To evaluate the quality of velvet antler samples, microscopic observation, DL measurements and chemical analysis were used to assess the LP, FP, XP and GP slices obtained from Cervus nippon Temminck and Cervus elaphus Linnaeu, respectively. We first focused on the experimental results of Cervus nippon Temminck. Figure 2 illustrates the images of LP, FP, XP and GP powder of Cervus nippon Temminck., and the selected slides of hair, bone fragments and non-ossified bone tissue. It was observed that the hair of LP samples was relatively smooth, and the bone density of LP was relatively higher from the image of bone fragments. Compared to LP, the hair of FP, XP and GP was relatively rough (FP), less (XP) and thin (GP), respectively; the bone fragments of FP, XP and LP possessed more bone lacunas; and the non-ossified bone tissue of FP, XP and LP demonstrated more irregular massive protuberances. Next, the LP, FP, XP and GP samples was measured by using DL. Figure 3a illustrates the significantly different DL decay curves of LP, FP, XP and GP powder. To extract the four parameters, a hyperbolic function was used to fit the DL decay curves. To analyze further the difference between DL parameters, a two-tailed, unpaired Student's t-test was used to compare the four DL parameters among the LP, FP, XP and GP samples. The results showed that the parameter I0 was significantly different in all velvet antler samples; T, Beta and Tau differed significantly among LP, FP and XP samples, but these parameters were not able to indicate the significant difference between XP and GP powder. To further evaluate the DL data, PCA was applied to the DL data for visualizing the variations among d LP, FP, XP and GP samples. The results illustrated that there were significant classifications among those samples (Fig. 3c). Many research supports that main prominent bioactive components of velvet antler are proteins and polypeptides [21]. Therefore, we tested the contents of total protein and 16 amino acids to assess the quality of LP, FP, XP and GP samples. The results showed that the total protein contents decreased gradually and significantly from LP to GP (Fig. 3d). Figure 3e is a heat-map of the tested amino acids in LP, FP, XP and GP samples. The contents of most amino acids in LP samples were significantly high, and the contents of most amino acids in GP samples were significantly low. LP, FP, XP and GP samples contained specific amino acids with characteristics in contents, respectively. For instance, the contents of Arg and Leu were highest in LP; the contents of Pro and Val were relatively lowest in FP; Tyr, Ala showed highest contents in XP; and Pro and Lys showed relatively highest and lowest contents, respectively, in GP. Next, we pooled DL and amino acids data of LP, FP, XP and GP samples together to establish a network in order to indicate the correlation between DL and specific chemicals in the velvet antler slices of Cervus nippon Temminck. The DL parameters possessed positive and strong correlation with Phe, Met, Glu, Asp and Arg (Fig. 3f).
Next, we focused on the experimental results of Cervus elaphus Linnaeu. In the results of microscopic observation, we found that the LP powder possessed more hair fragments compared to the FP, XP and GP powder. The hair fragments of FP powder were rough and thick, and the hair fragments of XP and GP powder were relatively thin (Fig. 4). It is worth noting that the morphology of bone fragments and non-ossified bone tissue of FP, XP and GP powder were different compared that of Cervus nippon Temminck. There were no significant bone lacunas and irregular massive protuberances, but darker colors on bone fragments and non-ossified bone tissue of XP powder and more fragments of bone and non-ossified bone tissue of GP powder (Fig. 4). In DL measurements results, the decay curves of LP, FP, XP and GP slices demonstrated differences, but the tails of curves overlapped together (Fig. 5a). The parameter I0 was able to distinguish LP, FP, XP and GP powder, but the other DL parameters were not able to indicate the significantly difference among them completely (Fig. 5b). In the PCA analysis, there was no clear separation between FP and LP samples, but XP and GP samples showed significantly different classification compared with LP and FP samples (Fig. 5c). The results of total protein contents also showed a gradual and significant decrease from LP to GP slices (Fig. 5d). LP samples demonstrated the most amino acids with high contents, however, GP samples also contained a lot of high contents amino acids compared to FP and XP slices (Fig. 5e). In additional, some characteristic amino acids were found such as Ser, Leu and Arg (LP); Gly and His (FP); Ala and Lys (XP); and Pro and Val (GP) (Fig. 5e). Moreover, the correlation network showed that there were nine positive and strong correlations between amino acids and DL data (Fig. 5d).
Next, the velvet antler samples of Rangifer tarandus were examined in order to test whether microscopic observation, DL measurements and chemical analysis could distinguish the samples which are not officially included in Chinese Pharmacopoeia. This is because of a lot of non-official velvet antler products exist on the market. Here, FP and XP samples obtained from Rangifer tarandus were used for this study. In the results of microscopic observation (Fig. 6), the hair fragments of FP and XP samples showed no significant difference compared to that of Cervus nippon Temminck and Cervus elaphus Linnaeu. The FP powder of Rangifer tarandus had no significant bone lacunas and irregular massive protuberances, but more non-ossified bone tissues. The XP powder of Rangifer tarandus had irregular massive protuberances on non-ossified bone tissue. Next, the DL decay curves of FP slices showed significant different photon emission kinetics among the three species of deer (Fig. 7a). Subsequently, all the DL parameters demonstrated significant difference of FP slices among those three species of deer (Fig. 7b). However, the DL decay kinetics of XP slices showed relatively minor difference (Fig. 7c), and no single DL parameter of XP slices was able to separate the three species of deer completely (Fig. 7d). In the analysis of chemical data, there was no significant difference on total protein contents of FP samples between Cervus nippon Temminck and Cervus elaphus Linnaeu, but the FP samples of Rangifer tarandus showed significantly lower contents of total protein compared to that of Cervus nippon Temminck and Cervus elaphus Linnaeu (Fig. 7e). In XP samples, there was no significant difference of total protein contents among those three species of deer (Fig. 7e). In addition, the results of heat-map illustrated that the contents of specific amino acids were significant difference on FP and XP samples, respectively, among the three species of deer (Fig. 7f and 7 g).