA-I | rat antibody, use at 1:500 concentration |
Antibody, anti-Keratin 17 | Cell Signaling | #4543 | rabbit antibody, use at 1:1,000 concentration |
Antibody, anti-Neurofilament | Cell Signaling | C28E10 | rabbit antibody, use at 1:500 concentration |
Betadine prep pads | Medline | MDS093917 | |
Carprofen (Rimadyl) | Zoetis | ||
Cordless rechargable clipper | Wahl | trimmer model 8900 | |
Corn Oil | Sigma-Aldrich | C8267 | |
Cryostat | Leica | CM1860 | |
DAPI | ThermoFisher Scientific | D1306 | use at 1:1000 concentration |
Deoxycholate | Sigma-Aldrich | D6750 | |
Depilatory Cream | Nair | N/A | |
Dimethylforamide | Sigma-Aldrich | 319937 | |
Dimethyl Sulfoxide (DMSO) | Sigma-Aldrich | D8418 | |
Glutaraldehyde | Sigma-Aldrich | G5882 | |
ImmEdge Pen | Vector Laboratories | H-4000 | |
Ketamine HCl | Hospira | NDC 0409-2051-05 | |
Magnesium chloride | Sigma | M8266 | |
Micro cover glass | VWR | 48404-454 | |
Micro Slides | VWR | 48311-703 | |
10% Neutral Buffered Formalin | VWR | BDH0502-4LP | |
6-0 nylon sutures | DemeTECH | NL166012F4P | |
Octylphenyl-polyethylene glycol | Sigma-Aldrich | I8896 | |
O.C.T. Compound | Sakura Tissue-Tek | 4583 | |
Paraformaldehyde | Sigma-Aldrich | 158127 | |
Pottasium ferrocyanide | Sigma-Aldrich | P9387 | |
Pottasium ferricyanide | Sigma-Aldrich | 702587 | |
Sodium phosphate monobasic | Sigma-Aldrich | P9791 | |
Sodium phosphate dibasic | Sigma-Aldrich | S5136 | |
Sucrose | Sigma-Aldrich | 84097 | |
Tamoxifen | Sigma-Aldrich | T5648-1G | |
Ultra fine forceps | Dumont | 0103-5-PO | |
Vectashield | Vector Laboratories | H1000 | |
X-gal | Roche | 10 651 745 001 | Disolve in dimethylforamide to create 50x stock prior to use |