Accepted Manuscript

Title: Determination of withaferin A and withanolide A inmice plasma using high-performance liquidchromatography-tandem mass spectrometry: Application topharmacokinetics after oral administration of Withaniasomnifera aqueous extract

Author: <ce:author id="aut0005"> Dada Patil<ce:authorid="aut0010"> Manish Gautam<ce:author id="aut0015">Sanjay Mishra<ce:author id="aut0020"> SureshKarupothula<ce:author id="aut0025"> SunilGairola<ce:author id="aut0030"> Suresh Jadhav<ce:authorid="aut0035"> Shrikrishna Pawar<ce:author id="aut0040">Bhushan Patwardhan

PII: S0731-7085(13)00111-8DOI: http://dx.doi.org/doi:10.1016/j.jpba.2013.03.001Reference: PBA 8995

To appear in: Journal of Pharmaceutical and Biomedical Analysis

Received date: 23-1-2013Revised date: 1-3-2013Accepted date: 6-3-2013

Please cite this article as: D. Patil, M. Gautam, S. Mishra, S. Karupothula, S. Gairola,S. Jadhav, S. Pawar, B. Patwardhan, Determination of withaferin A and withanolideA in mice plasma using high-performance liquid chromatography-tandem massspectrometry: Application to pharmacokinetics after oral administration of Withaniasomnifera aqueous extract, Journal of Pharmaceutical and Biomedical Analysis (2013),http://dx.doi.org/10.1016/j.jpba.2013.03.001

This is a PDF file of an unedited manuscript that has been accepted for publication.As a service to our customers we are providing this early version of the manuscript.The manuscript will undergo copyediting, typesetting, and review of the resulting proofbefore it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers thatapply to the journal pertain.

Page 1 of 35

Accep

ted

Man

uscr

ipt

1

Determination of withaferin A and withanolide A in mice plasma using high-performance 1

liquid chromatography-tandem mass spectrometry: Application to pharmacokinetics after 2

oral administration of Withania somnifera aqueous extract3

4567

Dada Patila, Manish Gautamb, Sanjay Mishraa, Suresh Karupothulab, Sunil Gairolab, 8

Suresh Jadhavb, Shrikrishna PawarC, Bhushan Patwardhand*9

10

11a Serum Institute of India Research Foundation, Hadapsar, Pune 411 028 (India)12

13b Serum Institute of India Limited, Hadapsar, Pune 411 028 (India)14

15c Synapse Labs Pvt Ltd. Pune- 411 014 (India)16

17d Interdisciplinary School of Health Sciences, University of Pune, Pune 411 007 (India))18

192021

* Address of author for correspondence22Professor and Director, Interdisciplinary School of Health Sciences23University of Pune24Pune 411 00725India26Tel: +91-020-2569017427Fax: +91-020-2569017428Email: bhushan@unipune.ac.in29

303132

33

34

35

36

Page 2 of 35

Accep

ted

Man

uscr

ipt

2

36Highlights37Withanolides are active constituents reported from Withania somnifera (WSE)38Quantitative LC-MS/MS method was developed for the withanolides in plasma 39The method was validated as per bioanalytical method validation guideline40PK parameters for the withanolides were evaluated after oral feeding of WSE41Withanolide A has shown more relative BA compared to Withaferin A 42

43444546

Abstract47Withania somnifera (WS) is one of the popular botanical medicines widely used in 48

Ayurveda. Withanolides such as withaferin A (WA) and withanolide A (WLD) are its 49

bioactive constituents reported as promising drug candidates in cancer and neurological 50

disorders respectively. A new, selective and rapid high performance liquid 51

chromatography-tandem mass spectrometry (HPLC-MS/MS) method has been developed 52

and validated for simultaneous determination of WA and WLD in mice plasma. Simple 53

liquid-liquid extraction procedure was followed using ter- butyl methyl ether (TBME) for 54

plasma sample pretreatment. Analytes were separated on Hypurity C18 column using 55

methanol and ammonium acetate (95:5, v/v) as a mobile phase and detected by 56

electrospray ionization in the multiple reaction monitoring (MRM) mode. The mass 57

transition ion-pair was followed as m/z 471.3→ 281.2 for WA; m/z 437.2→ 292.2 for 58

tianeptine (IS) and m/z 488.3→ 263.1 for WLD; m/z 315.9→ 270 for clonazepam (IS). 59

The method showed excellent linearity (r2 > 0.997) over the concentration range of 0.48460

-117.880 ng/ml for WA and from 0.476 -116.050 ng/ml for WLD. The lower limits of 61

quantification (LLOQs) were found to be 0.484 ng/ml and 0.476 ng/ml for WA and WLD 62

respectively. Precision (% CV) and accuracy (% bias) were found in the range of 3.7% -63

14.3 % and -14.4 to 4.0 % respectively. The validated method was successfully applied to 64

Page 3 of 35

Accep

ted

Man

uscr

ipt

3

a pharmacokinetic (PK) study for estimation of WA and WLD in mice plasma following 65

oral administration of W. somnifera root aqueous extract (WSE). The PK study suggested 66

rapid oral absorption of these withanolides. The PK study revealed that withaferin A has 67

one and half times more relative bioavailability as compared to withanolide A. 68

69

Key words:70

Withania somnifera71

Withaferin A72

Withanolide A73

Pharmacokinetic study74

HPLC-ESI-MS/MS75

Tandem mass spectrometry76

77

Page 4 of 35

Accep

ted

Man

uscr

ipt

4

771. Introduction78

Pharmacokinetic (PK) studies on bioactive constituents of herbal drugs (HD) provide 79

valuable information on bio-transformed metabolites, dosage form, doses and potential 80

herb-drug interactions [1]. PK studies are recommended by many international regulatory 81

agencies such as United States Food and Drug Administration (US FDA) and European 82

Medicines Agency (EMA) during various stages of HD development [2,3]. HD PK83

studies are challenging due to their complex physico-chemical properties, limited 84

information or redundancy on active principles and lack of sensitive bioanalytical 85

methods. Recent advances in analytical tools and imaging techniques made such studies 86

possible to the larger extend. Most of the reported studies on HD PK were based on 87

targeted or untargeted (global) metabolites profiling after oral administration crude or 88

single chemical components [4]. The targeted PK studies based on active constituents 89

such as steroidal saponins ( dioscin, protodioscin), triterpenoid saponins (boswellic acids,90

gymnemagenin, ginsenosides), polyphenolics (curcumin, resveratrol) and alkaloids ( 91

berberine, palmatine) have been reported [5,6,7].92

Withanolides are steroidal lactones such as withaferin A (WA) and withanolide A 93

(WLD) present in Withania somnifera (WS) (L.) Dunal, (family: Solanaceae, commonly 94

known as Indian ginseng; winter cherry or Ashwagandha). These withanolides are 95

considered as active constituents responsible for several pharmacological activities in 96

varying concentration from micrograms to milligrams range [8]. WA has been reported 97

as promising anti-cancer drug candidate due to its cytotoxic [9], apoptotic [10], anti-98

metastatic [11], anti-mitotic [12] and anti-angiogenesis properties [13]. Withaferin A was 99

reported for anti-cancer (3.5 mg/kg), anti-inflammatory (2.15 mg/kg), anti-angiogenesis 100

Page 5 of 35

Accep

ted

Man

uscr

ipt

5

(7-200 µg/kg) [14], anti-parasitic (0.3 mg/kg) [15] and hepatoprotective (10 mg/kg) [16] 101

potential. WLD has attracted interest due to its neuropharmacological properties of 102

promoting synaptic and outgrowth reconstruction at a dose of 4.7 µg/kg [17]. WLD is 103

therefore an important candidate for the therapeutic treatment of neurodegenerative 104

diseases, like Alzheimer disease (AD), Parkinson’s disease (PD), convulsions, cognitive 105

function impairment, as it is able to reconstruct neural networks [18]. The underlined 106

molecular mechanisms studies on WA and WLD demonstrated modulation of multiple 107

targets such as transcriptional factors, inflammatory cytokines, enzymes, growth factors, 108

receptors and other targets suggesting promising drug candidates for cancer and 109

neurological disorders [19]. The chemical structures and possible molecular mechanisms 110

of these withanolides are shown in Fig. 1. 111

Despite many mechanistic studies, PK data of these withanolides after oral 112

administration in animals or humans has not been reported yet. Recently, evaluation of 113

PK parameters of withanolides using in silico and in vivo models were attempted. In114

silico study showed log P value of 3.06 with prediction of 91.6 % of oral absorption for 115

WLD and such data was not reported for WA [20]. In another study, high performance 116

liquid chromatography with ultra violet visible detector (HPLC-UV) method was 117

reported for quantitative determination of WA in human plasma and showed sensitivity 118

upto 50 ng/ml [21]. However, PK data for targeted withanolides such as WA and WLD 119

has not reported.120

In present work, we developed a new rapid and sensitive HPLC coupled with 121

tandem mass spectrometry (HPLC-MS/MS) method for simultaneous determinations of 122

WA and WLD in mice plasma using multiple reaction monitoring (MRM) mode. HPLC-123

Page 6 of 35

Accep

ted

Man

uscr

ipt

6

MS/MS method provides several advantages over HPLC-UV and HPLC-MS techniques 124

in terms of sensitivity and selectivity etc. The method was thoroughly validated as per 125

international guidance on bioanalytical method validation and successfully applied for the 126

evaluation of PKs of WA and WLD after oral administration of standardized W. 127

somnifera aqueous extract (WSE) in mice. W. somnifera is one of the popular botanical 128

used in Indian traditional system of medicine known as Ayurveda for many therapeutic 129

and health benefits [22]. Its monograph is available in Indian Pharmacopoeia and 130

American Herbal Pharmacopoeia [23,24]. 131

132

Page 7 of 35

Accep

ted

Man

uscr

ipt

7

2. Experimental132

2.1. Materials and reagents133

Reference standards of WA (% purity ≥ 99.0) and WLD (% purity ≥ 99.0) were 134

purchased from Chromadex (CA, USA). Internal standards (IS) such as tianeptine and 135

clonazepam for WA and WLD respectively were procured from Synapse Labs Pvt. Ltd.136

(Pune, India). Methanol and ammonium acetate was purchased from Merck (Mumbai, 137

India) and tert-butyl methyl ether (TBME) and HPLC grade water were used from J. T. 138

Baker (Mumbai, India). Other reagents were of analytical grade if not otherwise stated. 139

2.2. Preparation of WSE and determination of withanolide contents140

Roots of W. somnifera were procured from Natural Remedies (Bangalore, India)141

and certified as authentic by National Institute of Science Communication and 142

Information Resources (NISCAIR), New Delhi, India (vide NISCAIR/RHM/F-143

3/2003/413). Powdered roots were extracted with distilled water as a decoction as per 144

method described in the Ayurvedic Pharmacopoeia of India. Dried aqueous extract 145

(WSE) was ensured to be free from pathogens, aflatoxins, pesticide residues and heavy 146

metals to meet WHO guidelines [25]. Chromatographic fingerprint of WSE was 147

developed and content of WA and WLD were determined using in-house HPLC-DAD 148

method [26]. The contents of WA and WLD were found to be 0.04585 g/100g and149

0.04785 g/100g of extract respectively.150

2.3. Liquid chromatography and tandem mass spectrometry151

The HPLC-ESI-MS/MS system consisted of HPLC-DAD system with automatic 152

injector, autosampler, binary pump, degasser and column oven from Shimadzu 153

Corporation (Kyoto, Japan). The HPLC system was coupled with Applied 154

Page 8 of 35

Accep

ted

Man

uscr

ipt

8

Biosystem/MDS SCIEX API-3000 triple quadrupole mass spectrometer equipped with 155

electrospray ionization (ESI) source (MA, USA). Mass spectrometer was operated in the 156

positive ion mode. Analytes were quantified using MRM. All the MS parameters were 157

manually fine-tuned to obtain the highest MRM signals. The following MRM transitions 158

of precursor ions to product ions were used: m/z 471.3→281.2 for WA; m/z 437.2→292.2 159

for tianeptine (IS); m/z 488.3→263.1 for WLD and m/z 315.9→270 for clonazepam (IS). 160

The optimized MS parameters for the detection of WA and WLD were listed in Table 1.161

Chromatographic separation was carried out on a reverse phase Hypurity C18 column (50 162

mm x 4.6 mm, 5 µm; Thermo scientific, Mumbai, India) using isocratic mobile phase 163

methanol and 2 mM ammonium acetate (95:5, v/v) at a flow rate of 0.4 ml/min. The 164

column and autosampler were maintained at 25ºC and 4ºC, respectively. The total 165

analytical run time was 3 min. Data acquisition was performed with Analyst version 1.4166

software.167

2.4. Preparation of calibration standards and quality control (QC) samples168

The stock of standard solution containing WA (1 mg/ml) and WLD (1 mg/ml) 169

was prepared in methanol. The IS solutions were prepared in deionized water. All 170

standard solutions were stored at 4ºC in polypropylene tubes in the dark when not in use. 171

Calibration standards were prepared by spiking appropriate amount of the standard 172

solution in drug free mice plasma samples to yield final concentrations of 0.484, 0.967,173

4.71, 23.58, 47.17, 70.73, 94.30 and 117.88 ng/ml for WA and 0.476, 0.952, 4.64, 23.22, 174

46.44, 69.63, 92.84 and 116.05 ng/ml for WLD. Four levels of QC samples containing 175

WA (0.505, 1.344, 58.94 and 90.68 ng/ml) and WLD (0.493, 1.316, 57.70 and 88.77176

ng/ml) were prepared in the same manner. 177

Page 9 of 35

Accep

ted

Man

uscr

ipt

9

2.5. Sample preparations178

One hundred microliter aliquot of the plasma sample was mixed with 100 µL of 179

IS solution on vortexer. The mixture was extracted with 2000 µL of TBME by vortex 180

mixing for 5 min at high speed and centrifuged at 600 x g for 5 min. The organic layer 181

transferred and evaporated to the dryness under nitrogen evaporator. The residues were182

dissolved in 100 µL of mobile phase and transferred to 1.5 ml autosampler vials. 20 µL 183

was used for injection in the HPLC-ESI-MS/MS system.184

2.6. Method validation185

A thorough and complete method validation was carried out according to the 186

industrial guideline for bioanalytical method validation from the US FDA [27]. 187

2.6.1. Specificity and selectivity188

The specificity and selectivity of the method toward endogenous plasma matrix 189

components was ascertained by analyzing blank mice plasma samples from six sources, 190

blank plasma samples spiked with WA and WLD at lower limit of quantification (LLOQ) 191

and plasma samples obtained from PK studies. 192

2.6.2. Linearity and LLOQ193

The linearity of the method was generated by analysis of five calibration curves 194

containing eight non-zero concentrations. Each calibration curves was analyzed 195

individually by fitting the area ratio response for analytes/IS as a function of standard 196

concentration, using no weighted or least square weighted (1/x or 1/x2 ) linear regression 197

and excluding the point of origin. The limits of detection (LOD) and lower limits of 198

quantitation (LLOQ) were determined at S/N of 3 and 10 under optimized 199

chromatographic conditions. The LLOQ was defined as the lowest concentration yielding 200

Page 10 of 35

Accep

ted

Man

uscr

ipt

10

a precision with coefficient of variance (CV) less than 20% and accuracy within 20% of 201

the nominal value. 202

2.6.3. Accuracy and precision203

The precision and accuracy were evaluated by assaying five replicates of four 204

different QC samples on the same day and three consecutive days. The accuracy was 205

calculated from the nominal concentration (Cnom) and the mean value of observed 206

concentration (Cobs) as follows: accuracy (bias, %) = [(Cnom-Cobs)/ Cnom] x 100. The 207

precision (relative standard deviation, RSD) was calculated from the standard deviation 208

and observed concentration as follows: precision (RSD, %) = [standard deviation 209

(SD)/Cobs] x 100. The inter-day and intra-day accuracy and precision value for the lowest 210

acceptable reproducibility concentrations were defined as being within ±15%. 211

2.6.4. Extraction recovery and matrix effect212

The percentage extraction recoveries of the analytes were determined at three QC 213

levels (low, mid and high) by calculating as the ratio of the mean peak area of analytes214

spiked before extraction (R1) to the mean peak area of analytes spiked post-extraction 215

(R2) x 100. As per acceptance criteria, recovery should be consistent, precise and 216

reproducible. The matrix effect was determined at three QC concentrations of analytes as 217

the ratio of the mean peak area of analytes spiked post-extraction (R2) to the mean peak 218

area of the same analytes neat standard solution (R3) x 100. It was considered negligible 219

if no more than a 15% difference was observed between plasma spiked with two 220

withanolides and neat standard solution. The value of matrix effect less than 85% 221

represented ionization suppression while, more than 115% depicted ionization 222

enhancement. 223

Page 11 of 35

Accep

ted

Man

uscr

ipt

11

2.6.5. Stability224

The stability of QC (low and high concentration level) samples was assessed by 225

analyzing samples stored room temperature for 6 h, three successive freeze (-70ºC) - thaw 226

(at room temperature) cycles and processed samples under autosampler condition (4ºC) 227

for 12 h. All stabilities were calculating as the ratio of average concentration and freshly 228

prepared samples.229

2.7. Application of the method to PK study230

Swiss albino female mice weighing from 20-24 g were used in the study. The 231

mice were housed and maintained under controlled environmental conditions 232

(temperature 22 ± 2ºC; humidity 55 ± 10%) with free access to food and drinking water 233

until 12 h prior to experiments. The experimental protocol was approved by the 234

Institutional Animal Ethics Committee of Serum Institute of India Ltd, Pune, India. All 235

mice were intragastric administered WSE suspended in 0.5% carboxymethyl cellulose 236

sodium aqueous solution at 1 g/kg. Blood samples were collected into potassium-237

ethylenediamine tetra-acetic acid (K2-EDTA) vaccutainer from each mouse by retro-238

orbital venous plexus. Blood (0.3-0.4 ml) was collected at time intervals of 0, 5, 10, 20, 239

30, 60, 90, 120, 180, 210 and 240 min after oral administration of WSE. Single mouse 240

was bled at every different time point to prevent hypovolemia and dehydration, which 241

may alter the pharmacokinetic parameters. Total six mice were used for each time point 242

to diminish individual variation. The plasma collected from six vehicle administered243

mice served as blank. All blood samples were immediately centrifuged (at 3000 x g for 244

10 min at 4ºC) and obtained plasma was stored at -70ºC until analysis. 245

Page 12 of 35

Accep

ted

Man

uscr

ipt

12

2.8. PK parameters and statistical analysis246

The pharmacokinetic parameters such as area under the plasma concentration-247

time curve from 0 to the last measured concentration (AUC0-t ) and area under the plasma 248

concentration-time curve extrapolated to infinity (AUC0-∞), terminal elimination half-life 249

(T1/2) and oral clearance (CL) were estimated by non-compartmental analysis using 250

pharmacokinetic program, WinNonlin version 3.0 from Pharmasight corporation, 251

(Mountain view, CA, USA). The maximum plasma concentration (Cmax) and the time to 252

reach Cmax (Tmax) were obtained directly from the plasma concentration-time curve. All 253

values were expressed as mean ± standard deviation (S.D.) except Tmax mentioned as 254

median.255

256257258259260261262263264265266267268269270271272273274275

Page 13 of 35

Accep

ted

Man

uscr

ipt

13

2753. Results and discussion276

3.1. Chromatographic and tandem mass spectrometric condition277

To optimize ESI conditions for detection of analytes, both the positive and 278

negative ion modes were investigated. Both the analytes showed the maximum response 279

in positive ionization mode. The full scan ESI spectra revealed higher signals at m/z 280

471.3 [M+H]+ and m/z 488.3 [M+NH4]+ for WA and WLD respectively as shown in Fig.281

2. The product ion spectrum of precursor ions produced a fragmentation patterns lead to 282

ions at m/z 281.2 and m/z 263.1 for WA and WLD respectively as shown in Fig. 3. The 283

product ion at m/z 281.2 was prominently formed from [M+H]+ ions of WA due to 284

cleavage of C17-C20 position to lose lactone moiety, subsequently undergoes 285

rearrangement (-2H) and loss of one water molecule (Fig. 3 A). Similarly, [M+NH4]+286

ions of WLD undergoes fragmentation with additional loss of one water molecule to form 287

the product ion at m/z 263.1 (Fig. 3 B). The fragmentation pattern was found to be 288

consistent with the mass fragmentation of withanolides published earlier [28,29]. The 289

quantification of analytes was performed using MRM data acquisition mode due to its 290

sensitivity and selectivity, where the precursor and product ions are monitored. The 291

chromatographic conditions with respect to mobile phase, column and isocratic elution 292

were optimized on the basis of peak shape, response and resolution of WA and WLD. 293

The use of methanol and ammonium acetate improved the peak shape and signal intensity 294

in HPLC-ESI-MS/MS analysis. The use of simple isocratic elution of methanol and 295

ammonium acetate (95:5, v/v) yielded retention time less than 3 min, allowing high 296

sample through put. The representative HPLC-ESI-MS/MS chromatograms of WA and 297

Page 14 of 35

Accep

ted

Man

uscr

ipt

14

WLD in mice plasma samples are shown in Fig. 4 and Fig. 5 respectively. Retention 298

times (tR) for WA and WLD were found to be 1.81 min and 1.87 min respectively. 299

3.2. Optimization of sample pre-treatment300

Plasma sample preparation is an important step, employed with an aim to remove 301

interferences from biological samples using simple procedure having suitable recovery. 302

Liquid–liquid extraction, protein precipitation and solid phase extraction are the most 303

widely used sample preparation techniques – in bioanalytical method development. In the 304

present study, liquid–liquid extraction with TBME resulted in consistent good recoveries305

of analytes (>55%). The advantage of using liquid–liquid extraction was that it minimizes 306

chances of errors, saves time and simplifies the sample preparation methodology. In 307

quantitative analysis of analytes in biological samples, an appropriate IS is needed. 308

Several compounds were investigated to find a suitable IS. Tianeptine and clonazepam309

were used as IS for WA and WLD respectively because of similar chromatographic 310

behaviors and extraction characteristics. 311

3.2. Method validation312

3.2.1. Specificity and selectivity313

Under the optimized HPLC-ESI-MS/MS conditions, the representative 314

chromatograms of blank plasma, blank plasma spiked with analytes and IS, and plasma 315

sample obtained at 3.5 h after oral administration WSE are shown in Fig. 4 and Fig 5. 316

The mass transition ion-pair was followed as m/z 471.3→ 281.2 for WA; m/z 437.2→ 317

292.2 for tianeptine (IS) and m/z 488.3→ 263.1 for WLD; m/z 315.9→ 270 for 318

clonazepam (IS). No significant interferences from endogenous substances were observed 319

Page 15 of 35

Accep

ted

Man

uscr

ipt

15

at the tR of WA, WLD and IS in the matrix, which indicated that the elaborated 320

procedures was specified and selective.321

3.2.2. Linearity, LOD and LLOQ322

The linearity of calibration curves were demonstrated by correlation coefficients 323

(r2) values obtained for the regression line. The calibration curves were obtained over the 324

concentration range of 0.484-117.880 ng/ml for WA and 0.476-116.050 ng/ml for WLD 325

in mice plasma showed r2 ≥ 0.997. Linear regression analysis with a weighting of 1/ x2326

gave the optimum precision (% CV) and accuracy (% bias) of the corresponding 327

calculated concentrations at each level as mentioned in the Table 2. The low % CV value 328

for the slope of WA and WLD showed 10.50 % and 12.90 % respectively indicated the 329

repeatability of the method. The LOD for WA and WLD was 0.1 ng/ml in mice plasma. 330

Sensitivity was evaluated by the LLOQ determinations of the method and was found to 331

be 0.484 and 0.476 ng/ml for WA and WLD respectively allowing sufficient for PK study 332

of the two withanolides following oral administration of WSE to mice. 333

3.2.3. Precision and accuracy334

The precision (intra and inter-day) and accuracy data on four different levels of335

QC samples containing WA and WLD are summarized in Table 3. The intra and inter-336

day precision (% CV) values ranged between 4.80-12.80% and 5.30-14.30 % for WA and 337

3.70-7.60 and 5.30-12.70 % for WLD. The accuracy (% bias) values ranged within -338

14.40 to -2.10 % and -10.50 to 0.30 % for WA and -11.10to 3.90 % and -3.80 to 4.00 % 339

for WLD at four QC levels. The data indicated that the precision and accuracy of this 340

method was acceptable.341

342

Page 16 of 35

Accep

ted

Man

uscr

ipt

16

3.2.4. Extraction recovery and matrix effect 343

The extraction recoveries and matrix effect from mice plasma are shown in Table 344

4. Extraction recovery (relative recovery) of analytes from the mice plasma after the 345

extraction procedure was assessed in three QC samples. The mean extraction recoveries 346

of 55.63 ± 0.42 and 62.85 ± 1.51 were found for WA and WLD respectively with % RSD 347

less than 15% and shown to be consistent, precise and reproducible. Matrix effect was 348

measured at three different QC levels for WA and WLD. All calculated values were 349

between 85.92 % ± 3.24 and 89.05 % ± 4.12. Slight ion suppression for all the analytes 350

was observed but they were all within the acceptable criteria. 351

3.2.5. Stability352

Stability of WA and WLD was evaluated under conditions mimicking situations 353

likely to be encountered during the experiment, storage and sample preparation 354

processes. Stability under conditions such as three freeze-thaw cycles, short term room 355

temperature storage and autosampler stability were evaluated. All established stabilities 356

for WA and WLD are shown in Table 4. Deviation of the mean observed concentrations 357

of the samples from the nominal concentration was within 15%, indicating no significant 358

loss of analytes. The reanalysis of the constituted solutions stored for 24 h at 4ºC showed 359

the acceptable accuracy and precision.360

3.3. Application of the method to study PKs of WA and WLD in mice361

The validated method was successfully applied for the determination of WA and 362

WLD after oral administration of WSE at a dose of 1000 mg/kg (equivalent to 0.4585 363

mg/kg of WA and equivalent to 0.4785 mg/kg of WLD) in mice. Although few 364

quantitative methods specifically for WA have been reported and showed less selectivity 365

Page 17 of 35

Accep

ted

Man

uscr

ipt

17

and sensitivity as their major limitations (17, 24). Therefore, aiming at a more selective 366

method, HPLC-ESI-MS/MS technique was chosen in the present study for simultaneous 367

estimation of WA and WLD with its high resolution, selectivity and sensitivity resulted in 368

the optimized LC conditions and triple-quadrupole tandem mass spectrometer in MRM 369

mode. The mean plasma concentration-time curves of WA and WLD are illustrated in 370

Fig. 6 and the pharmacokinetics parameters are shown in Table 5. The almost parallel PK 371

profiles of these withanolides were observed and it could be because of similar chemical 372

nature. 373

Maximum mean plasma concentrations (Cmax) of 16.69 ± 4.02 ng/ml and 26.59 ± 374

4.47 ng/ml for WA and WLD with observed Tmax of 10 and 20 min respectively. The 375

results indicated that the absorption of these withanolides might be rapid. The values 376

LLOQs of WA and WLD are less than Cmax/20 ratio, suggests that the method has 377

adequate sensitivity to measure these withanolides in the plasma samples. The area under 378

the plasma concentration-time curve from 0-4h (AUC0-4h) of 1572.27 ± 57.80 min ng/ml379

and 2458.47 ± 212.72 min ng/ml and extrapolated to infinity (AUC0-∞) of 1673.10 ± 380

54.53 min ng/ml and 2516.41 ± 212.10 min ng/ml for WA and WLD respectively. The 381

ratios of AUC0-4h to AUC0-∞ are greater than 0.8 and closer to 1.00 indicates that selection 382

blood sampling intervals and calculated terminal elimination half life (T1/2) were 383

appropriate. The T1/2 of 59.92 ± 15.90 min and 45.22 ± 9.95 min and clearance (CL) of 384

274.10 ± 9.10 and 191.10 ± 16.74 for WA and WLD respectively were observed. WLD 385

showed relatively higher extend of oral absorption (the higher value of Cmax, AUC0-∞ and 386

lower T1/2 and CL) compared with WA, although slightly different contents of 0.4585 387

mg/kg and 0.4785 mg/kg for WA and WLD respectively were quantified in WSE. The 388

Page 18 of 35

Accep

ted

Man

uscr

ipt

18

relative oral bioavailability of WLD showed 1.44 times greater than the WA calculated389

on the basis of dose normalization. The results might be presumably due to different 390

physicochemical properties of these withanolides or the PK interaction of the prescribed 391

chemical constituents. Overall, the information described in the present study might be 392

helpful for future studies on the PK evaluation of WSE or its formulations and beneficial 393

for application of this herb in the clinical therapy. 394

4. Conclusions395

A new HPLC-ESI-MS/MS method has been developed and validated for 396

simultaneous determination of WA and WLD in mice plasma. Simple liquid-liquid 397

extraction sample pre-treatment procedures using TBME was followed. The assay 398

provided adequate recovery and matrix effect with acceptable accuracy and precision for 399

bioanalytical method validation. The pharmacokinetic properties of withanolides were 400

characterized as rapid oral absorption following oral administration of WSE. These PK401

data on withanolides could help to elucidate the absorption mechanism of WSE and its 402

formulations, which might help to design newer rationale formulations and to study herb-403

drug interactions potentially influencing absorption and metabolism during concurrent 404

administration with other prescription drugs. In conclusion, this is the first study 405

reporting the PK evaluation of WA and WLD after oral administration. The established 406

method demonstrated good performance in terms of specificity, linearity, precision and 407

accuracy and was successfully applied for quantitative estimation of WA and WLD in 408

mice plasma to support PK studies. 409

410

411

Page 19 of 35

Accep

ted

Man

uscr

ipt

19

Acknowledgments 412413

This work was supported by Department of Science and Technology, Government 414

of India under Drugs and Pharmaceuticals Development Programme. 415

416

Page 20 of 35

Accep

ted

Man

uscr

ipt

20

416Figure captions417

418419

Figure 1. Chemical structures of withaferin A (A) and withanolide A (B) and their underline molecular 420mechanisms in cancer and neurological disorders respectively. The interaction of WA with the targets 421proteins regulates several cellular responses such as apoptosis, inflammation, angiogenesis and cell 422proliferation either by up regulation ( ) or down regulation ( ) or variable effect ( ) as shown in 423(C). WLD is known to modulate multiple synergistic targets such as down regulation ( ) of beta-site 424APP-cleaving enzyme 1 (BACE1); enzyme responsible of proteolytic cleavage of amyloid precursor 425protein (APP) and up regulation ( ) of disintegrin and metalloproteinase domain-containing protein 10426(ADAM 10); which involved in amyloidogenic and non-amyloidogenic processing of APP. In addition, it 427significantly up-regulated ( ) insulin degrading enzyme (IDE) levels, which has role in degrading 428excess β amyloid from the AD brain [30].429

430431

Figure 2. ESI MS of withaferin A (WA; mol wt., 470) and withanolide A (WLD; mol wt., 470) showed the 432presence of molecular ions at m/z 471.3 [M+H]+ and ammonium adducted molecular ions at m/z 488.3 433[M+NH4]

+ as shown in A and B respectively.434435436

Figure 3. ESI-MS/MS of withaferin A (WA) and withanolide A (WLD). WA undergoes cleavage at C17-437C20 position with a loss of lactone moiety, subsequently undergoes rearrangement (-2H) and loss of one 438water molecule resulted in formation daughter ion at m/z 281 (A). Similarly, WLD undergoes cleavage with 439a additional loss of water molecule resulted in formation of daughter ion at m/z 263.1 (B).440

441442

Figure 4. Representative HPLC-ESI-MRM chromatograms of withaferin A (WA) in blank mice and 443experimental mice plasma samples. A and B represents chromatograms of blank mice plasma without IS; C 444and D represents the chromatograms of blank mice plasma spiked with IS; E and F represents 445chromatograms of blank mice plasma spiked with 0.484 ng/ml of WA (at LLOQ level) with IS and 446chromatograms G and F showed the presence WA in mice plasma sample obtained at 3.5 h after oral 447administration of WSE at a dose of 1000 mg/kg in mice.448

449

Figure 5. Representative HPLC-ESI-MRM chromatograms of withanolide A (WLD) in blank mice plasma450and experimental mice plasma samples. A and B represents chromatograms of blank mice plasma without 451IS; C and D represents the chromatograms of blank mice plasma spiked with IS; E and F represents 452chromatograms of blank mice plasma spiked with 0.476 ng/ml of WLD (at LLOQ level) with IS and 453chromatograms G and F showed the presence WLD in mice plasma sample obtained at 3.5 h after oral 454administration of WSE at a dose of 1000 mg/kg in mice.455

456457

Figure 6. Mean plasma concentration versus time curve for withaferin A (▲) and withanolide A (■) after 458oral administration of Withania somnifera root aqueous extract at dose of 1000 mg/kg. Values are 459expressed as means ± SD (n = 6). 460

461462463464465466

Page 21 of 35

Accep

ted

Man

uscr

ipt

21

466Table 1. Optimized ESI-MS/MS parameters for withaferin A, withanolide A and internal 467standards (tianeptine and clonazepam)468

469ValuesParameters

Withaferin A Withanolide A Tianeptine Clonazepam Nebulizing gas (psi) 12 12 12 12Curtain gas (psi) 8 8 8 8CAD 8 8 8 8Ion source (V) 5000 5000 5000 5000Ion source temp (0C) 500 500 500 500Q1 mass (precursor ion) 471.30 488.30 437.20 315.90Q3 mass (product ion) 281.20 263.10 292.20 270DP (V) 44 30 35 45FP (V) 160 100 128 160EP (V) 4 5 10 8CEM (V) 2200 2200 2200 2200CE (V) 26 28 18 35CXP (V) 6 5 13 16MRM (amu) 471.3/281.2 488.3/263.1 437.2/292.2 315.9/270Dwell time (ms) 180 180 180 180

470CAD: collision-activated dissociation; DP: declustering potential; EP: entrance potential; FP: focusing 471potential; CEM: channel electron multiplier; CE: collision energy; CXP: collision cell exit potential; MRM: 472multiple reaction monitoring, V: volts; ms: milliseconds, amu: atomic mass unit.473

474

Page 22 of 35

Accep

ted

Man

uscr

ipt

22

Table 2. Slope and correlation coefficients (r2) values for withaferin A and withanolide A474475

Parameters Withaferin A Cnominal 0.48 0.96 4.71 23.58 47.17 70.73 94.30 117.88 Slope r2

Cobserved 0.49 0.90 4.98 26.54 48.29 67.25 92.07 113.76 0.0027CV (%) 2.20 6.60 8.50 4.60 3.70 2.70 2.20 4.40Bias (%) 2.40 - 6.10 5.60 12.50 2.40 - 4.90 - 2.40 - 3.50

10.50 0.997

Withanolide A Cnominal 0.47 0.95 4.64 23.22 46.44 69.63 92.84 116.05 Slope r2

Cobserved 0.47 0.94 4.59 23.85 45.89 67.97 92.71 118.71 0.0051CV (%) 5.00 9.60 5.80 1.50 5.80 2.70 4.60 4.50Bias (%) 0.40 - 0.60 - 1.10 2.70 - 1.20 - 2.40 -0.10 2.30

12.90 0.998

476Values are expressed as mean concentration (ng/mL; n= 5), Precision and accuracy were expressed in terms 477of % CV and % bias respectively, Correlation coefficient (r2) was used for determination linearity of 478calibration curves479

480481482

Page 23 of 35

Accep

ted

Man

uscr

ipt

23

Table 3. Precision and accuracy data for withaferin A and withanolide A482

Intra-day (n = 5) Inter-day (n =15)Cnom (ng/ml) Cobs

(ng/ml)Precision (% CV)

Accuracy (% bias)

Cobs

(ng/ml)Precision (% CV)

Accuracy (% bias)

Withaferin A0.50 0.48 12.80 -3.40 0.46 14.3 -6.801.34 1.31 5.20 -2.10 1.34 5.30 0.3058.94 53.55 4.80 -9.10 55.02 8.40 -6.6090.68 77.62 7.30 -14.40 81.15 5.90 -10.50

Withanolide A0.49 0.43 7.60 -11.10 0.47 12.70 -3.801.31 1.35 6.20 3.00 1.36 5.70 4.0057.70 59.93 3.70 3.90 59.81 7.20 3.7088.77 88.39 5.90 -0.40 89.60 5.30 0.90

483Values are expressed as mean concentration, Cnom and Cobs are the nominal and observed concentrations 484respectively, Precision and accuracy were expressed in terms of % CV and % bias respectively. 485

486

487488

Page 24 of 35

Accep

ted

Man

uscr

ipt

24

Table 4. Extraction recovery, matrix effect and stability of withaferin A and withanolide 488A in mice plasma489

490Extraction recovery and matrix effect (%)

Cnom (ng/ml) R1 R2 R3 Matrix effect RecoveryWithaferin A

1.34 0.65 ± 0.08 1.12 ± 0.10 1.38 ± 0.18 86.73 ± 3.60 55.08 ± 2.5658.94 31.36 ± 3.62 51.42 ± 5.61 59.05 ± 7.14 87.07 ± 5.47 55.59 ± 1.6190.68 44.70 ± 8.60 80.93 ± 8.12 90.98 ± 7.12 88.96 ± 4.25 55.23 ± 3.63

Withanolide A1.31 0.73 ± 0.09 1.16 ± 0.11 1.35 ± 0.11 85.92 ± 3.24 63.92 ± 2.8857.70 33.03 ± 3.18 51.48 ± 4.05 57.81 ± 3.13 89.05 ± 4.12 61.78 ± 3.2888.77 52.01 ± 6.02 77.81± 7.14 88.97 ± 6.18 87.46 ± 5.08 66.85 ± 5.13

Mice plasma stability Cnom (ng/ml) Freeze-thaw stability Short-term stability Autosampler stability

Withaferin A1.34 -13.10± 1.09 4.20 ± 1.01 4.10 ± 1.1290.68 -6.00 ± 0.88 -13.70 ± 2.19 -0.50 ± 0.09

Withanolide A1.31 14.20 ± 1.22 -8.20 ± 2.11 6.10 ± 2.0488.77 -9.00 ± 1.19 -6.60 ± 2.94 5.60 ± 1.29

491Data are expressed as means ± SD (n= 3), Cnom is the nominal concentration, R1 represents analytes spiked 492before extraction, R2 represents analytes spiked after extraction, R3 represents analytes prepared in 493injection solvent. Recovery (%) was calculated as R1/R2 x 100, Matrix effect (%) was calculated as R2/R3 494x 100.495

496497498

Page 25 of 35

Accep

ted

Man

uscr

ipt

25

Table 5. PK parameters of WA and WLD in mice plasma after oral administration of 498WSE 499

500Parameter Withaferin A Withanolide ACmax (ng/mL) 16.69 ± 4.02 26.59 ± 4.47Tmax (min) 20 (20-30) 10 (10-30)T1/2 (min) 59.92 ± 15.90 45.22 ± 9.95AUC 0-t (ng/mL min) 1572.27 ± 57.80 2458.47 ± 212.72AUC 0-∞ (ng/mL min) 1673.10 ± 54.53 2516.41 ± 212.10CL(mL/min/kg) 274.10 ± 9.10 191.10 ± 16.74

501Data expressed as mean ± SD (n= 6) except Tmax mentioned as median, Cmax: the maximum plasma 502concentration, Tmax: the time required to the Cmax, T1/2: elimination half life, AUC 0-t and AUC 0-∞ area under 503the concentration –time curve from 0 to t and 0 to ∞ respectively. CL: clearance. 504

505506507

References508 [1] B. Kamble, S. Talreja, A. Gupta, D. Patil, D. Pathak, I. Moothedath, B. Duraiswamy, Development and biological evaluation of Gymnema sylvestre extract-loaded nonionic surfactant-based niosomes. Nanomedicine (Lond). (2012).DOI: 10.2217/NNM.12.162 (in Press).

[2] Guidance for Industry Botanical Drug Products U.S. Department of Health andHuman Services Food and Drug Administration Center for Drug Evaluation andResearch (CDER) June 2004; 1-47.

[3] Committee on Herbal Medicinal Products (HMPC), Guideline on the use of the CTD format in the preparation of a registration application for traditional herbal medicinal products. European MedicinesAgency. August 2012. 1-25. EMA/HMPC/71049/2007 Rev. 1.

[4] W. Xiao-Dong, L. E-Hu, J. Yang, L. Chang-Yin, G. Wen, Q. Lian-Wen, Chong-Zhi Wang, Y. Chun-Su, L. Ping, Identification of metabolites of Buyang Huanwu Decoction in rat urine using liquid chromatography-quadrupole time-of-flight massspectrometry. J. Pharm. Biomed. Anal. 67-68 (2012) 114-122.

[5] K. Yu, F. Chen, C. Li, Absorption, disposition, and pharmacokinetics of saponinsfrom Chinese medicinal herbs: what do we know and what do we need to know more? Curr. Drug Metab. 13 (2012) 577-598.

[6] T. Lu, Y. Liang, J. Song, L. Xie, G.J. Wang, X.D. Liu, Simultaneous determination of berberine and palmatine in rat plasma by HPLC-ESI-MS after oral administration of traditional Chinese medicinal preparation Huang-Lian-Jie-Du decoction and the pharmacokinetic application of the method. J. Pharm. Biomed. Anal. 40 (2006) 1218-1224.

Page 26 of 35

Accep

ted

Man

uscr

ipt

26

[7] B. Kamble, A. Gupta, D. Patil, L. Khatal, S. Janrao, I. Moothedath, B. Duraiswamy,Determination of gymnemagenin in rat plasma using high-performance liquid chromatography–tandem mass spectrometry: application to pharmacokinetics after oral administration of Gymnema sylvestre extract. Biomed Chromatogr. 2012. DOI 10.1002/bmc.2845 (in Press).

[8] M. Mirjalili, E. Moyano, M. Bonfill, R.M. Cusido, J. Palazzos, Steroidal lactones from Withania somnifera, an ancient plant for novel medicine. Molecules 14 (2009) 2373-2393.

[9] S.K. Yousuf, R. Majeed, M. Ahmad, P. Sangwan, B. Purnima, A.K. Saxsena, K.A.Suri, D. Mukherjee, S.C. Taneja, Ring A structural modified derivatives of withaferin A and the evaluation of their cytotoxic potential. Steroids 76 (2011) 1213-1222.

[10] E. Mayola, C. Gallerne, D.D. Esposti, C. Martel, S. Pervaiz, L. Larue, B. Debuire, A. Lemoine, C. Brenner, C. Lemaire, Withaferin A induces apoptosis in human melanoma cells through generation of reactive oxygen species and down-regulation of Bcl-2. Apoptosis 16 (2011) 1014-1027.

[11] J.T. Thaiparambil, L. Bender, T. Ganesh, E. Kline, P. Patel, Y. Liu, M. Tighiouart, P.M. Vertino, R.D. Harvey, A .Garcia, A.I. Marcus, Withaferin A inhibits breast cancer invasion and metastasis at sub-cytotoxic doses by inducing vimentin disassembly and serine 56 phosphorylation. Int. J. Cancer 129 (2011) 2744-2755.

[12] S.D. Stan, Y. Zeng, S.V. Singh, Ayurvedic medicine constituent withaferin A causes G2 and M phase cell cycle arrest in human breast cancer cells. Nutr. Cancer 60 (2008) 51-60.

[13] R. Mohan, H.J. Hammers, P. Bargagna-Mohan, X.H. Zhan, C.J. Herbstritt, A. Ruiz, L. Zhang, A.D. Hanson, B.P. Conner, J. Rougas, V.S. Pribluda, Withaferin A is a potent inhibitor of angiogenesis. Angiogenesis 7 (2004) 115-122.

[14] R. Mohan, H.J. Hammers, P. Bargagna-Mohan, X.H. Zhan, C.J. Herbstritt, A. Ruiz, L. Zhang, A.D. Hanson, B.P. Conner, J. Rougas, V.S. Pribluda. Withaferin A is a potent inhibitor of angiogenesis. Angiogenesis 7 (2004) 115-122.

[15] S. Kushwaha, V.K. Soni, P.K. Singh, N. Bano, A. Kumar, R.S. Sangwan, S. Misra-Bhattacharya, (2011) Withania somnifera chemotypes NMITLI 101R, NMITLI 118R, NMITLI 128R and Withaferin A protect Mastomys coucha from Brugia malayi infection. Parasite Immunol. doi: 10.1111/j.1365-3024.2011.01352.x (in press).

[16] S. Sudhir, R.D. Budhiraja, Comparison of the protective effect of Withaferin-'A' and hydrocortisone against CCL4 induced hepatotoxicity in rats. Indian J. Physiol.Pharmacol. 36 (1992) 127-129.

Page 27 of 35

Accep

ted

Man

uscr

ipt

27

[17] T. Kuboyama, C. Tohda, K. Komatsu, Neuritic regeneration and synaptic reconstruction induced by withanolide A. Br. J. Pharmacol. 144 (2005) 961-971.

[18] C. Tohda, T. Kuboyama, K. Komatsu, Search for natural products related to regeneration of the neuronal network. Neurosignals 14 (2005) 34-45.

[19] W. Vanden Berghe, L. Sabbe, M. Kaileh, G. Haegeman, K. Heyninck, Molecular insight in the multifunctional activities of Withaferin A. Biochem. Pharmacol. 84 (2012) 1282-1291.

[20] N. Santhi, S. Aishwarya, Insights from the molecular docking of withanolide derivatives to the target protein PknG from Mycobacterium tuberculosis. Bioinformation 7 (2011)1-4.

[21] P. Patial, V. Gota, Rapid and sensitive method for determination of withaferin-A in human plasma by HPLC. Bioanalysis 3 (2011) 285-289.

[22] B. Patwardhan, A. Vaidya, M. Chorghade, Ayurveda and natural product drug discovery. Curr. Sci. 86 (2004) 789–799.

[23] Govt. of India, The Ayurvedic Pharmacopoeia of India, Ministry of Health and family welfare, Department of AYUSH, first ed., Vol. I (2007) pp. 15-17.

[24] Anonymous, American Herbal Pharmacopoeia. Therapeutic Compendium: Ashwagandha Root (Withania somnifera) – Standards of Analysis, Quality Control, and Therapeutics. (2000).

[25] World Health Organization. Quality control guidelines for medicinal plant material. (1998). Pp. 111.

[26] D. Patil, M. Gautam, U. Jadhav, S. Mishra, S. Karupothula, S. Gairola, S. Jadhav, B Patwardhan, Physicochemical stability and biological activity of Withania somniferaextract under real-time and accelerated storage conditions. Planta Med. 76 (2010) 481-488.

[27] US Food and Drug Administration. FDA Guidance for Industry: Bioanalytical Q10 Method Validation. US Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research: Rockville, MD (2001).

[28] D. Lavie, E. Glotter, Y. Shvo, Constituents of Withania somnifera Dun. Part IV. The structure of withaferin A. Journal of American Chemical Society 30 (1965) 7517-7531.

[29] B. Kamble, A. Gupta, D. Patil, S. Janrao, L. Khatal, B. Duraiswamy, Quantitative estimation of gymnemagenin in Gymnema sylvestre extract and its marketed formulations

Page 28 of 35

Accep

ted

Man

uscr

ipt

28

using the HPLC-ESI-MS/MS Method. Phytochem. Anal. (2012) doi: 10.1002/pca.2392 (in press).

[30] S.P. Patil, S. Maki, S.A. Khedkar, A.C. Rigby, C. Chan, Withanolide A and asiatic acid modulate multiple targets associated with amyloid-beta precursor protein processing and amyloid-beta protein clearance. J. Nat. Prod. 73 (2010) 1196-1202.

Page 29 of 35

Accep

ted

Man

uscr

ipt

Page 30 of 35

Accep

ted

Man

uscr

ipt

Page 31 of 35

Accep

ted

Man

uscr

ipt

Page 32 of 35

Accep

ted

Man

uscr

ipt

Page 33 of 35

Accep

ted

Man

uscr

ipt

Page 34 of 35

Accep

ted

Man

uscr

ipt

Page 35 of 35

Accep

ted

Man

uscr

ipt