(a) Double immunohistochemistry for ASMACy3 and K5 (an epidermal keratin detected by anti-keratin 5) on P1 and P3 thoracic skin sections from wild-type mice. NEM organogenesis takes place between P1 and P3 (arrowheads). ASMACy3+ tissues at P1 are blood vessels. (b) Triple immunohistochemistry for ASMACy3, TH and K5 on P5, P7 and P11 NEM sections from wild-type mice. Note that TH+ axons innervating the NEM increase in density from P5 to P11 (arrowheads). (c) Triple immunohistochemistry for NPY, TH and RET on NEM sections from adult wild-type mice. Note TH+NPY+RET+ sympathetic axons (insets). Insets are single optical sections selected from a Z-stack of the region of interest, acquired at 2 μm intervals. (d) Triple immunohistochemistry for TH, GFP and PGP9.5 (protein gene product 9.5, encoded by ubiquitin-C-terminal hydrolase 1, to detect all nerve fibers) on adult NEM sections from RetCreERT2;R26RmGFP shows TH and RetmGFP expression in different location of the PGP9.5+ fiber. Presence of both markers suggests a RET+TH+ origin. (e) Innervation at the base of NEM (ASMACy3+) is initiated at P6 from TH+RetCFP+ fibers (inset). (f) Double immunohistochemistry for ASMACy3 and K5 on P1 and P3 dorsal skin sections from wild-type mice. PEM organogenesis takes place between P1 and P3 (arrowhead). (g-j) Triple immunohistochemistry for ASMACy3, TH and K5 or CD31 on P3, P5, P7 and P11 PEM sections from wild-type mice. Note TH+ sympathetic axons (arrows) in close association with CD31+ blood vessels (g-h, arrowheads), located in proximity of the PEM (asterisks), suggesting that PEM innervation is received from nerves associated with nearby blood vessels. Innervation of PEM is initiated at P7 (i, insets) and it is robust at P11 (j). (k) Triple immunohistochemistry for TH, NPY and ASMACy3 on P20 PEM sections from RetCFP mice. TH+ sympathetic axons (arrows) innervating the PEM do not show immunoreactivity for NPY, showing that PEM is not a target of NA2 or NA3 neurons. Insets are max intensity projections of a Z-stack of the region of interest, acquired at 2 μm intervals. (l) Innervation of PEM is initiated at P6; TH+RetCFP+ fibers are present on blood vessels (arrows, inset) and on the base of PEM (arrowheads, inset). (m) Sympathetic DbhTOM+ traced NEM fibers (in DbhCre/+;R26RTOM/+ mice) express general noradrenergic markers such as TRKA (inset) and NA1-NA3 markers, such as GFRα3 (arrowheads). (n) RetCFP+ (in RetCFP mice, P20) PEM fibers contain TRKA (inset). (o-s) FastBlue (FB) retrograde tracing of axons innervating NEM (o, p) and PEM (q-s). (o) NEM injection site. Double immunohistochemistry for ASMA and TH on NEM samples from Wnt1-Cre;R26RTOM/+ adult mice injected with FB (green signal). The nipple is outlined, arrowheads point at NEM. (p)Triple immunohistochemistry for NPY, RET and TOMATO on SG of traced Wnt1-Cre;R26RTOM/+adult mice. Traced cells express both NPY and RET, matching the molecular signature of NA2 neurons. (q) PEM injection site. Double immunohistochemistry for ASMACy3 and K5 on back skin samples from TrkACre;R26RTOM/+ adult mice injected with FB (green signal). Arrowheads point at PEMs. (r, s) Triple immunohistochemistry for RET, NPY and TOMATO (recapitulating TrkA expression, r) and for ENC1, GFRα2 and TOMATO (s) on SG from traced TrkACre/+;R26RTOM/+ adult mice. Note traced FB+TrkATOM+ neurons in (r) express RET (arrowheads) but are negative for NPY and that FB+TrkATOM+GFRα2+ neurons in (s) are negative for ENC1. (t, u) Quantification of FB+ traced cells following injection of NEM (t) or PEM (u). FB+ cells expressing NPY and RET (NPY+RET+) were classified as NA2, those expressing only NPY (NPY+RET-) as NA3 and those expressing only RET (NPY-RET+) as NA4-5 neurons. FB+ cells that were negative for both markers were labeled as ÒunassignedÓ. Neurons classified as NA2 were virtually absent when the tracer was injected in the dorsal skin PEMs (t) while when injected into the NEMs traced neurons represent more than one third of all traced cells (u). The skin is richly innervated by autonomic neurons of various types, including cutaneous blood vessels, nipple and hair follicle erector muscles. Thus, as shown in (o) and (q), injecting 0.25 ml FB results in some spread of the tracer, resulting in traced neurons also from nearby autonomic targets. Therefore, in (t), we conclude that the remaining 70% of FB+ neurons are traced from blood vessels (NA3 neurons) or from ventral hair follicle PEMs (NA4-5 neurons) of the tissue surrounding the nipple; in (u), the greatest part of FB+ cells are NA4-5 and the second most abundant group is represented by presumed blood vessel NA3 neurons. (a-m), (n), (o) and (q) are representative images from experiments carried out on at least two animals. Scale bars: 100 μm in (o) and (r); 20 μm in (a-c) and (e-l); 10 μm in (n), (p), (q), (s); 2 μm in (d), (m) and all insets.