We aim to attain the sustainable use of longgu and have investigated the significance of longgu in Keishikaryukotsuboreito (KRB) decoction. We have already reported that longgu alters compound profiles in KRB decoction and hypothesized that it does so by adsorbing foreign organic compounds into its superficial pores. In the present study, we focused on the adsorbability of organic materials onto longgu surface as the cause of component profile alteration. We analyzed the physical changes in longgu through the decoction process by measuring the adsorbed water on longgu surface. ¹H magic angle spinning NMR (¹H-MASNMR) spectroscopic analysis revealed that raw longgu (R-raw) as well as decocted longgu [whether single (R-r) or KRB (R-krb) decoction] adsorbed water. However, the amount of adsorbed water in R-krb was smaller than that in R-raw and R-r. The nitrogen adsorption isotherms of longgu samples indicated that longgu was macroporous. The Brunauer–Emmett–Teller (BET) surface area of R-krb was smaller than that of R-raw and R-r. Further, thermogravimetric analysis of longgu samples showed that R-krb adsorbed matter that R-raw and R-r did not adsorb. The above findings and the ¹H-MASNMR analysis of heated longgu samples suggested that longgu adsorbed organic compounds into the pores. We considered that longgu adsorbed organic compounds during KRB decoction into its superficial pores through the decoction process.

Along with economic development, domestic demand for crude drugs has been increasing in People’s Republic of China (PRC). However, there are some problems in crude drug supply such as price rise and export restrictions because of the depletion of natural sources. In Japanese market, around 80% of crude drugs are from PRC.¹⁾ The cultivation of botanical crude drug sources has been attempted as a countermeasure against resource exhaustion. However, the production of mineral and fossil crude drug sources has seldom been studied.

In Japan, mammalian fossil bones are used as a sedative drug named ‘longgu’ (‘Ryu-kotsu’ in Japanese; ‘Fossilia Ossis Mastodi’ in Latin), and are listed in the Japanese Pharmacopoeia.²⁾ The longgu traded in Japan is imported from PRC, where longgu resources are facing the threat of depletion. In previous studies, orally administrated longgu powder and 80% methanol extract have shown sedative effects in rats and mice.^3,4) However, longgu is a component of Kampo formulas and their decoctions or extracts are generally used in clinical practice. Therefore, patients rarely take longgu powder and methanol extracts of longgu. It is necessary to reveal the role of longgu in Kampo preparation and to develop a countermeasure against the resource depletion.

We have investigated the role of longgu in Keishikaryukotsuboreito (KRB). In our previous study, we showed that the profiles of organic and inorganic components of KRB and longgu-free KRB (KB) decoctions were different. Scanning electron microscopy/energy dispersive X-ray spectrometry (SEM/EDX) results have shown that organic materials exist in the superficial pores of longgu decocted as KRB.⁵⁾ We hypothesized that longgu adsorbed foreign organic compounds in KRB decoction on its superficial pores. However, it is impossible to observe the pore interiors directly by surface analysis methods such as EDX. In this study, we aimed to reveal our hypothesis indirectly by three different analyses; ¹H magic angle spinning NMR (¹H-MASNMR) spectroscopic analysis, thermogravimetric analysis (TG) and nitrogen adsorption isotherm. Longgu contains hydroxyapatite (HA) that adsorbs water.^5–7) If longgu adsorbs organic compounds, the pores will be filled and the reduction in adsorbed water will be detected by ¹H-NMR spectroscopy. We used ¹H-MASNMR to obtain highly resolved NMR spectra. TG was performed to confirm the behavior of adsorbed water. We also estimated the pore size of longgu by nitrogen adsorption isotherm and investigated the changes in the status of the pores after decoction by Brunauer–Emmett–Teller (BET) surface area mesurements. Through those results, we observed the transition in longgu surface with decoction process of KRB.

All crude drugs used in this study were purchased from Tochimoto Tenkaido Co., Ltd. (Osaka, Japan): cinnamon bark (Cinnamomi Cortex, lot: 002810010), peony root (Paeoniae Radix, lot: 005310009), ginger (Zingiberis Rhizoma, lot: 005810001), jujube (Ziziphi Fructus, lot: 007110012), glycyrrhiza (Glycyrrhizae Radix, lot: 002010008), oyster shell (Ostreae Testa, lot: 010210010), and longgu (Fossilia Ossis Mastodi, lot: 010805006, 010810002, 010810001). These drugs were cut or crushed according to the criteria of the Japanese Pharmacopoeia, 17th edition.^2,8) Each longgu sample was named R1 (lot: 010805006), R2 (010810002), and R3 (010810001), respectively. Longgu were crushed into smaller pieces between 0.5 and 1 mm, and the longgu particles were selected using sieves with 0.5 and 1 mm meshes.

Longgu were decocted as KRB and single decoction (R) according to the reference⁵⁾: 3 g of longgu was decocted with 400 mL of distilled water (and other crude drugs in KRB, Table 1) for 60 min using a Kampo decocting machine (EK-SA10, Tochimoto Tenkaido Co., Ltd.). Longgu in the residues were collected and dried at 60°C for 12 h for subsequent analysis. Each longgu sample is named as follows: R-raw (non-decocted longgu), R-r (decocted as R), and R-krb (decocted as KRB).⁵⁾ We prepared these sample sets for 3 different lots of longgu (R1, R2, R3) and performed experiments on each sets.

Longgu samples were powdered to pass a sieve with 0.15 mm mesh. The powdered samples were packed into a 4-mm zirconia MAS rotor. The ¹H-MASNMR experiment was performed using DSX-200 and ASX-200 spectrometers (Bruker Analytik Gmbh, Germany) with a superconducting magnet (4.7 T) operated at the Larmor frequency of 200.13 MHz. The samples were spun at 8 kHz, and the free induction decay (FID) signals were gained using a 4-µs single pulse rf excitation with 4 s recycling time and 256 scans for the ¹H spectra of longgu samples. All the measurements were carried out at ambient temperature. The spectra were fitted using the dmfit program.⁹⁾

Powdered longgu samples (10 mg) that pass a sieve with 0.075 mm mesh were examined using a TG-DTA2000SA analyzer (Bruker AXS K.K., Kanagawa, Japan). Each sample was heated in a flow of nitrogen (gas flow rate: 200 cm³/min) at a heating rate of 5°C/min from ambient temperature to 500°C.

Nitrogen adsorption isotherm was measured using a Gemini 2375 v5.00 system (Micromeritics Inc., GA, U.S.A.) at 77 K. The longgu samples were powdered to pass a sieve with 0.15 mm mesh, and the BET method was used to obtain the specific surface areas (BET surface area) of the longgu samples.¹⁰⁾ The BET surface areas were calculated from the nitrogen isotherm data in 0.04–0.31 relative pressure range.

Significant differences among the sample types were analyzed by one-way ANOVA followed by Tukey’s honest significant difference (HSD) test using SPSS Statistics 21 (IBM Japan Ltd., Tokyo, Japan).

The ¹H-MASNMR spectra are shown in Figs. 1a–c. All spectra show signals at 5 ppm and 0 ppm that can be assigned to water and hydroxide ion (OH⁻) of HA, respectively. This assignment is in agreement with the previous studies on bone, HA, and ivory.^11–14) The spectrum of R-krb exhibited a smaller water peak than the other samples did, whereas OH⁻ peak were equivalent among all groups. The result of fitting simulation shows that each peak is consisted of two peaks; two Lorentzian peaks (Peaks 1 and 2) for water and two Gaussian peaks (Peaks 3 and 4) for OH⁻. Peaks 2 and 4 are broader than peaks 1 and 3, respectively. Peaks 2, 3, and 4 have no significant difference between their widths, heights, and areas among all the sample types. However, peak 1 of R-krb has significantly smaller height and area than those of peak 1 of R-raw and R-r (Table 2).

Fig. 1. ¹H-MASNMR Spectra of Longgu Samples and Their Fitting Simulated Peaks

(a) R-raw, (b) R-r, and (c) R-krb. The spinning side bands are indicated by asterisks (*). The spectra of longgu have two peaks at 5 and 0 ppm that are derived from H₂O and OH⁻, respectively (bold black line). Each peak is fitting simulated with two peaks: Lorentzian peaks 1 (black line) and 2 (dashed black line) for H₂O around 5 ppm, and Gaussian peaks 3 (dark gray line) and 4 (dashed gray line) for OH⁻ around 0 ppm.

The thermogravimetric (TG) curves of the longgu samples are illustrated in Fig. 2. R-raw and R-r demonstrated similar TG curves with two inflection points at 100 and 350°C. The TG curve of R-krb also exhibited inflection points at 100 and 350°C. However, it showed an additional inflection point at 250°C, and its slope became steeper after this point. The NMR spectra of the heated longgu samples are shown in Figs. 3a–c. The spectra have two peaks at 5 and 0 ppm as shown in Figs. 1a–c. The water peak (5 ppm) reduced upon heating whereas OH⁻ peak was constant. The result of fitting simulation is similar to former spectra; two Lorentzian peaks (Peaks 1 and 2) for water and two Gaussian peaks (Peaks 3 and 4) for OH⁻. Transitions of peak areas according to the heating are shown as relative value against the peak 4 at room temperature (Figs. 3d–f). The water peaks in the spectra of all the samples reduced upon heating, particularly peak 1. The peak 2 remained for temperatures up to 350°C.

Fig. 2. Typical Thermogravimetric (TG) Curves

(a) R-raw, (b) R-r, and (c) R-krb. The dashed arrows indicate the inflection points in the TG curves.

Fig. 3. Transitions upon Heating in ¹H-MASNMR Spectra and in Peak Areas of the Fitting Simulated Peaks

(a, d) R-raw, (b, e) R-r, and (c, f) R-krb. The spectrum of each sample (a–c) was measured at room temperature (25°C, black line), 100°C (dashed black line), 250°C (gray line), 350°C (dashed gray line). The spinning side bands are indicated by asterisks (*). Each peak was fitting simulated into peak 1–4 as shown in Figs. 1a–c, and the transitions in peak areas according to the temperature rise are shown as relative value based on the peak 4 at room temperature (d–f): each peak is indicated as mean ± S.D. (n = 3) with circle (peak 1), square (peak 2), triangle (peak 3), and cross (peak 4).

The typical nitrogen adsorption isotherm of longgu is shown in Fig. 4a; it is initially concave shaped in the lower relative pressure range (P/Po < 0.3), followed by a linear region, and is finally convex shaped in the higher relative pressure range (P/Po > 0.7). This type of isotherm is classified as type II according to IUPAC, and indicates that longgu is a macroporous (pore diameter >50 nm) material.¹⁵⁾ The BET surface area of R-krb samples (33.2 cm³/g) is significantly smaller than that of R-raw and R-r (39.9 and 38.2 cm³/g, respectively, Fig. 4b).

Fig. 4. Nitrogen Adsorption Isotherm Measurement Results

(a) Nitrogen adsorption isotherm of R-raw sample. (b) Specific surface area (BET surface area) of each sample type (mean ± S.D.). The asterisks (*) indicate significant difference (p < 0.05, n = 5–10, Tukey’s HSD test).

The peak for water of R-krb of ¹H-MASNMR spectroscopic analysis was significantly smaller in height and area than those of R-raw and R-r. This shows that R-krb adsorbed little water compared to the other samples and it suggests that adsorption of water was prevented by some adsorbates on longgu after decoction.

Next, we investigated the presence of organic adsorbates indicated by previous SEM/EDX and above-mentioned NMR analyses.⁵⁾ In TG curve analysis, we found the inflection point at 250°C in R-krb, which could not be found in R-raw and R-r. R-krb may have adsorbed materials that R-raw and R-r did not adsorb. In order to assign the adsorbent, longgu samples were heated at 100, 250, and 350°C for 30 min and subsequently analyzed by NMR spectroscopy (Figs. 3a–c). Wilson et al. reported the presence of water peak in the ¹H-NMR spectrum of a heated bone up to 225°C and ascribed it to tightly bound structural water.¹¹⁾ According to Votyakov et al., fossil bones lose adsorbed water at 25–230°C and structural water at 230–400°C.¹⁶⁾ Moreover, the proton bonded to other nuclides generates a broader NMR peak than the non-bonded proton does. Therefore, we assigned the peaks of longgu spectra as follows: adsorbed water (peak 1), structural water in lattice (peak 2), structural OH⁻ in HA (peak 3), and H-bonded OH⁻ in HA (peak 4). These results indicate that the mass loss (Fig. 2) of longgu mainly occurred due to the loss of both adsorbed and structural water. The transition upon heating in R-krb NMR spectrum is the same with those in R-raw and R-r. Therefore, we consider that the inflection point at 250°C in the TG curve characteristically seen in R-krb resulted from the desorption of adsorbents.

The nitrogen adsorption isotherm and BET surface area measurements of longgu further evidenced the adsorption by longgu. If longgu adsorbs compounds in KRB, its BET surface area will reduce. Our results suggest that longgu is a macroporous (pore diameter >50 nm) material. This is in agreement with our previous SEM observation.⁵⁾ The reduction in BET surface area of R-krb indicates that the macropores on the longgu surface are filled by adsorbates.

Although further study is needed, we consider the adsorbates organic compounds in KRB decoction for three reasons: (1) the white surface of longgu dyed brown, which is the color of the decoction, by the decoction process (data is not shown); (2) the stain was charred by heating; and (3) our previous SEM/EDX study indicated that longgu adsorbs organic components in KRB.⁵⁾ Bone char, which is one of adsorbents composed of bone, has a similar BET surface area (about 40–50 m²/g) as that of longgu.^17,18) Therefore, it is valid to consider longgu an adsorbent.

We have investigated the role of longgu in KRB. Our previous and present results suggest that longgu adsorbs organic compounds into its pores during the decoction process of KRB as illustrated in Fig. 5. Namely, there are numerous macropores on the surface of longgu, and they are filled with water from the air initially. But when longgu decocted as Kampo prescription, that water in the pores are pushed out and replaced with organic compounds derived from other crude drugs. We suppose that this is the main role of longgu in the prescription which results the adjustment of the components in the decoction.

Fig. 5. Schematic Illustration of the Role of Longgu in the KRB Decoction Process

Longgu adsorbs organic materials from other crude drugs into its pores during the decoction process. The photograph of Kampo decocting machine is provided by Tochimoto Tenkaido Co., Ltd.

We designed two approaches against the depletion of longgu resources: the development of (1) substitutes and (2) reutilization methods. However, there is a problem with approach (1): Japanese medical doctors using Kampo preparations will avoid using substitutes made from artificial substances such as ceramics. Thus, the substances should be made from natural resources, i.e., mammalian bones. Therefore, approach (1) requires clarification and imitation of the fossilization process (e.g., decomposition and diagenesis of carcass), and subsequent clinical trial of the developed substitute. Although approach (2) also requires further investigations (e.g., development of appropriate methods to remove adsorbates, evaluation of reusability and durability, and proof of equivalency to original longgu), it seems to be more practical than approach (1) that involves the use of a different material. Our results suggest the role of longgu in Kampo decoction and provide a clue for seeking both an alternative material and a reutilization method. However, further study for both the approaches is required to prepare for the near-future shortage of fossil medicinal materials.

We observed the transition in longgu surface with decoction process of KRB. Longgu naturally adsorbs water on its surface, and the adsorbed water is suggested to be expelled by compounds in Kampo prescription by ¹H-MASNMR spectroscopy and TG analysis. This is in agreement with our previous SEM/EDX results and the present nitrogen isotherm measurement.⁵⁾ Although further study of the adsorbed compounds and their correlation with the sedative effect of longgu is required, we suggest that the significance of longgu is in adjusting the component profiles in Kampo decoctions as an adsorbent.

We are grateful to Dr. Kayoko Shimada-Takaura (the Museum of Osaka University, and Graduate School of Pharmaceutical Sciences, Osaka University) for her advice. We thank Dr. Keisuke Miyakubo (the Museum of Osaka University) for his help in NMR experiments. We sincerely thank to the late Mr. Fumio Tochimoto (Tochimoto Tenkaido Co., Ltd.) for his advice on KRB prescription. This work was supported by Grant-in-Aid for Scientific Research (B) (No. 25282071) and Grant-in-Aid for Scientific Research (A) (No. 17H00832) from the Japan Society for the Promotion of Science (JSPS).

The authors declare no conflict of interest.

• 1) Japan Kampo Manufacturers Association, “The report of the research for the consumption of crude drugs (4).”: ‹http://www.nikkankyo.org/serv/pdf/shiyouryou-chousa04.pdf›, cited 16 March 2017.
• 2) Ministerial Notification No. 64, “The Japanese Pharmacopoeia, 17th edn.,” Ministry of Health, Labour and Welfare, Tokyo, 2016.
• 3) Tsuda T., Sugaya A., Kaneko E., Ohguchi H., Katoh K., Jin W., Sugaya E., Nat. Med., 52, 300–309 (1998).
• 4) Ha J. H., Lee M. G., Chang S. M., Lee J. T., Biol. Pharm. Bull., 29, 1414–1417 (2006).
• 5) Oguri K., Kawase M., Harada K., Shimada-Takaura K., Takahashi T., Takahashi K., J. Nat. Med., 70, 483–491 (2016).
• 6) Yesinowski J. P., Eckert H., J. Am. Chem. Soc., 109, 6274–6282 (1987).
• 7) Cho G., Wu Y., Ackerman J. L., Science, 300, 1123–1127 (2003).
• 8) Ministerial Notification No. 348 “Supplement I to the Japanese Pharmacopoeia, 17th edn.,” Ministry of Health, Labour and Welfare, Tokyo, 2017.
• 9) Massiot D., Fayon F., Capron M., King I., Le Calvé S., Alonso B., Durand J. O., Bujoli B., Gan Z., Hoatson G., Magn. Reson. Chem., 40, 70–76 (2002).
• 10) Brunauer S., Emmett P. H., Teller E., J. Am. Chem. Soc., 60, 309–319 (1938).
• 11) Wilson E. E., Awonusi A., Morris M. D., Kohn D. H., Tecklenburg M. M., Beck L. W., Biophys. J., 90, 3722–3731 (2006).
• 12) Zhu P., Xu J., Sahar N., Morris M. D., Kohn D. H., Ramamoorthy A., J. Am. Chem. Soc., 131, 17064–17065 (2009).
• 13) Jäger C., Welzel T., Meyer-Zaika W., Epple M., Magn. Reson. Chem., 44, 573–580 (2006).
• 14) Bracco S., Brajkovic A., Comotti A., Rolandi V., Period. Mineral., 82, 239–250 (2013).
• 15) Sing K. S. W., Pure Appl. Chem., 57, 603–619 (1985).
• 16) Votyakov S., Kiseleva D., Shchapova Y., Smirnov N., Sadykova N., J. Therm. Anal. Calorim., 101, 63–70 (2010).
• 17) Chen Y. N., Chai L. Y., Shu Y. D., J. Hazard. Mater., 160, 168–172 (2008).
• 18) Siebers N., Leinweber P., J. Environ. Qual., 42, 405–411 (2013).