The TRHergic cells were identified using Whole-Mount Fluorescent In Situ Hybridization (FISH) and Immunohistochemistry (IHC). IHC was performed using a custom antibody (Wood, 2020; Cocurullo et al., 2023). The TRHergic cells first arise as one or two cells located right at the connection between the ALA and the apical plate (Figures 2A–H’), where the peculiar shaped cells were observed using SEM (Figure 1G). These cells seem to project axons toward the apical organ plate and show an elongated morphology with IHC (Figures 2C–D’). Their number quickly increases and already at 48 h post fertilization (hpf) it is possible to observe 3 or 4 cells labelled (Figures 2E–H). The localization of these cells is better appreciated by IHC, since during the FISH procedure the larval arms, lacking the support from the skeleton, often collapse (e.g., Figures 2I,M). At 72 hpf, the number of TRHergic cells is stabilized at four, two at the base of the arms, on the dorsal side, and two more external facing the oral side (Figures 2I–P’). Their localization can be fully appreciated in an oral view, like in Figures 2O–P’. A full projection of confocal pictures taken from dorsal or ventral views might suggest that they are localized on the same plane (e.g., Figures 2K–L’). Nonetheless, when only the individual stacks are analyzed it becomes obvious that these groups of cells are not localized on the same focal plane (Supplementary Figure S1). Intriguingly, at 72 hpf, these cells have long projections which colonize the area between the ALA, on the dorsal side of the pluteus, where the apical organ is located. Double FISH and IHC on 48 hpf confirm colocalization of Trh mRNA and TRH peptide in the cell body, while TRH peptide only is stained in the cell projections (Figures 2Q–S’’’).

To follow the fate of the TRHergic cells during sea urchin larva development, IHC was performed on older larval stages. To describe the morphology of late P. lividus developmental stages, we used the detailed developmental atlas produced by Formery et al. (2022).

At late 4-armed pluteus stage, 8 days post fertilization (dpf), the network of TRHergic cells is widely extended, and even some interneurons localized in the apical organ plate appear to be stained (Figures 3A–D). In some cases, TRH positive cells are localized in the ALA. Occasionally, a projection originating from the cell at the corner with the oral ectoderm integrates into the arm region and runs underneath the ciliary band (Figures 3C,D). At the 6-armed pluteus stage, the number of TRHergic cells and the network of connections between them increases (Figures 3E–L), and they colonize the whole apical plate. In particular, while the number of cell bodies observed in single stacks is limited (Supplementary Figure S2), the network of connections between TRHergic cells is extensive and fully colonize the oral hood. In one case, projections from the apical organ extending down the mouth epithelium were also observed (Figures 3I–L and Supplementary Figure S2C). Moreover, from the sides of the apical plate, the projections branch out. A first TRHergic axon runs underneath the ciliary band covering the ALA, goes down the epithelium at the side of the mouth, and innervates also the PO ciliary band, while a second group of axons branches from the apical organ area and innervates the buds of the second pair of ALA that will soon rise, which are called preoral arms or PRA (Figures 3G,H,K,L).

To highlight the ciliary band, larvae at different stages were stained using an anti-acetylated tubulin antibody. At 72 hpf and 5 days post fertilization (dpf), numerous cilia are localized not only in the ciliary band, but also in the area between the ALA and on the oral hood (Figures 4A,E,I and Figure 1E). Double staining showed that the TRHergic cells are ciliated cells (Figures 4A–L, see green arrows and white arrows in Supplementary Figure S3). In plutei at 6-armed stage, the ciliary band extends on all the arms, and the TRHergic axons run underneath (Figures 4M–P). In older 6–armed larvae, two ciliated structures begin to form below the PO arms, these are called epaulettes and are responsible for larval swimming at later stages (Smith et al., 2008). Intriguingly, the TRHergic axons do not run underneath the epaulettes (Figures 4Q–T and Supplementary Figures S4B,C for a close-up). Moreover, no TRHergic axons connect to the rudiment that is forming on the left side of the larva (Supplementary Figures S4D,E).

Since no TRHergic cells were observed in the rudiments, we set out to investigate if the TRH-type neuropeptide is expressed after metamorphosis in P. lividus. Therefore, immunostaining was performed on juveniles a few weeks post metamorphosis, corresponding to stage J5-J6 in Thompson et al. (2021). Sea urchin juveniles at this stage look very much like adult sea urchins on a smaller scale. They have five sets of three tube feet on the oral side, in correspondence with each radial nerve, and five sets of four spines between them. Moreover, they have five peristomial podia located around the mouth, in correspondence with the radial nerves. On the contrary, the aboral surface does not have spines yet at this stage. More spines and tube feet are located on the lateral side of the juvenile. The TRHergic cells appear to be quite widespread. In fact, TRHergic projections were stained on the aboral side of the juvenile (Figures 5A,B). TRHergic axons also innervate the stalk of tube feet (Figures 5C–F). In the tube feet observed from the aboral side, TRH cell bodies can be observed at the base of the disk, and additional projections seem to branch from this, up into the disk (Figures 5C–F). At juvenile stage, the anti-acetylated tubulin stains not only cilia, but also neuronal fibers, therefore the anti-acetylated tubulin antibody was used to highlight juvenile morphology. Another group of TRHergic cells is located underneath the spines, forming a ganglion (Figures 5G–N). Moreover, a net of TRHergic projections extends to the spines (Figures 5G–N). On the oral side, a group of TRHergic neurons are located at the tip of the radial nerve (Figure 6A, see yellow arrows). In addition, two groups of TRH positive cells are located at the tip of the peristomial podia (Figures 6B,M,N). Ganglia containing TRHergic cells are also localized at the base of the tube foot disks (Figures 6C–J). From these ganglia, TRHergic projections branch to the sides of the discs, where cell bodies of elongated bipolar cells are located (Figures 6C–J). Additionally, TRH projections are located at the base of the tube feet (Figures 6K,L). In order to check if the juvenile’s TRHergic cells are neuronal cells, immunostaining against an echinoderm pan neuronal marker (synaptotagmin-B or SynB) was performed using the antibody 1E11 (Burke et al., 2006b). In our juveniles, 1E11 effectively stained the radial nerves, the axons at the base of the spines, and the ganglia located in the peristomial podia, and at the base of the disks in the tube feet, in agreement to what previously described at slightly earlier juvenile stage by Burke et al. (2006a), Formery et al. (2021), Paganos et al. (2022c). Intriguingly, not all the TRHergic cells were found positive to SynB (Figures 7C–H). In particular, the tube feet ganglia showed co-expression only in some cells located around the center. Also, in the peristomial podia, only one of the of the two TRHergic groups of cells looks to be positive to 1E11 (Figures 7I,J). Since 1E11 is commonly considered a pan-neuronal marker, this could suggest that not all the TRHergic cells observed at juvenile stage are neurons or utilizing different synaptotagamins. Indeed, it has been shown by recent studies that SynB is not the only synaptotagamin present in the echinoderm nervous system (Meyer et al., 2023) and that 1E11 is not sufficient to label all the neuros present in the tube foot disk (Paganos et al., 2022c). Nonetheless, further analysis, possibly including scRNA-seq, are necessary to confirm the possibility that non-neuronal population(s) of TRHergic cells also exists.

Furthermore, we set out to investigate to what extent TRHergic cell molecular identity and signaling known in S. purpuratus (Wood, 2020; Paganos et al., 2021; Cocurullo, 2022; Cocurullo et al., 2023) are conserved in P. lividus. In S. purpuratus larvae, TrhR is expressed at the tip of the ALA and PO (Wood, 2020; Cocurullo, 2022). To determine if this expression pattern is conserved in P. lividus larvae, FISH was performed at 48 hpf. This staining confirmed that in both species, TrhR is expressed by the arm epithelium (Figures 8A–D). Double staining with Trh suggests that TRHergic cells do not express TrhR, or, if they do, it is at a very low level (Figures 8C,D). This is especially clear when looking at the individual stacks rather than at the full projections (Supplementary Figure S5A). Double FISH of Trh and FbsL_2 (marker for the ciliary band; Paganos et al., 2021), shows no co-expression (Figures 8E–H and Supplementary Figure S5B), similarly to S. purpuratus (Paganos et al., 2021).

Next, we aimed at comparing the molecular signature of P. lividus’ TRHergic cells with the one described for S. purpuratus larvae (Cocurullo et al., 2023). In S. purpuratus, the TRHergic cells are photoreceptors expressing the Go-Opsin3.2. Despite the fact that we could identify both Opsin3.1 and Opsin3.2 in the P. lividus’ genome (Opsin 3.2 is pliv13652/ PL31706, Opsin 3.1 is Plv13839/PL31732, https://genome-euro.ucsc.edu; for orthology assignment see D’Aniello et al., 2015), no Opsin3.2 expression was detected at early larva stage by PCR and RT-PCR (data not shown). However, we could detect Opsin3.1, which is not expressed in S. purpuratus at early pluteus stage. Opsin3.1 probe gave a diffused signal throughout the larva, but the signal is mainly accumulated in an area between the two ALA. Since data from our previous work (Cocurullo et al., 2023) suggests that Sp-Opsin3.2 could signal through a Gi-mediated phototransduction cascade, we also performed a FISH to detect the expression pattern of Pl-Gi. Intriguingly, although this Gi seems to be broadly expressed, transcripts accumulate in the TRHergic cells and in cells located in the dorsal area between the two ALA, where the apical organ is located (Figures 8M,N). Double Gi/Trh FISH confirmed expression of Gi in the TRHergic cells at 72 hpf (Figures 8O–T). In summary, whether P. lividus TRHergic cells are photoreceptors remains a mystery and further analysis is required.

In S. purpuratus, TRHergic cells represent a distinct cell cluster, which connects to the apical organ (Paganos et al., 2021). Based on the evidence that in P. lividus the first TRHergic cells to appear are located closer to the apical organ than in S. purpuratus, we were interested in determining if there is any co-expression of TRH and TPH/Serotonin at early P. lividus larva stages. Tryptophan hydroxylase, or TPH, is an enzyme involved in the synthesis of the monoamine neurotransmitter serotonin. At 48 hpf, there are about 2–3 Tph huge flask-shaped positive cells located in the apical organ (Figures 9E–H). A number of smaller round shaped Tph positive cells are distributed nearby and between the bigger cells. Of these cells, none was also Trh positive, although some Trh + cells were located in very close proximity to the Tph + ones (Figures 9A–D). At 72 hpf, we used an anti-serotonin antibody to identify the serotonergic cells. IHC of Serotonin and TRH at 72 hpf, stained two serotonergic neurons and four TRHergic cells. The TRH positive ones are located two at the base of the ALA and two more externally, below the ALA (Figures 9I–T). The serotonergic cells were located between the more external TRHergic cell and the one located at the arm connection. TRHergic and serotonergic axons in the apical organ seem to run very close to each other. At 96 hpf, often one or more serotonergic cell bodies are stained deep in the ALA (Figures 9M–P). These findings suggest that the TRHergic cells do not belong to the serotonergic cell type in P. lividus, similarly to what found in S. purpuratus (Paganos et al., 2021; Cocurullo et al., 2023).

Furthermore, we tested if Trh + neurons at early pluteus stage are also cholinergic, like in S. purpuratus (Paganos et al., 2021). To do this, we performed a double FISH for Trh and Chat. Choline acetyltransferase (commonly abbreviated as ChAT) is a key enzyme responsible for the synthesis of the neurotransmitter acetylcholine and therefore its expression is used as marker to identify cholinergic neurons. In P. lividus, at early 48 hpf, no Trh/Chat co-expression was detected (Figures 10A–H). Similarly, we tested if TRHergic neurons are dopaminergic using the expression of DOPA decarboxylase (DDC) as marker.

As previously described, TRHergic cells in S. purpuratus are located at the sides of the serotonergic neurons, at least up to 96 hpf (Figures 11A–B’). Moreover, the number of TRHergic cells appears to increase at a slow rate and at 6 dpf larvae having still two TRHergic neurons can be observed (Figures 11C–D’). Based on these differences observed between TRHergic cells in S. purpuratus and P. lividus, we set out to investigate the localization of TRHergic cells in two other species.

First we chose Arbacia lixula, which is another Mediterranean sea urchin species sharing the same habitat of P. lividus. TRH IHC stained a group of projections and cell bodies located between the ALA, resembling TRH expression in P. lividus after 1 week post fertilization (Figures 11E,F). Serotonergic cells, on the contrary, show a different organization. Most cell bodies could be stained only at the sides of the area between the ALA (Figures 11G–L). Another serotonergic cell is usually located inside the ALA, next to the tip, that sends a projection toward the cells at the base of the arms. Moreover, from the basal group of cells, axons run into the oral hood and toward the PO arms.

The second species we investigated is Heliocidaris tuberculata, a species of the Pacific Ocean, like S. purpuratus, but since it comes from Australia, larvae from this species grow at a higher temperature, more similar to the Mediterranean species. In comparison, Heliocidaris plutei have long arms, like P. lividus, but a more compact and large body, similar to S. purpuratus. In H. tuberculata, our TRH antibody stained about two cells located at each side of the apical plate (Figures 11M–R), which resemble the topology of the TRHergic cells in S. purpuratus plutei.

Similarly to S. purpuratus (Wood et al., 2018; Cocurullo et al., 2023), TRHergic cells in P. lividus first appear at early pluteus stage as one or two cells bilaterally located above the mouth (Figure 2). However, while in S. purpuratus these first two cells are located in the oral ectoderm just above the mouth, in P. lividus they are more dorsal and located right at the corner of the ALA. Very quickly and in a non-stereotypic way, the number of TRHergic cells increases and most larvae at 72 hpf already possess four of them. From this point on, also the net of connections between the cells increases and at late 6-armed larval stage they colonize the oral hood (Figures 3, 4). At this stage, projections from the oral hood extend underneath the ciliary band of all six arms and of the buds of the additional two ALA forming in front of the primary ones. No TRHergic neurons or axons were stained in the rudiment, but juveniles showed an extended network of TRHergic cells and projections (Figures 5–7), raising questions on when the first TRHergic cells appears in the rudiment/juvenile. To answer this question, a more detailed analysis of larvae at 8-armed stage and juveniles during and right after metamorphosis is needed.

Based on the differences observed in terms of topology and number of TRHergic cells at comparable developmental stages in S. purpuratus and P. lividus, we decided to further investigate the molecular identity of the TRHergic cells in P. lividus.

First we investigated the expression pattern of the TRHR. Similarly to what was observed for S. purpuratus (Cocurullo, 2022), the receptor at early larva stages is expressed on the epithelium which covers the tip of ALA and PO, suggesting that the function of the TRH neuropeptide is conserved between P. lividus and S. purpuratus (Figure 8). Here, we also set out to compare the molecular fingerprint of the TRHergic cells of P. lividus and S. purpuratus. Single cell transcriptomics of S. purpuratus larvae at early pluteus stage (72 hpf) identified the TRHergic cells as an independent neuronal subcluster (called distal apical plate neurons) (Paganos et al., 2021). In fact, these cells, despite their proximity to the ciliary band and the serotonergic neurons of the apical organ, did not express genes such as Sp-FbsL_2 and Sp-Tph, which are markers for the ciliary band and serotonergic neuron clusters, respectively. Intriguingly, also P. lividus plutei at 48 hpf (which is the corresponding early pluteus stage) do not co-expressed FbsL_2 or Tph (Figures 8, 9), suggesting that the TRHergic cells could form an independent neuronal cluster in P. lividus too. Moreover, the relative position of TRHergic and serotonergic neurons seems to be slightly different between the two species. While in S. purpuratus, serotonergic neurons are located in between the two lateral clusters of TRHergic cells (Figures 11A,B; Wood et al., 2018), in P. lividus, the serotonergic cells look to be located more at the sides of the oral hood, and from later stages on (72–96 hpf), some cells are also localized in the ALA (Figure 9). Nonetheless, up to 96 hpf, TRH and Tph/Serotonin do not co-express in the same cells (Figures 11A,B), although their axons and cell bodies appear to be in very close proximity in the central area between the ALA. Furthermore, we investigated if the TRHergic cells in P. lividus are also secreting neurotransmitters, in particular acetylcholine and dopamine. For this, we performed a double FISH of Trh and Chat (key enzyme for acetylcholine synthesis) and Ddc (enzyme for dopamine synthesis). In both cases, no co-expression with TRH was observed, although some dopaminergic neurons appear to be located in very close proximity to the TRHergic cells (Figure 10).

An interesting characteristic of the TRHergic neurons of the S. purpuratus larva is their dual sensory/neurosecretory role as they also express the Go-Opsin3.2 and deploy an ancient regulatory module responsible for photoreceptor specification (Cocurullo et al., 2023). Therefore, we investigated if the TRHergic cells in P. lividus are also photoreceptors. Interestingly, it was impossible for us to detect the expression pattern for the Opsin3.2, both by immunostaining or FISH (data not shown). Nonetheless, S. purpuratus genome encodes a second Go-Opsin named Opsin3.1, which is not expressed at early pluteus stage. Also P. lividus genome encodes an orthologous Opsin3.1. Our data suggest that Opsin3.1 is expressed at the pluteus stage, but the expression pattern is different from the Sp-Opsin3.2 (Figure 8). In fact, Opsin3.1 transcripts are found in a few cells located between the two ALA in P. lividus, while the co-expression with Trh is not yet clear. Indeed, single-cell transcriptomic is a powerful technique that will allow us to answer this question in the near future. Nonetheless, this finding shows a major difference between S. purpuratus and P. lividus larvae since it means that these two species deploy a different Go-Opsin at the pluteus stage, and most likely with different functions. Also, the expression pattern of these Go-Opsins is different, suggesting a fascinating evolutionary history for Go-Opsins in sea urchins. We could hypothesize that after gene duplication, Go-Opsins were differentially deployed. Nonetheless, the Opsin3.1 expression profile in P. lividus represents the first evidence for this process, since a Go-Opsin of another sea urchin species (Hemicentrotus pulcherrimus), closely related to S. purpuratus, appears to have the same expression pattern as Opsin3.2 in S. purpuratus (Yaguchi and Yaguchi, 2021).

Nonetheless, TRHergic cells in P. lividus could be sensory as they are ciliated. To further investigate if these cells could be photoreceptors, we also tested the presence in these cells of transcripts derived from homologues of retinal genes found to be expressed by the Sp-Go-Opsin3.2+ cells of 72 hpf and 5 dpf S. purpuratus larvae (Paganos et al., 2021; Cocurullo et al., 2023). In S. purpuratus, the phototransduction cascade activated by Opsin3.2 might be mediated by the Gi protein; therefore, we investigated the expression of Gi in the 4-armed P. lividus larva. Intriguingly, in these larvae, Gi transcripts are accumulated in cells located between the ALA, in a similar position to Opsin3.1 and Tph positive cells, but also in the TRHergic cells (Figure 8).

Finally, we performed a survey to assess the localization of TRHergic cells also in other sea urchin species, namely A. lixula, H. tuberculata, and compared their topology with that observed in P. lividus and S. purpuratus. We also attempted to reconstruct the relative localization in relation to the serotonergic neurons in these different larvae. We summarized these results in the illustration shown in Figure 12, with the integration of a phylogenetic tree showing evolutionary relations of these four sea urchin species. It is important to note that data regarding serotonergic neuron organization in A. lixula, H. tuberculata, and P. lividus are extremely limited or absent, and need to be expanded (Bisgrove and Raff, 1989; Byrne et al., 2007; Rendell-Bhatti et al., 2021; Tsironis et al., 2021). Nonetheless, our comparison suggests that the relative organization of TRHergic and serotonergic cells slightly changes in different sea urchin species, being the two types of neurons more intermingled in P. lividus and A. lixula as compared to the more clustered distribution of these cells in S. purpuratus and H. tuberculata at the early larva stage. The mechanistic meaning of such subtle differences need to be further investigated. On the other hand, the comparison of different species would provide new evidence to understand how the TRHergic/photoreceptor cells evolved.

Paracentrotus lividus gametes were obtained by the Stazione Zoologica Anton Dohrn animal models facility. Spawning was induced by intracoelomic injection of 0.5 M KCl. The sperm was collected dry using a Pasteur pipette and stored at 4°C until usage. To collect eggs, females were inverted over a beaker filled with Mediterranean Sea seawater (FMSW). About 20 mL of eggs were fertilized, adding a few drops of sperm diluted 1:10000. Embryos were transferred in FMSW and reared at 19°C under a 12 h light/12 h dark cycle. To grow larvae until metamorphosis, specimens were reared in 2 or 4 L glass beakers and fed three times per week with a mix of 70% of D. tertiolecta and 30% Isochrysis galbana. The amount of algae was increased from 5,000 cells/ml to 80,000–10,000 cells/ml, according to the needs of the cultures as described in Zupo et al. (2018). Cultures were cleaned two times per week removing most of the sea water and replacing it with fresh MFSW. After metamorphosis, juveniles were fed with seaweed.

Adult S. purpuratus were obtained from Patrick Leahy (Kerckhoff Marine Laboratory, California Institute of Technology, Pasadena, CA, USA) and maintained in circulating seawater at Stazione Zoologica Anton Dohrn in Naples. Gamete release was induced by vigorously shaking the adults. Gametecollection, fertilization and culture maintenance were performed as described for P. lividus but using Mediterranean Sea water diluted 9:1 to adapt to the different salinity of the Pacific Ocean.

A. lixula gametes were obtained as for P. lividus and the larvae reared in the same conditions, however, they were not maintained to metamorphosis, but they were collected at 4-armed pluteus stage.

Heliocidaris erytrogramma samples, fixed at 72 hpf for in situ were kindly gifted by Prof Maria Byrne, University of Sydney.

Specimens for immunostaining (IHC) were fixed in 4% PFA made in MFSW at Room Temperature (RT). The incubation time was adapted to the size of the specimens. For pluteus stages, 10–15 min were used, while at later pluteus stages the incubation time was increased up to a 20 min. After the incubation, the samples were thoroughly washed with PBS and stored at 4°C. Juveniles were fixed for 15 min in 4% PFA, washed once with PBS to remove the fixative and then incubated in cold methanol for 15 min on ice. Finally, they were washed 3–4 times with PBS and stored at 4°C. Specimens for in situ hybridization were fixed in Fixative I (4% PFA in 0.1 M MOPS and 0.5 M NaCl) for at least one night at 4°C. Subsequently, samples were washed 3 times with MOPS buffer (0.1 M MOPS, 0.5 M NaCl, 0.01% Tween 20) for 15 min at RT, dehydrated in 70% ethanol and finally stored at-20°C until usage.

Whole mount RNA Fluorescent in situ hybridization (FISH) and combined FISH-IHC were performed as described in details in Perillo et al. (2021) and Paganos et al. (2022a). Summarizing, antisense probes were transcribed from linearized DNA and labelled during transcription using Digoxigenin (Roche) or Fluorescein (Roche) labelled ribonucleotides following the manufacturer’s instructions. Pl-Trh sequence was obtained by blasting the Sp-Trh sequence against the Trinity transcriptome, the full sequence is included in the Supplementary materials (TRH CDS sequence in P. lividus) together with the translated protein sequence (TRH peptide sequence in P. lividus) and the TRH precursor protein sequences from P. lividus and S. purpuratus aligned and annotated (Aligned TRH precursor peptide sequences in P. lividus and S. puprupartus). Primer sequences used for cDNA isolation and probes synthesis and the type of ribonucleotides used are included in Supplementary Table S1. Fluorescent signal was developed via using fluorophore conjugated tyramide technology (Perkin Elmer, Cat. #NEL752001KT). For combined FISH-IHC, after tyramide amplification step, samples were incubated in blocking (containing 1 mg/ml Bovine Serum Albumin and 4% Sheep Serum in PBS) for 1 h at RT, then transferred in primary antibody diluted 1:400 in blocking O/N at 4°C. The list and details of the primary antibodies used can be found in the Supplementary Table S2 Samples were washed 4–6 times with PBS 1x, then stained with appropriate Alexa Fluor secondary antibodies diluted 1:1000 in blocking, and finally washed 4–6 times with PBS 1x. DAPI (10 mg/mL stock) was added to the samples at a final dilution of 1:10000 to stain nuclei. Specimens were imaged using a Zeiss LSM 700 confocal microscope and pictures analyzed using ImageJ 1.53v.

Paracentrotus lividus larvae at 72 hpf were collected and processed as described in Paganos et al. (2023). In summary, whole plutei were fixed in 2.5% glutaraldehyde in MFSW at 4°C overnight. The day after, samples were washed several times with MFSW and infiltrated with DMSO by subsequent incubations in 25% DMSO in double distilled water (ddH₂0) for 1 h at RT followed by a second incubation in 50% DMSO in ddH₂0 for 1 h at RT. During the last 30 min of incubation a metal key, a pair of forceps, an aluminum tray and a razor blade were placed in a polyurethane ice bucket and covered with liquid nitrogen to pre-cool. Approximately 100 μl of specimens were quickly transferred inside the groove of a pre-chilled metal key and spread throughout the groove surface. The key was quickly placed inside a pre-chilled aluminum tray and, using a hammer, a razor blade placed on the key’s groove was stricken multiple times until complete fracturing of the pellet. The fractured sample was then collected and transferred in a clean Eppendorf tube. DMSO solution was removed, and specimens were dehydrated in a graded ethanol series. Samples were then subjected to the standard processing for SEM including critical point drying and sputtering and were analyzed using a Jeol field emission scanning electron microscope (JSM 6700-F).