Identification of Phyllanthus niruri by FTIR spectroscopy with chemometrics

Phyllanthus niruri (Indonesian: meniran) is a crude drug used in scientific jamu for hyperuricemia, mild hypertension, osteoarthritis, hemorrhoids, and hypercholesterolemia. This plant contains various bioactive compounds such as flavonoids, terpenes, coumarins, lignans, tannins, saponins, alkaloids, etc. Phyllanthus niruri is a wild plant growing in the tropics and sub-tropics of Asia, America, and China. This herb grows well from the lowlands to the highlands, such as bushes, terrain, yards, roadsides, gardens, and rivers. The levels of active compounds in a plant can vary depending on many factors, such as growing location, temperature, humidity, rainfall, and others. This research aimed to evaluate and differentiate P. niruri herbs according to their geographical location using FTIR fingerprint coupled with chemometrics, Principal Component Analysis (PCA) and Cluster Analysis (CA). The results of this study showed that 14 samples of P. niruri are classified into two groups. The first group consists of P. niruri powder originating from Tawangmangu, Kediri, Surabaya, Bangkalan, Gresik, Mojokerto, Kertosono, Krian, Blitar, Nganjuk, unknown 1, unknown 2, and unknown 3, whereas the second group consists of only one sample from Pasuruan. In conclusion, FTIR fingerprint analyzed with chemometrics is adequate to differentiate the powder of P. niruri collected from different locations. FTIR fingerprints combined with chemometrics can be further considered as a method in the quality control process of P. niruri .


INTRODUCTION
Jamu is a traditional Indonesian medicine with plant, animal, and/or mineral as its active ingredients.It has been used empirically for hundreds of years.Jamu has been developed into standardized herbal medicine and phytopharmaca through pre-clinical and clinical tests, respectively.Another way taken by the Indonesian government to develop jamu is through the saintifikasi jamu program.This program is scientific proof of jamu through research based on health services (Siswanto, 2012).Until 2016, there were seven "scientific jamu" used for the treatment of hyperuricemia, mild hypertension, osteoarthritis, hypercholesterolemia, hemorrhoids, and dyspepsia.Those were also applied as a hepatoprotector (Triyono, 2016).
Previous studies have shown that the levels of active compounds in a plant can vary according to the cultivation area, such as its temperature, humidity, rainfall, altitude, and others (Kartini et al., 2019;Pisano et al., 2015;Yudthavorasit et al., 2014).These variations cause inconsistency in the quality and efficacy of herbal products.Therefore, it is necessary to confirm the geographical origin of the P. niruri to maintain consistency in quality, safety, and efficacy.It can be conducted through several approaches.Determination of the concentration of one or several bioactive compounds can be carried out for this purpose.However, this method has limitations.There are various bioactive compounds contained in P. niruri and there may be only a slight difference in concentration at each location of growth.Hence, just determining the concentration of a few compounds is not appropriate.For this reason, an analytical technique which permits us to observe the amount of phytochemicals in a crude drug is needed.Determining the profiles of bioactive constituents through chromatography or spectroscopy is better because it will describe the chemical properties of the plant entirely (Tistaert et al., 2011).
Fourier Transform Infrared (FTIR) is a rapid and sensitive spectroscopic method to be implemented in the determination of the fingerprints of herbal medicine raw materials.FTIR spectrum contains multivariate data in the form of absorbance or transmittance in the range of wavenumber 4000-400 cm -1 .Each sample will have a different spectrum profile where the transformations in the wavenumber and peak intensity of the FTIR spectrum suggest that the composition of chemical substances in the sample does change.Thus, the same plant species originated from different locations can be distinguished by their FTIR spectrum.This way, the analysis can be done even though we do not know the overall chemical compounds contained in the plant (Rohman et al., 2014;Sun et al., 2010).Distinguishing the FTIR spectra of a species with different geographical locations visually will be very difficult since the difference is only in the signal intensity.Accordingly, chemometrics is needed to distinguish them easily (Gad et al., 2013).This study aimed to analyze P. niruri herbs based on their geographical sources using FTIR fingerprint coupled with chemometrics.In this study, the chemometrics of Principal Component Analysis (PCA) and Cluster Analysis (CA) are applied.

MATERIALS AND METHOD Materials
The plant material used was Phyllanthus niruri herb.Determination of plants was carried out by the Center for Information and Development of Traditional Medicine, University of Surabaya (Determination number: 1405/D.T/VIII/2019) and by B2P2TOOT Tawangmangu (determination number: YK.01.03/2/289/2020).The collection of plants was taken in June 2019 from 10 locations in East Java, 1 location in Central Java, and 3 samples with unknown sources (Table 1).Chemicals include KBr spectroscopy grade (Merck, Darmstadt, Germany) and demineralized water.Equipment consisted of spectrophotometer Jasco FT/IR-4200 (Japan) with DTGS (deuterated triglycine sulfate) detector, moisture content analyzer (Moisture Analyzer HB43 Mettler-Toledo GmbH, Laboratory & Weighing Technologies, Switzerland), analytical balances (Ohaus), freeze dryer, and Ultrasonic bath (Powersonic bath AS ONE, Osaka, Japan).

Preparation of P. niruri powder
Aerial parts of P. niruri were used in this research (Figure 1).Plants were pulled out and then the roots and other foreign materials were removed.It was then washed with tap water, drained, and dried under indirect sunlight.P. niruri was then ground and sieved to the mesh of 45.

Extract preparation
Five grams of powdered P. niruri herbs was sonicated with 60 mL of demineralized water using an ultrasonic bath (42 KHz, 50°C) for 10 minutes.Water was used as the extraction solvent to facilitate the drying process of the extract using freeze-drying.The residue was re-extracted in the same way using 40 ml of water.Extracts were subsequently dried using a freeze dryer.

FTIR spectrum
The FTIR spectrum was made using a spectrophotometer Jasco FT/IR-4200, with a DTGS (deuterated triglycine sulfate) detector.A total of 5 mg of P. niruri powder or dried water extract was mixed with 95 mg of KBr, finely crushed in an agate mortar before being pressed with a manual hydraulic press to become a pellet.It was then placed in a sample holder.Spectrum was recorded in the range of 4000-400 cm -1 with a resolution of 4 cm -1 and operated with Spectra Manager Version 2 software.

Principal component analysis (PCA)
The FTIR spectra of P. niruri from different origins were further analyzed by applying chemometric methods, namely PCA (Principal Component Analysis) and CA (Cluster Analysis) using Minitab v.16 software (Minitab Inc., USA).PCA converts spectrophotometer FTIR data from ndimensional to principal component (PC) where there is one score describing each spectrum in a new, much lower-dimensional space in the data set.PCs are organized by their eigenvalues which quantify the number of variants sought by the PC.Then the PCs that capture the largest variation in the FTIR spectrum (the PC with the largest eigenvalue) are maintained while the PC containing only noise is discarded (the PC with the smallest eigenvalue).PCA score plots were used to objectively classify the sample from the measured traits.The sample distribution on this graph can reveal patterns that may correlate with the general characteristics of the sample.As for the loadings plot, the most varied variables will have the greatest effect on the orientation of the PCs (Yang et al., 2007).

Cluster analysis (CA)
Cluster analysis was carried out to group samples according to their chemical similarity.The samples are grouped in a high-dimensional space and thus form a dendrogram.Firstly, each sample forms a cluster, and the two closest objects are grouped.Then, the third sample joins the first two samples or the other two samples join together in a different cluster.Each stage produces one less cluster than the previous stage until all samples are finally gathered in one cluster.In this work, the mean relationship was used and the Euclidean distance was calculated (Yang et al., 2007).

RESULTS AND DISCUSSION Organoleptic feature of the sample
Phyllanthus niruri used in this work were obtained from 11 locations.One of these samples was from Tawangmangu.P. niruri from this origin was used as a reference for other origins because it is used in the Saintifikasi Jamu service by B2P2TOOT Tawangmangu.There were also 3 samples from unknown origins (commercially found) which their origins were then analyzed using FTIR fingerprint data.Organoleptic features and moisture content of each P. niruri powder are presented in Figure 2 and Table 2, respectively.The moisture content of all of the samples was not more than 10%.It indicates that the drying process that has been applied on P. niruri is sufficient.The low water and/or moisture content in plant material is important to inhibit microbial growth, the presence of fungi or insects, and deterioration following hydrolysis (WHO, 1998).Indonesian Herbal Pharmacopoeia required the LOD (Loss On Drying) of P. niruri is not more than 14% (RI, 2017).

FTIR fingerprint of Phyllanthus niruri
FTIR, a spectroscopic technique, is increasingly used in herbal fingerprint analysis.FTIR spectroscopy coupled with chemometrics is a fast and straightforward method and is widely used to analyze different plant species, to predict the main active constituents, and also to ensure their quality and authenticity.However, this method has drawbacks when compared to chromatography methods.There is no separation process; thus how many compounds are contained in the samples, including a crude drug as starting material of herbal medicine, cannot be predicted.
FTIR spectrum of a plant can be recorded in several ways, either by using its extract or directly using the plant powder.In this study, as a screening phase, the two methods were carried out.Phyllanthus niruri herb was extracted with water and dried under a freeze dryer.This extract and its plant powder were then crushed with KBr and pressed into pellets.The FTIR spectra of these two samples (Figure 3) were then compared.
Figure 3 exhibited that the spectrum of the two samples is similar.However, P. niruri powder showed a higher number of signals.Phyllanthus niruri contains various chemical constituents (Bagalkotkar et al., 2006), both polar and non-polar.Non-polar compounds such as terpenes and alkaloids are not well extracted using water; therefore, the FTIR spectrum of the water extract showed fewer bands.According to the general principle of fingerprint analysis, the constituents or sample characters are portrayed and evaluated as much as possible; hence it will be possible to take accurate conclusions.Therefore, the FTIR spectrum of all samples was carried out on P. niruri powder for the next steps.Those steps follow the previous studies (Christou et al., 2018;Man et al., 2015;Saraf et al., 2018).Technically, preparing KBr pellets using plant powder is more manageable, economical, and faster than extracts.The FTIR profiles of P. niruri from various regions were very identical, with slight differences in the transmittance of each peak.It indicates that their phytochemicals are similar, only differing in concentration (Figure 4).All samples have some distinctive bands.A reasonably wide band at wavenumber 3351 cm -1 indicates the presence of strain vibrations of the O-H group.The bands' appearance shows stretching vibrations C-H and C = O at 2917 cm -1 , 1714 cm -1 , and 1644 cm -1 , respectively.Based on these bands, it is predicted that P. niruri contains phenolic and flavonoid compounds.Note: Tawangmangu (1), Kediri (2), Surabaya (4), Bangkalan (5), Gresik (6), Mojokerto (7),Kertosono (8), Krian (9), Blitar (10), Nganjuk (11), unknown 1 (12), unknown 2 (13), and unknown 3 ( 14)

FTIR fingerprint analysis by PCA
To analyze the FTIR fingerprint of P. niruri, FTIR spectrum data were then tabulated.Rows reflect the origin of the samples, whereas the columns indicate the transmittance value of every peak at every wavenumber.The spectra at the wavenumbers of 1800-400 cm -1 were tabulated to form 14 x 24 matrix data.These multivariate data were then analyzed using PCA.The first and second PCs described 95.6% and 2.3% of the variability in the original observations respectively, and both PCs accounted for 97.9% of the total variance.Thus, the first two PCs concentrated the multidimensional information into a 2-D dataset to classify the samples.Figure 5 shows that P. niruri grouped into two groups, where the samples from Tawangmangu, Kediri, Surabaya, Bangkalan, Gresik, Mojokerto, Kertosono, Krian, Blitar, Nganjuk, unknown 1, unknown 2, and unknown 3 were group one, while sample from Pasuruan as group two.As a statistical tool that is driven by data, PCA works by identifying the similarity of origin's fingerprint.Although they come from different zone, if two or more origins have similar FTIR fingerprint, PCA will groups them into one group.In this research PCA excludes other variables, such as weather or geographical characteristics, it only considers FTIR fingerprint data.Figure 5 shows that different zones can have similar FTIR fingerprint, only Pasuruan origin has significant FTIR fingerprint difference, so it can not be grouped with others.

FTIR fingerprint analysis by CA
Although PCA can exhibit groups like multiple objects, PCA does not always do this successfully.Therefore, in this study, CA-Cluster Analysis-was created to acquire groups from the auto-scaled data set and compare those two with previous ones from PCA.As the basis for classification, two main components are taken.The clusters formed (Figure 6) indicated distinct characteristics in P. niruri.In this research, CA uses a single linkage with Euclidean distance to measure the similarity among origins.This analysis also found that P. niruri can be classified into two clusters as in PCA.The first cluster is marked in red, while the second cluster is marked in green.The first cluster consists of P. niruri originating from Tawangmangu (1), Kediri (2), Surabaya (4), Bangkalan (5), Gresik (6), Mojokerto (7), Kertosono (8), Krian (9), Blitar (10), Nganjuk (11), unknown 1 (12), unknown 2 (13), and unknown 3 (14).The second cluster consists of only one sample, which was from Pasuruan (3).P. niruri from Tawangmangu is a standard material for Saintifikasi Jamu practice.Therefore, this research indicated that all samples except Pasuruan have a similar quality to those from Tawangmangu.
In this case, the CA result was not different from PCA.Only Pasuruan origin has a diverse group (cluster), and this result strengthens previous analysis that FTIR fingerprint data would provide an alternative conclusion in grouping origins.Figure 6 shows how CA steps work; the analysis is started by grouping the first two most similar origins into a cluster, continued with other high similarity ones.Formed clusters with high similarity are then also grouped into new parent clusters.This algorithm finally showed by the dendogram.Similar to this research, CA analysis with dendogram for origins grouping is indicated in the previous study (De Luca et al., 2011).

CONCLUSION
Based on the FTIR fingerprint coupled with PCA and CA, Phyllanthus niruri from eleven origins in East and Central Java as well as three unknown samples can be grouped into two clusters.The first cluster consists of P. niruri from Tawangmangu, Kediri, Surabaya, Bangkalan, Gresik, Mojokerto, Kertosono, Krian, Blitar, Nganjuk, unknown 1, unknown 2, and unknown 3, whereas the second cluster consists of sample from Pasuruan.FTIR fingerprint analyzed with chemometrics was adequate to differentiate the powder of P. niruri collected from different locations.FTIR fingerprints combined with chemometrics can be further considered as a method in the quality control process of P. niruri.

Table 1 . Geographical origin of Phyllanthus niruri
m a.s.l: meter above sea level; CJ: Central Java; EJ: East Java