Vivid biofluorescence discovered in the nocturnal Springhare (Pedetidae)

Biofluorescence has been detected in several nocturnal-crepuscular organisms from invertebrates to birds and mammals. Biofluorescence in mammals has been detected across the phylogeny, including the monotreme duck-billed platypus (Ornithorhyncus anatinus), marsupial opossums (Didelphidae), and New World placental flying squirrels (Gluacomys spp.). Here, we document vivid biofluorescence of springhare (Pedetidae) in both museum specimens and captive individuals—the first documented biofluorescence of an Old World placental mammal. We explore the variation in biofluorescence across our sample and characterize its physical and chemical properties. The striking visual patterning and intensity of color shift was unique relative to biofluorescence found in other mammals. We establish that biofluorescence in springhare likely originates within the cuticle of the hair fiber and emanates, at least partially, from several fluorescent porphyrins and potentially one unassigned molecule absent from our standard porphyrin mixture. This discovery further supports the hypothesis that biofluorescence may be ecologically important for nocturnal-crepuscular mammals and suggests that it may be more broadly distributed throughout Mammalia than previously thought.


Results
While studying biofluorescence in New World flying squirrels (Glaucomys spp.) and members of Anomaluromorpha at the Field Museum of Natural History (FMNH) in Chicago, Illinois, USA, in April 2018 19 and November 2019 22 , we discovered springhare biofluorescence and subsequently documented the trait in multiple specimens and captive individuals. We examined a total of 14 museum specimens, including eight museum specimens of P. capensis collected from Angola (n = 2) and Botswana (n = 6), and six specimens of P. surdaster collected from Kenya (n = 3) and Tanzania (n = 3) (Fig. 1, Supplementary Table S1). We examined individuals of both sexes, including four males (P. capensis, n = 2; P. surdaster, n = 2) and ten females (P. capensis, n = 6; P. surdaster, n = 4), and specimen collection dates ranged from 1905 to 1963. For museum specimens, we observed and photographed biofluorescence on both the dorsal and ventral side of each specimen following the methods of Anich et al. 22 (Canon EOS 50D, Canon USA Inc., Melville, NY, USA; Sigma 17-70 mm f 2.8-4 DC Macro) under visible light (Canon Speedlite 430EX) and then separately under 395 nm UV light (LED UV flashlight, iLumen8 100 LED). We photographed a subset of specimens using a 470 nm longpass filter (K&F Concept, Guangdong Sheng, China; Tiffen Yellow 2 #8, Hauppauge, New York, USA) under UV illumination to absorb any residual light in the blue wavelengths. We also captured UV reflectance using a Nurugo SmartUV camera (Union community Co., Ltd., Seoul, Republic of Korea), which suggested that very little UV light was being absorbed on either dorsal or ventral surfaces of a subset of springhare specimens. Following the methods of Anich et al. 22 for fluorescence spectroscopy (Ocean Optics USB2000 + , Largo, Florida, USA), we identified two peaks of fluorescence at 500 and 650 nm for a section of highly fluorescent fur on the ventral surface ( Figs. 1 and 2).
We also observed and photographed biofluorescence on five living captive-bred P. capensis individuals (three males and two females) at Omaha's Henry Doorly Zoo and Aquarium in Omaha, Nebraska and one deceased (due to natural causes) individual (female) at the Mesker Park Zoo & Botanic Garden in Evansville, Indiana ( Fig. 3 We photographed a human hair under the same conditions for comparison. We observed fluorescence of individual springhare hair fibers and variation in the presence of fluorescence within individual hair fibers, suggesting that the fluorescence is distributed through the thickness of the cuticle and absent from the core and tips of hair fibers (Fig. 4). Washing the hair or fur with Dawn dish soap (Cincinnati, Ohio, USA) did not remove or diminish fluorescence or result in a transfer of the fluorescence.
We used thin layer chromatography to separate fluorescent extracts from hair samples. Using high performance liquid chromatography (HPLC), we revealed the presence of fluorescent porphyrin species including uroporphyrin-I, uroporphyrin-III, heptacarboxylporphyrin, hexacarboxylporphyrin, and coproporphyrin-I ( Fig. 5; Supplementary Figs S1 and S2). We also detected an unassigned molecule absent from our standard www.nature.com/scientificreports/ mixture of porphyrins that peaked at approximately two minutes, which may also play a role in the biofluorescence observed ( Fig. 5; Supplementary Figs S1 and S2). For P. capensis and P. surdaster, all individuals exhibited orange to red biofluorescence, although we did observe variation in the intensity of biofluorescence across individuals. Biofluorescence was pronounced on both the dorsal and ventral surfaces (Fig. 1), and fluorescence intensity for the dorsal surfaces was often more intense in the head and posterior region (Figs. 1 and 3). Ventrally, biofluorescence occurred predominantly along the inner thigh and tail (Fig. 1). We observed notable patchiness in the biofluorescence in both museum and captive specimens ( Figs. 1 and 3).

Discussion
We have discovered a funky and vivid porphyrin-based biofluorescence in Pedetidate, representing the first welldocumented biofluorescence of an Old World eutherian mammal. While we do not have a large enough sample size to draw conclusions about the frequency of this trait in wild populations, we did consistently observe the trait in six captive individuals, as well as, 14 museum specimens collected at different times over a 58-yr. period and across seven separate locations in four countries (Fig. 3). Due to the spatial, temporal, and contextual (i.e., captive or wild) diversity of our specimens, we suspect this trait is not environmental. Both male and female specimens fluoresced in the same regions and with the same intensity, generally; therefore, we suspect that the trait is not sexually dimorphic. The fact that biofluorescence was not easily removed via washing and was present on museum specimens from 1905 suggests that the biofluorescence is a part of the physical anatomy of the hair fibers for Pedetidae. Biofluorescence appeared more vivid in living individuals than in museum specimens, potentially indicating some degradation over time (Figs. 1 and 3). www.nature.com/scientificreports/ We detected multiple species of porphyrin in extracts from springhare hair samples. Porphyrin-based biofluorescence has been suspected or confirmed in many marine invertebrates 26,27 , the plumage of many bird species 18,[28][29][30] , the bones of at least one species of rodent 31 , and at least one species of Platyhelminthes 9 (Platydemus manokwari). Here we detected uroporphyin-I, -III, heptacarboxylporphyrin, hexacarboxylporphyrin, coproporphyrin-I, and one unassigned molecule absent from our standard porphyrin mixture. The isolated porphyrins are formed by oxidation of porphyrinogens, which are intermediates in the biosynthetic pathway of heme 32 . Uroporphyrin and coproporphyrin fluoresce between 570 and 720 nm in various conditions 33,34 , suggesting that at least both uroporphyrin and coproporphyrin play a role in causing biofluorescence in springhare. We recommend future studies be done to determine whether this biofluorescence is an advantageous evolutionary trait or a disease, such as porphyrias as seen in fox squirrels (Sciurus niger) 31,35,36 , canefield rats (Rattus sordidus) 37 , and humans (Homo sapiens) 32,38 .
Biofluorescence in both species of Pedetes was notably patchy (Figs. 1 and 3), and observations of captive individuals indicated that the areas most impacted by grooming and intra-specific interactions, i.e., reproduction, appear to overlap relatively strongly with areas most consistently exhibiting biofluorescence; this would suggest that the fluorescence might be externally applied to the fur during certain behaviors. However, thorough    Fig S1) which allowed for identification of uroporphyrin-I, uroporphyrin-III, heptacarboxylporphyrin, and coproporphyrin-I in the sample. The peak centered around two minutes has not been assigned and does not correspond to any of the porphyrins in the standard mix (uroporphyrin, heptacarboxylporphyrin, hexacarboxylporphyin, pentacarboxylporphyrin, coproporphyrin, and mesoporphyrin). www.nature.com/scientificreports/ nocturnal-crepuscular [19][20][21][22] and UV-sensitive 7,39 species, and UV-color vision appears to be ecologically important to many nocturnal-crepuscular mammals 1 . While we cannot determine why Pedetidae exhibits biofluorescence, our observations add further support for the hypothesis that biofluorescence and UV wavelengths of light may be ecologically important for nocturnal-crepuscular mammals 1,9,19,22 . Our observations also suggest that biofluorescence may be more broadly distributed throughout Mammalia than previously thought 22 .

Methods
All We selected an intense fluorescent spot on a subset of the Pedetidae specimens and placed a probe holder directly on that spot with the probe at 45° relative to the sample. Fluorescence spectra were taken at five different places within that spot, and these five spectra were averaged. We recorded the light source spectrum against a polytetrafluoroethylene diffuse reflectance standard. . We separated the solution using thin layer chromatography and a mixture of 4 mL DMF, 35 mL MeOH, 6 mL ethylene glycol, 0.4 mL glacial acetic acid, 18 mL 1-chlorobutane, and 20 mL CHCl 3 as the solvent 40 . The thin layer chromatography plates had aluminum backs and were coated with silica gel 60 F 254 . After separation, three distinct pink bands were visible under a handheld UV light. We collected and washed the pink silica with acetone and deionized water until the silica was no longer pink. Then, we filtered and reduced the extracts under pressure until a yellow oil remained which was sent for HPLC analysis. We dissolved extracted materials in 1 mL of 1 M hydrochloric acid, centrifuged and transferred the supernatant to amber colored auto-sampler vials. For the identification of porphyrin carboxylic acids, we used a chromatographic marker kit containing 5 micromole mixture of each octa-, hepta-, hexa-, penta-, tretra-, and di-carboxyl porphyrin acids (Frontier Scientific, Inc., Salt Lake City, UT, USA). Octa-and tetra-carboxyl porphyrin acids are conventionally called as Uropophyrin and Coproporphyrin, respectively. We dissolved the standards mixture of porphyrins in the tube in 10 mL of 1 M hydrochloric acid (High Purity grade from Fisher Scientific, Inc., Salt Lake City, UT, USA), and considered this mix as the HIGH standards mix. We created a ten times dilution of the HIGH standards mix and considered this the LOW standards mix.

Extraction of biofluorescent compounds.
We used a gradient elution program with an injection volume set to 100 µl and total run time 36 min 41,42 (Table 1). Two vials of porphyrin carboxylic acid of the (HIGH and LOW concentrations) standards were also analyzed with each batch of samples. We identified the HPLC peaks in samples by matching retention times (min) of sample peaks with carboxyl porphyrin acid peaks (Supplementary Fig S1). Table 1. Gradient elution program for high-performance liquid chromatography analysis of extracts from hair samples to determine the basis of biofluorescence in springhare (Pedetes capensis). Step