Stress degradation study of bortezomib: Effect of co-solvent, isolation and characterization of degradation products by UHPLC-Q-TOF-MS/MS and NMR and evaluation of the toxicity of the degradation products

Article abstract of DOI:10.1039/d0nj05781a

Bortezomib (BTZ) is a first-in-class, potent reversible inhibitor of proteasome used in the treatment of multiple myeloma, the second most common hematological cancer. Stress degradation studies were performed to investigate the inherent stability of the drug according to ICH recommended guidelines Q1A (R2). Stress experiments were carried out in two ways using acetonitrile and methanol as co-solvents under various conditions. A selective stability-indicating LC-MS method has been developed to separate all degradation products of the drug on a Hibar-Purospher STAR, C18 (250 × 4.6 mm, 5 μm) column using a mobile phase consisting of 0.1% formic acid and acetonitrile in the gradient mode. BTZ was found to undergo degradation under acidic, basic, neutral hydrolysis and oxidative conditions, whereas it was stable under other conditions. Thirteen degradation products (DP-1-DP-13) were identified using acetonitrile as a co-solvent. Additionally, three (DP-14-DP-16) degradation products were found where methanol was used as a co-solvent. A total of 16 (DP-1-DP-16) degradation products were characterized by liquid chromatography-tandem mass spectrometry (LC-ESI-Q-TOF/MS/MS) and high-resolution mass spectrometry (HRMS). Major degradation products, DP-3, DP-6, DP-9, DP-10, DP-11 and DP-12, formed under oxidation conditions were isolated using preparative HPLC and characterized by 1D and 2D NMR experiments. Furthermore, in vitro cytotoxicity of isolated DPs was tested on normal cell lines such as CHO-K1, HEK-293 and NRK-49F by MTT assays. This study revealed that they were around 2-6 times less toxic as compared with the standard control of the drug and DP-10 showed relatively more toxicity than other isolated DPs against rat kidney cells at 18.20 μM. In silico toxicity studies suggested that BTZ and its DPs can be hepatotoxic and genotoxic resulting in severe toxicity.

Full text of DOI:10.1039/d0nj05781a

NJC
View Article Online
View Journal | View Issue
PAPER
Stress degradation study of bortezomib: effect of
co-solvent, isolation and characterization of
degradation products by UHPLC-Q-TOF-MS/MS
and NMR and evaluation of the toxicity of the
degradation products†
Cite this: New J. Chem., 2021,
45, 8178
c
c
Suresh Udutha,ab Roshan M. Borkar,
G. Shankar,a T. Sony,a Aishwarya Jala,
E. Vamshi Krisna,d T. Kiran Kumar,e S. Misra,d S. Prabhakarab and R. Srinivas
*
ab
Bortezomib (BTZ) is a first-in-class, potent reversible inhibitor of proteasome used in the treatment of
multiple myeloma, the second most common hematological cancer. Stress degradation studies were
performed to investigate the inherent stability of the drug according to ICH recommended guidelines
Q1A (R2). Stress experiments were carried out in two ways using acetonitrile and methanol as co-
solvents under various conditions. A selective stability-indicating LC-MS method has been developed to
separate all degradation products of the drug on a Hibar-Purospher STAR, C18 (250 Â 4.6 mm, 5 mm)
column using a mobile phase consisting of 0.1% formic acid and acetonitrile in the gradient mode. BTZ
was found to undergo degradation under acidic, basic, neutral hydrolysis and oxidative conditions,
whereas it was stable under other conditions. Thirteen degradation products (DP-1–DP-13) were
identified using acetonitrile as a co-solvent. Additionally, three (DP-14–DP-16) degradation products
were found where methanol was used as a co-solvent. A total of 16 (DP-1–DP-16) degradation
products were characterized by liquid chromatography-tandem mass spectrometry (LC-ESI-Q-TOF/MS/
MS) and high-resolution mass spectrometry (HRMS). Major degradation products, DP-3, DP-6, DP-9,
DP-10, DP-11 and DP-12, formed under oxidation conditions were isolated using preparative HPLC and
characterized by 1D and 2D NMR experiments. Furthermore, in vitro cytotoxicity of isolated DPs was tested on
normal cell lines such as CHO-K1, HEK-293 and NRK-49F by MTT assays. This study revealed that they were
around 2–6 times less toxic as compared with the standard control of the drug and DP-10 showed relatively
more toxicity than other isolated DPs against rat kidney cells at 18.20 mM. In silico toxicity studies suggested
that BTZ and its DPs can be hepatotoxic and genotoxic resulting in severe toxicity.
Received 28th November 2020,
Accepted 28th March 2021
DOI: 10.1039/d0nj05781a
rsc.li/njc
stress conditions which are more severe than accelerated con-
ditions. The International Council for Harmonization of Tech-
1. Introduction
Stress degradation is a process that involves stability of drug nical Requirements for Pharmaceuticals for Human Use (ICH)
products and the drug substances are examined under various guidelines states the requirement of stability testing data to
understand how the quality of a drug molecule changes
with time under the influence of various environmental factors
a Analytical & Structural Chemistry Department, CSIR-Indian Institute of Chemical
such as light, pH, moisture, and heat. The drug stability
test guidelines Q1A (R2) issued by ICH1–4 require the stress
Technology, Hyderabad, 500 007, India. E-mail: srini@iict.res.in,
sragampeta@yahoo.co.in; Fax: +91-40-27193156; Tel: +91-40-27193122
stability studies to be done on a drug to establish its inherent
stability characteristics. Identification and characterization of
degradation products are useful to understand the degradation
mechanisms under various conditions such as hydrolysis,
oxidation, thermolysis and photolysis of drug substance or
drug products.1,5 Furthermore, degradation products can be
synthesized or isolated and subsequently their toxicity can be
evaluated with the help of in vivo and in vitro toxicity studies.
b Academy of Scientific and Innovative Research (AcSIR), Ghaziabad-201 002, India
c Department of Pharmaceutical Analysis, National Institute of Pharmaceutical
Education and Research (NIPER), Guwahati, Changsari, 781101, India
d Centre for Natural Products & Traditional knowledge, CSIR-Indian Institute of
Chemical Technology, Hyderabad, 500 007, India
e Applied Biology Division, CSIR-Indian Institute of Chemical Technology,
Hyderabad, 500007, India
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0nj05781a
8178
|
New J. Chem., 2021, 45, 8178–8191
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
Paper
NJC
This phenomenon helps in drug discovery and development acetonitrile and methanol as co-solvents which includes: (i)
research. performing stress experiments on drugs under various stress
Proteasome inhibitors are a new class of drugs containing conditions (hydrolysis, oxidation, photolysis and thermolysis).
boronic acid moieties, effective on a wide variety of tumors. (ii) Developing a LC-MS method for separation of the drug
Bortezomib (BTZ) is a dipeptide consisting of phenylalanine and its degradation products. (iii) Structural elucidation of
and leucine with boronic acid at the C-terminal instead of DPs by a fragmentation pathway (MS/MS) and accurate mass
a carboxylic acid. Bortezomib, [(1R)-3-methyl-1-[{(2S)-1-oxo-3- measurements (HRMS). (iv) Establishing mechanistic pathways
phenyl-2-[(pyrazinyl carbonyl amino)]propyl}amino]butyl] boro- for the formation of degradation products. (v) Isolation of
nic acid, is the first-in-class therapeutic proteasome inhibitor. major degradation products using preparative HPLC and char-
BTZ binds to proteasome via boronic acid and therefore, the acterization by NMR experiments. (vi) In addition, in vitro
presence of this moiety is necessary to achieve therapeutic toxicity evaluation of isolated DPs by cytotoxicity assays on
activity.6,7 BTZ is approved in the US mainly for the treatment normal cell lines.
of relapsed multiple myeloma where the disease is refractory to
conventional lines of therapy. Multiple myeloma is a B-cell
malignancy of plasma cells, representing the second most
common hematological malignancy. Approximately, 11 200
deaths per year are due to this disease, representing 2% of all
2. Experimental
2.1 Chemicals and reagents
Bortezomib (BTZ) was obtained as a gift sample from a local
cancer deaths. It is a malignancy of bone marrow in which
manufacturing company, Hyderabad, India. LC-MS Grade acet-
cancerous plasma cells grow out of control and create a tumor
onitrile (ACN) and methanol (MeOH) were purchased from
and recent studies indicate that it is active towards other non-
Merck, India. HPLC grade water was prepared by filtrating
Hodgkin lymphoma such as mantle cell lymphoma. Recently it
through a Millipore Milli-Q-plus system (Millipore, Milford,
MA, USA). All analytical grade reagents: formic acid, sodium
was found that use of BTZ provides a novel strategy to treat
patients with breast cancer osteolytic lesions, and represents an
hydroxide, hydrochloric acid, and 30% (w/w) hydrogen peroxide
approach for protecting the entire skeleton from metastatic
bone disease, and even helps in bone formation.8
were purchased from SD Fine Chemicals Pvt. Ltd (Mumbai,
India).
A thorough literature search revealed that several HPLC and
UV visible spectrophotometric methods were reported for the
determination of BTZ in bulk and its formulations.9–11 A few
2.2 Instrumentation
validated LC-MS/MS methods have been reported on the assay
2.2.1 UPLC-Q-TOF mass spectrometer. The LC-MS analysis
of BTZ in human plasma with various sample preparation was carried out using an Agilent 1290 infinity series LC system
procedures.12,13 Pekol et al.14 identified BTZ metabolites in (Agilent Technologies, USA) coupled to a quadrupole time-of-
human plasma by LC-MS/MS and NMR. Hsieh et al.15 identified flight (Q-TOF) mass spectrometer (Q-TOF LC/MS 6545 series-
two potential biomarkers (apolipoprotein C-I and apolipopro- G6545A, Agilent Technologies, USA) equipped with an electro-
tein C-I1) by mass spectral analyses and confirmed using spray ionization (ESI) source. The UPLC system was configured
authentic protein standards based on MALDI-TOF-MS/MS with an auto sampler (G7129B), a quaternary pump (G7104A), a
sequencing of proteolytic peptides. Byrn16 analyzed commer- diode array detector (G7117A),
a column compartment
cially available VELCADEs and bortenat products and reported (G7116B) and a degasser. Data acquisition was under the
the differences in assays of active agents and impurities using control of Mass Hunter workstation software.
different analytical techniques. Bolognese et al.17 characterized
2.2.2 Preparative HPLC system. Isolation of major degra-
two DPs by NMR. By considering safety aspects, ICH guidelines dation products of bortezomib was performed by using a
Q3A and Q3B recommend for performing toxicity evaluation preparative HPLC system (GILSON, USA). The major compo-
of degradation products which includes identification, charac- nents of this instrument are a binary pump (331 and 332), a
terization, and qualification thresholds.18,19 A few of the toxi- liquid handler (GX-241), a degasser, a column compartment,
city studies were reported on degradation products that can be and a diode array detector (157 UV-VIS). Data acquisition was
found in many pharmaceutically active substances and under the control of TRILUTION LC software.
products.20–22 Recently, we have reported the applications of
2.2.3 Nuclear magnetic resonance spectroscopy. To eluci-
LC-MS/MS experiments and accurate mass measurements for date the exact structure of isolated major degradation products,
the structural characterization of unknown degradation pro- different types of NMR experiments (1H, 13C and 2D) were
ducts of active pharmaceutical drugs formed under various performed on a 500 MHz NMR (AVANCE III HD-500, Bruker,
stress conditions, like sumatriptan succinate, omeprazole, Billerica, Massachusetts, United States) spectrometer. NMR
rabeprazole and lansoprazole.23–26
spectra were recorded in deuterated solvent (CDCl3) and
As there are no reports available in the literature on the chemical shift values were measured in parts-per-million
degradation behavior of the drug according to the ICH recom- (PPM) on a horizontal d scale. Tetra methyl silane (TMS) is
mended guidelines under various stress conditions, we have used as an internal standard and adjusted to 0 ppm on the
undertaken this study. Hence, the purpose of the present d scale. The data acquisition and processing of NMR spectra
work is to investigate the degradation behavior of drugs using were done using Top spin software (3.2 version).
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 8178–8191 | 8179

View Article Online
NJC
Paper
2.3 Stressed degradation conditions
2.6 Preparative HPLC for isolation of major degradation
products
To understand the drug degradation behavior, bortezomib was
exposed to various stress conditions according to ICH guide- Isolation of major oxidation degradation products (DP-3, DP-6,
lines Q1A (R2). The stress experiments were performed in two DP-9, DP-10, DP-11 and DP-12) was carried out in a reverse-
ways by taking acetonitrile and methanol as co-solvents. All phase mode by preparative HPLC. All these degradation
stress reactions were maintained at 1.0 mg mLÀ1 concentration products were separated on preparative column Luna
of the drug by dissolving it with a combination of acetonitrile/ phenomenex-C8 (150 Â 21.2 mm, 10 mm) using the mobile
methanol and water. Bortezomib was subjected to stress acidic phase consisting of 0.1% formic acid in water and acetonitrile
hydrolytic conditions (1 N HCl) at room temperature for 24 h. (100%). The gradient solvent program was set as follows (T min/
However, basic and neutral hydrolysis was conducted by reflux- % proportion of solvent B): 0/5, 2/5, 23/60, 28/95, 33/95, and
ing the drug with 1 N NaOH and water at 80 1C for 12 h and 35/5, and flow rate at 10.0 mL minÀ1. 50 mg of crude sample
24 h, respectively. An oxidative degradation was carried out for (obtained from reaction mixture) was loaded on a prep
3 days under dark conditions at room temperature by dissol- column and fractions were collected at different time intervals
ving the drug in 3% H2O2. For thermal degradation, the solid with respect to peak elution order. Chromatograms were mon-
form of the drug was spread on a 1.0 mm thick Petri dish and itored at a wavelength of 254 nm. The collected fractions were
kept at 90 1C for 3 days. Photolytic studies were performed in evaporated using a lyophilizer to get the pure compound. The
the photo stability chamber by exposing the drug in solid and isolated degradation products were subjected to various NMR
solution forms to 200 W h mÀ2 UV-light and 1.2 Â 106 lux h experiments.
visible light.27 All the stress samples were withdrawn and stored
2.7 Cytotoxicity test (MTT assay)
in a refrigerator at 4 1C until the analysis.
To evaluate the preliminary toxicity potential of the test com-
pounds, cytotoxicity assay (MTT) was carried out as per the
procedure published earlier (Javvaji et al., 2018).28 For compar-
ison of the cytotoxicity efficacy of the test compounds (DP-3,
DP-6, and DP-9–12), bortezomib (BTZ) was used as the standard
control; and mitomycin C and doxorubicin were used as a
positive control. Three normal cell lines viz., CHO-K1, Chinese
hamster ovary cells (ATCCs CCL-61t); HEK-293, Homo sapiens
embryonic kidney cells (ATCCs CRL-1573t) and NRK-49F,
Rattus norvegicus kidney cells (ATCCs CRL-1570t) were
employed in this experiment. These cell lines were obtained
from the ATCC (Bethesda, MD, USA) and maintained in DMEM
supplemented with 10% FBS, 2 mM L-glutamine, 100 U mLÀ1
penicillin, and 100 mg mLÀ1 streptomycin at 37 1C in a 5% CO2
incubator. After seeding of cells in a 96 well culture plate, the
cells were allowed to attach properly. Test compounds of
different concentrations (from 1 to 100 mM) were added in
triplicates and incubated for 24 h. The cells were then incu-
bated with MTT (0.5 mg mLÀ1) for 3 h. To dissolve the insoluble
formazan crystals, 100 mL DMSO was added to each well and
finally, the absorbance of the plates was measured using a
Synergy H1 multi-mode plate reader, USA.
2.4 Sample preparation
The collected stress samples at different time intervals were
diluted to 10 times with a solution of acetonitrile and water
(1 : 1). All solutions were filtered through a 0.22 mm pore size
membrane syringe filter before LC/MS analysis.
2.5 Method development and optimization of LC-MS
conditions
A Hibar-purospher STAR, C18 column (250 Â 4.6 mm, 5 mm)
(Merck Lichro, Switzerland) was found to be a suitable
column to separate the drug and its degradation products.
Initial trials were performed on this column with a combi-
nation of an organic modifier (ACN/MeOH) and ammonium
formate/formic acid as a buffer in different proportions.
Finally, good peak symmetry and resolution were achieved
with a mobile phase consisting of 0.1% formic acid (A) and
acetonitrile (B) in a gradient elution method. A linear gra-
dient program was set as follows (T min/% of solvent B): 0/10,
2/10, 17/75, 22/80, 25/95, 29/95 and 30/10 at a flow rate of
0.8 mL minÀ1. Chromatograms were monitored with a sui-
table wavelength of 254 nm. The column temperature was
maintained at 30 1C and the injection volume was 5 mL. The
typical operating source conditions for MS scan were opti-
mized in a positive mode of ESI as follows: capillary voltage
3. Results and discussion
3.1 Degradation behavior of bortezomib (BTZ)
3500 V, gas temperature 325 1C, nebulizer 40 psig, drying The degradation behavior of BTZ was studied under various
gas 8.0 l minÀ1, sheath gas flow 11.0 l minÀ1, nozzle voltage stress conditions by using LC-MS/MS. The drug undergoes
1000 V, fragmentor 80 V and skimmer voltage 65 V. For the degradation under acidic, basic, neutral and oxidation condi-
full scan MS mode, all the spectra were recorded at an tions whereas it was found to be stable under photolytic
average of 20–25 scans in the mass range of m/z 100–1000 and thermal conditions. Thirteen degradation (DP-1–DP-13)
under identical experimental conditions. For MS/MS experi- products were identified when acetonitrile was used as a
ments, keeping MS1 static, the precursor ion of interest was co-solvent. In addition, three more degradation products
selected using a quadrupole analyzer and the product ions (DP-14–DP-16) were observed where methanol was used as a
were analyzed using a time-of-flight (TOF) analyzer.
co-solvent. A total of 16 DPs were identified and their total ion
8180
|
New J. Chem., 2021, 45, 8178–8191
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
Paper
NJC
Fig. 2 UPLC/ESI/MS total ion chromatogram of the drug under oxidation
(a) using acetonitrile as a co-solvent, photolytic (b) and thermal (c)
conditions.
bortezomib were subjected to UV and fluorescent light for
one week.
3.1.4 Thermal degradation. The total ion chromatogram of
BTZ (Fig. 2c) showed the absence of degradation products
under thermal conditions where it was heated to 90 1C for
3 days.
Fig. 1 UPLC/ESI/MS total ion chromatogram of bortezomib (a) drug
under acidic (b), basic (c) and neutral (d) conditions using acetonitrile as
a co-solvent.
chromatograms under various stress conditions are presented
in Fig. 1–3. A few of the degradation products, DP-1, DP-2, DP-4,
DP-5, DP-9 and DP-10, were formed in both cases. Based on the
m/z values of the protonated degradation products and their
3.2 Characterization of BTZ and its degradation products by
LC-MS/MS experiments
3.2.1 MS/MS of BTZ ([M + H–H2O]+, m/z 367). The positive
MS/MS fragmentation combined with accurate mass measure- ion ESI-HRMS spectrum of BTZ shows an abundant [M +
ments, most plausible structures have been proposed for the H–H2O]+ ion at m/z 367.1945 with an elemental composition
of C19H24BN4O3 and a moderately abundant [M + Na]+ ion at
+
DPs (Scheme 1 and Table 1).
3.1.1 Hydrolysis. LC-ESI/HRMS analysis of BTZ showed m/z 407.1859. The LC-ESI-MS/MS spectrum of BTZ ([M + H–
that the formation of 3 DPs (DP-1, DP-2 and DP-3), 1 DP H2O]+, m/z 367) showed an abundant product ion at m/z
(DP-2) and 2 DPs (DP-5 and DP-5) in acidic (Fig. 1b), basic 226 (loss of C6H10BNO) and low abundance ions at m/z 349
(Fig. 1c) and neutral hydrolysis (Fig. 1d), respectively, where (loss of H2O), m/z 276 (loss of C7H7 radical from m/z 367), m/z
acetonitrile was used as a co-solvent. In the case of methanol as 253 (loss of C5H9BO), m/z 244 (loss of C5H5N3O), m/z 208 (loss of
a co-solvent, 2 DPs (DP-2 and DP-14), 2 DPs (DP-1 and DP-2) and H2O from m/z 226), m/z 199 (loss of C2H2N2 from m/z 253), m/z
2 DPs (DP-4 and DP-5) were observed in acidic (Fig. 3a), basic 188 (loss of C4H8 from m/z 244), m/z107 (loss of C9H10N2), m/z
(Fig. 3b) and neutral hydrolysis (Fig. 3c) respectively.
91 (loss of C5H4N2O) and m/z 79 (loss of carbonyl from m/z 107)
3.1.2 Oxidation. A total of nine degradation products (Scheme 2 and Fig. 4a). The product ions at m/z 79 and m/z
(DP-3, DP-6, DP-7, DP-8, DP-9, DP-10, DP-11, DP-12, and DP- 91 are diagnostic peaks for the presence of a pyrazine ring and
13) were found in 3% H2O2 solution where acetonitrile used as benzyl group in the drug molecule respectively. It can be noted
a co-solvent (Fig. 2a). Four degradation products (DP-9, DP-10, that the product ion at m/z 226 (Scheme 2) is also a key
DP-15 and DP-16) were observed in the case of methanol as a structure that helps in structural determination of degradation
co-solvent (Fig. 3d).
products derived from BTZ. The elemental compositions of
3.1.3 Photolytic degradation. No degradation products all these ions were confirmed by HRMS data (Table S1, see
were observed (Fig. 2b) when the solid form and solution of the ESI†).
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 8178–8191 | 8181

View Article Online
NJC
Paper
C9H11N2O) (Scheme S2 and Fig. 4c, see the ESI†). The product
ions at m/z 142 and m/z 120 are characteristic peaks, whereas
a neutral loss of 17 Da represents the presence of a free
amine group in degradation product DP-2. The elemental
compositions of protonated DP-2 and its products ions were
confirmed by accurate mass measurements. These data were
clearly compatible with the structure of DP-2 and named
2-amino-N-(1-hydroxy-4-methylborolan-2-yl)-3-phenylpropanamide
(Table S2, see the ESI†).
3.2.4 DP-3 ([M + H]+, m/z 272). The LC-ESI-HRMS spectrum
of DP-3 (Rt = 9.9 min.) shows an abundant ion [M + H]+ at m/z
272.1023 with an elemental composition C14H14N3O3+ (Fig. 1b).
The MS/MS spectrum of protonated DP-3 displayed an abun-
dant product ion at m/z 226 (loss of formic acid) and low
abundance ions at m/z 208 (loss of water), m/z 146 (loss of
pyrazine), m/z 120 (loss of C5H2N2O), m/z 109 (loss of C9H9NO2),
+
m/z 91 (loss of C7H7N3O3 ) and m/z 81 (loss of a carbonyl group)
(Scheme S3 and Fig. 4d, see the ESI†). A mass difference of 46
Da between DP-3 and its corresponding product ion at m/z 226
indicates the presence of a free carboxylic group (–COOH).
However, the product ion at m/z 81 is diagnostic for a pyrazine
ring in DP-3. The proposed structures of all these fragment ions
were confirmed using HRMS data. Based on these data, DP-3 was
predicted as 3-phenyl-2-(pyrazine-2-carboxamido)propanoic acid
(Table S2, see the ESI†).
3.2.5 DP-4 and DP-5 ([M + Na]+, m/z 379). Degradation
products DP-4 and DP-5 are stereoisomers and eluted their
retention times at 13.12 min and 13.62 min, respectively
(Fig. 1d). These two degradation products have the same
elemental composition C19H24N4O3Na+ and exhibit similar
fragmentation pathways. The MS/MS spectra of sodiated degra-
dation products (m/z 379) showed an intense product ion at m/z
293 (loss of C5H8 from m/z 361) and low abundance product
ions at m/z 361 (loss of water), m/z 276 (loss of ammonia from
m/z 293), m/z 248 (loss of CH3NO from m/z 293), m/z 226 (loss of
Fig. 3 UPLC/ESI/MS total ion chromatogram of the drug under acidic (a),
basic (b), neutral (c) and oxidation (d) conditions using methanol as a co-
solvent.
3.2.2 DP-1 ([M + H]+, m/z 166). The acidic degradation C6H11NO from m/z 361), m/z 208 (loss of water from m/z 226)
+
product DP-1 with an elemental composition C9H12NO2 was and m/z 142 (loss of C5H2N2O from m/z 248). The absence
eluted at Rt = 2.91 min (Fig. 1b). The ESI-MS/MS spectrum of of a boron pattern in the ESI-HRMS spectra of these two
[M + H]+ ions (m/z 166) of DP-1 shows an intense product ion at degradation products (DP-4 and DP-5) and loss of water indicates
m/z 120 (loss of formic acid) and low abundance ions at m/z the presence of an ‘OH’ group in these compounds (Scheme S4 and
149 (loss of ammonia), m/z 131 (loss of water), m/z 103 (loss of Fig. 5a, see the ESI†). The elemental compositions of all these
ammonia), m/z 93 (loss of C2H3NO2) and m/z 77(loss of fragment ions were confirmed by accurate mass measurements
methane) (Scheme S1 and Fig. 4b, see the ESI†). A simultaneous (Table S2, see the ESI†). Based on these data, stereoisomers of DP-4
loss of 46 Da and 17 Da from protonated DP-1 is indicative of and DP-5 were named N-(1-((1-hydroxy-3-methylbutyl)amino)-1-oxo-
the presence of carboxylic and amine groups. The proposed 3-phenylpropan-2-yl)pyrazine-2-carboxamide.
structures of all of these ions were supported with accurate
mass measurements. Based on these data, the structure of DP-1 protonated DP-6 (Rt = 7.9 min) with elemental composition
3.2.6 DP-6 ([M + H]+, m/z 271). The MS/MS spectrum of
+
was identified as 2-amino-3-phenylpropanoic acid (Table S1,
C
14H15N4O2 shows an intense product ion at m/z 226
see the ESI†).
(loss of CH3NO) and low abundance ions at m/z 254 (loss of
3.2.3 DP-2 ([M + H]+, m/z 261). DP-2 (Rt = 6.2 min), formed ammonia), m/z 208 (loss of water from m/z 226), m/z 199 (loss of
in acidic and basic hydrolysis, shows [M + H]+ at m/z 261.1765 HCN from m/z 226), m/z 107 (loss of C9H12N2O), m/z 103 (loss of
with an elemental composition of C14H22BN2O2+ (Fig. 1b and c). C5H5N3O from m/z 226), and m/z 79 (loss of C9H7N from m/z
Its MS/MS spectrum yields highly abundant product ions at m/z 208) (Scheme S5 and Fig. 5b, see the ESI†). Loss of a formamide
142 (loss of C8H9N) and low abundant ions at m/z 244 (loss of group (45 Da) from [M + H]+ giving rise to the product ion at m/z
ammonia), m/z 188 (loss of C4H11N), m/z 170 (loss of benzyl 226 supports the proposed structure for DP-6. The elemental
radical), m/z 120 (loss of C6H12BNO2), and m/z 98 (loss of compositions of DP-6 and its products ions were confirmed by
8182
|
New J. Chem., 2021, 45, 8178–8191
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
Paper
NJC
Scheme 1 Proposed structures of protonated degradation products of bortezomib.
Table 1 UPLC/ESI/HRMS data of the drug and its degradation products
Degradation products
Retention time (min)
Proposed molecular formula [M + H]+
Calculated (m/z)
Observed (m/z)
Error (ppm)
+
BTZ
11.52
2.91
6.23
C19H24BN4O3
C9H12NO2
367.19312
166.08621
261.17681
272.10235
379.17402
379.17402
271.11892
287.11387
287.11387
395.16892
395.16893
411.16381
411.16381
355.17647
286.11862
393.18971
393.18971
367.19341
166.08606
261.17704
272.10231
379.17193
379.17193
271.11769
287.11358
287.11358
395.16702
395.16702
411.1653
À0.78
0.90
À0.88
0.14
5.51
5.51
4.53
2.41
2.41
4.8
+
DP-1
DP-2
DP-3
DP-4
DP-5
DP-6
DP-7
DP-8
DP-9
DP-10
DP-11
DP-12
DP-13
DP-14
DP-15
DP-16
+
C14H22BN2O2
+
9.92
C14H14N3O3
13.12
13.64
7.97
6.32
7.02
14.31
15.32
12.8
13.7
16.40
13.9
C19H24N4O3Na+
C19H24N4O3Na+
+
C14H15N4O2+
C14H15N4O3+
C14H15N4O3
C19H24N4O4Na+
C19H24N4O4Na+
C19H24N4O5 Na+
C19H24N4O5+Na+
C19H23N4O3+
4.8
À3.62
À3.62
0.1
1.56
2.12
2.12
411.1653
355.17646
286.11747
393.19014
393.19014
C15H16N3O3
17.2
17.5
C20H26N4O3Na+
C20H26N4O3Na+
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 8178–8191 | 8183

View Article Online
NJC
Paper
Scheme 2 Proposed fragmentation pathway of the protonated drug (m/z 367).
HRMS data (Table S3, see the ESI†). Based on these data DP-6 Fig. 5c, see the ESI†). The mass difference between DP-6 and
was named N-(1-amino-1-oxo-3-phenylpropan-2-yl)pyrazine-2- DP-7 or DP-8 is 16 Da, which suggests the addition of an oxygen
carboxamide.
atom to the degradation product (DP-6). Nitrogen atoms in the
3.2.7 DP-7 and DP-8 ([M + H]+, m/z 287). The oxidation pyrazine ring are highly active sites for N-oxide formation. The
degradation products, DP-7 (Rt = 6.32 min) and DP-8 (Rt = 7.02 product ions at m/z 118 and m/z 224 (Scheme S6, ESI†) are
+
min), have the same elemental composition of C14H15N4O3
diagnostic peaks for the formation of N-oxide in the pyrazine
(Fig. 2a) and possibly isomeric to each other or isomers. The ring. The elemental compositions of all these fragment ions
ESI-MS/MS spectra of DP-7 and DP-8 (m/z 287) yield highly were confirmed with accurate mass measurements. All these
abundant ions at m/z 242 (loss of a formamide group) and low data were compatible with the proposed structures for DP-7 and
abundance product ions at m/z 270 (loss of ammonia), m/z 224 DP-8 (Table S3, see the ESI†).
(loss of water from m/z 242), m/z 199 (loss of HCNO from m/z
3.2.8 DP-9 and DP-10 ([M + Na]+, m/z 395). DP-9 (Rt =
242), m/z 146 (loss of C4H4N2O from m/z 242), m/z 118 (loss of 14.31 min) and DP-10 (Rt = 15.32 min) are oxidation degradation
+
carbonyl from m/z 146), m/z 91 (loss of C7H5N3O3 from m/z 270), products with the same elemental composition C19H24N4NaO4
and m/z 52 (loss of C9H9NO from m/z 199) (Scheme S6 and (Fig. 2a). The MS/MS spectrum of DP-9 and DP-10 (m/z 395) yields
8184
|
New J. Chem., 2021, 45, 8178–8191
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
Paper
NJC
Fig. 5 UPLC/ESI/MS/MS spectra of (a) DP-4/DP-5 [M + Na]+, (b) DP-6 [M +
H]+, (c) DP-7/DP-8 [M + H]+ and (d) DP-9/DP-10 [M + Na]+ at different
collision energies of 8 eV, 5 eV, 5 eV and 11 eV respectively.
Fig. 4 UPLC/ESI/MS/MS spectra of (a) BTZ [M + H–H2O]+, (b) DP-1 [M +
H]+, (c) DP-2 [M + H]+ and (d) DP-3 [M + H]+ at different collision energies
of 15 eV, 6 eV, 16 eV and 18 eV respectively.
highly abundant ions at m/z 321 (loss of C4H10O3) and low from m/z 309), m/z 264 (loss of formamide from m/z 309), m/z
abundance product ions at m/z 377 (loss of water), m/z 361 (loss 199 (loss of C2H2N2ONa radical from m/z 292), and m/z86 (loss of
of O from m/z 377), m/z 293 (loss of C5H8 from m/z 361), m/z 275 C14H10N3NaO3 from m/z 377) (Scheme S8 and Fig. 6a, see the ESI†).
(loss of water from m/z 293), m/z 248 (loss of HCN from m/z A mass difference of 32 Da between DP-11/DP-12 and DP-4/DP-5
275), m/z 226 (loss of NaCN from m/z 275), m/z 199 (loss of indicates di oxidation of degradation products DP-4/DP-5. The
C2H5N3Na radical from m/z 293), and m/z 142 (loss of C5H2N2O product ions at m/z 325, m/z 199 and m/z 86 are diagnostic peaks
from m/z 248) (Scheme S7 and Fig. 5d, see the ESI†). The mass which support the presence of two oxygen atoms in the pyrazine
difference between DP-9/DP-10 and DP-4 or DP-5 is 16 Da which ring. HRMS data confirm the elemental composition of DP-11/DP-
suggests that these two are mono oxidation products of DP-4 and 12 and its fragment ions (Table S4, see the ESI†).
DP-5. The product ions at m/z 321 and m/z 142 support the
3.2.10 DP-13 ([M+ H]+, m/z 355). DP-13 (Rt = 16.4 min) was
occurrence of N-oxidation in the pyrazine ring. The elemental also observed under oxidation conditions and its [M + H]+ ion
+
compositions of DP-9/DP-10 and its product ions were confirmed (m/z 355) has an elemental composition C19H23N4O3 (Fig. 2a).
by accurate mass measurements (Table S3, see the ESI†). Based on The MS/MS spectrum of protonated DP-13 (m/z 355) shows a
these data, most plausible structures of DP-9 and DP-10 were highly abundant ion at m/z 271 (loss of C5H8O) and low
proposed as given Scheme S7 (ESI†).
abundance product ions at m/z 254 (loss of ammonia from
3.2.9 DP-11 and DP-12 ([M + Na]+, m/z 411). DP-11 (Rt = m/z 271), m/z 226 (loss of formamide from m/z 271), m/z 208
12.8 min) and DP-12 (Rt = 13.7 min) are the oxidative degradation (loss of water from m/z 226), m/z 181 (loss of HCN from m/z
+
products with the same elemental composition C19H24N4NaO5
208), m/z 148 (loss of C5H5N3O from m/z 271), m/z 107 (loss
(Fig. 2a). The MS/MS spectrum of [M + Na]+ of DP-11 and DP-12 of C8H9N from m/z 226), and m/z 85 (loss of C14H14N4O2)
displayed an abundant product ion at m/z 337 (loss of C4H8 from (Scheme S9 and Fig. 6b, see the ESI†). The ion at m/z 85 is
m/z 393) and low abundance product ions at m/z 393 (loss of water), a diagnostic peak for the presence of a CO group in the
m/z 377 (loss of H2O2), m/z 325 (loss of C5H8 from m/z 393), m/z molecule. The elemental compositions of DP-13 and its frag-
309 (loss of C5H8 from m/z 377), m/z 292 (loss of ammonia ment ions were confirmed by accurate mass measurements
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 8178–8191 | 8185

View Article Online
NJC
Paper
They probably exist as stereoisomers and possess the same
+
elemental composition C20H26N4NaO3 . The MS/MS spectrum
of the sodiated degradation product (m/z 393) yielded an
intense product ion at m/z 248 (loss of formamide from m/z
293) and low abundance product ions at m/z 361(loss of
methanol), m/z 293 (loss of C5H8 from m/z 361), m/z 276 (loss
of ammonia from m/z 293), m/z 226 (loss of C6H10NNaO from
m/z 361), m/z 208 (loss of water from m/z 226), m/z 108 (loss of
C
14H11N3O2 from m/z 361). m/z 86 (loss of C14H10N3 NaO2 from
m/z 361), and m/z 69 (loss of ammonia from m/z 86) (Scheme
S11 and Fig. 6d, see the ESI†). A mass difference of 14 Da
between DP-15/DP-16 (m/z 393) and DP-4/DP-5 (m/z 379)
indicates O-methylation (–CH2–) of DP-4 and DP-5 respectively.
Furthermore, the product ion at m/z 361 was also formed by
loss of methanol suggesting the presence of methoxy groups in
DP-15 and DP-16. The fragmentation pathways of both proto-
nated DP-15 and DP-16 were found to be similar. Based on
these data, structures for the isomeric DP-15 and DP-16 were
proposed and the elemental compositions of all these fragment
ions were confirmed by HRMS data (Table S5, see the ESI†).
3.3 Mechanistic pathway for the formation of BTZ
degradation products
3.3.1 Hydrolytic conditions. The formation of DP-1 can be
explained by acid catalyzed hydrolysis of two amide groups of
BTZ. The reactions may involve protonation of amide carbonyl
followed by an attack of H2O on the carbonyl carbons as shown
in Scheme S12 (ESI†). Similarly, DP-2, DP-3 and DP-14 may also
proceed with amide hydrolysis reactions by involving only one
amide group. Furthermore, DP-2 and DP-14 may involve dehy-
dration and esterification reactions, respectively. Probable
mechanistic pathways for the formation of DP-1, DP-2, DP-3
Fig. 6 UPLC/ESI/MS/MS spectra of (a) DP-11/DP-12 [M + Na]+, (b) DP-13
[M + H]+, (c) DP-14 [M + H]+ and (d) DP-15/DP-16 [M + Na]+ at different
collision energies of 13 eV, 5 eV, 7 eV and 23 eV respectively.
(Table S4, see the ESI†). All these data are highly compatible with and DP-14 are shown in Schemes S12 and S15 (see the ESI†).
the structure of DP13 and named N-(1-(3-methylbutanamido)-1-
oxo-3-phenylpropan-2-yl)pyrazine-2-carboxamide.
Formation of DP-4 and DP-5 under neutral conditions can
be explained by the nucleophilic substitution reaction (SN1).
3.2.11 DP-14 ([M + H]+, m/z 286). DP-14 (Rt = 13.9 min) was The mechanism may proceed with the formation of a carboca-
formed under acidic conditions using methanol as a co-solvent tion as an intermediate by the elimination of a boron group
and its protonated molecule has an elemental composition (B(OH)2). The solvent water itself can act as a nucleophile and
C
15H16N3O3+ (Fig. 3a). The MS/MS spectrum of protonated DP- attack the electrophilic carbocation to form DP-4 and DP-5
14 (m/z 286) displayed a highly abundant ion at m/z 226 (loss (Scheme S13, see the ESI†).
of C2H4O2) and low abundance product ions at m/z 254 (loss
3.3.2 Oxidative conditions. DP-9 and DP-10 were
of methanol), m/z 208 (loss of water from m/z 226), m/z 109 probably formed from DP-4 and DP-5 respectively by a
(loss of C10H11NO2), m/z 91 (loss of C8H9N3O3), and m/z 81 (loss N-oxidation pyrazine ring.29,30 The mechanism may involve
of carbonyl from m/z 109) (Scheme S10 and Fig. 6c, see the an electrophilic attack of OH+ (from H2O2) on the nitrogen
ESI†). A mass difference of 14 Da between DP-14 (m/z 286) and atom of the pyrazine ring followed by abstraction of a proton
DP-3 (m/z 272) indicates methylation (–CH2–) on the carboxylic with OHÀ (Scheme S14, ESI†). Dealkylation of DP-4 and DP-5
group of DP-3. The product ions at m/z 254 and m/z 226 are can lead to the formation of DP-6, which on oxidation may lead
characteristic peaks, formed by loss of methanol and methyl to DP-7 and DP-8. DP-9/DP-10 and DP-4/DP-5 on further oxida-
formate groups, respectively. All these fragment ions were tion can give rise to DP-11/DP-12 and DP-13 respectively. DP-4
confirmed by accurate mass measurements. Based on all these and DP-5 can also produce DP-15 and DP-16 respectively
data, DP-14 was identified as methyl 3-phenyl-2-(pyrazine-2- when methanol was used as a co-solvent. The mechanism
carboxamido)propionate (Table S4, see the ESI†).
may involve loss of water followed by the O-methylation reac-
3.2.12 DP-15 and DP-16 ([M + Na]+, m/z 393). DP-15 tion. Mechanistic pathways for the formation of all oxidation
(Rt = 17.2 min) and DP-16 (Rt = 17.5 min) were formed under degradation products are illustrated in Schemes S13–S15 (see
oxidation conditions using methanol as a co-solvent (Fig. 3d). the ESI†).
8186
|
New J. Chem., 2021, 45, 8178–8191
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
Paper
NJC
Fig. 7 Schematic diagram of chemical structures of BTZ, DP-3, DP-6, DP-9, DP-10, DP-11, and DP-12 with number depiction.
3.4 Structural characterization of DP-3, DP-6, DP-9, DP-10, DP- The proton (1H) and carbon (13C) chemical shifts of these com-
11 and DP-12 by 1D and 2D NMR experiments
pounds are presented in Tables 2 and 3.
The 1H NMR spectrum of BTZ (Fig. S1, see the ESI†)
was compatible with its structure, and supports that the
degradation study proceeded with the pure form of the drug
(BTZ). The 13C and DEPT135 spectra revealed that it has 17
different carbons, two secondary (–CH2–) carbons at C-11 and
The structural characterization of BTZ and isolated degradation
products, DP-9, DP-10, DP-11 and DP-12 (Fig. 7), was carried
out using various NMR experiments such as 1H, 13C, DEPT135
(Distortionless Enhancement by Polarization Transfer), COSY (COr-
relation SpectroscopY), HSQC (Heteronuclear Single Quantum Cor-
relation) and HMBC (Heteronuclear Multiple Bond Correlation). C-12, and four quaternary (C) carbons at C-1, C-7, C-12 and C-18
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 8178–8191 | 8187

View Article Online
NJC
Paper
Table 2 NMR data of BTZ, DP-9 and DP-10
BTZ
DP-9
13C ppm Atom
DP-10
13C ppm Atom
Atom
number
Type of
atom
Type of
atom
Type of
atom
13C ppm
(d)
1H ppm (d)
(d)
number
1H ppm (d) (d)
number
1H ppm (d)
1
2
4
5
7
9
10
11
C
—
143.72
144.34
142.70
147.61
163.05
—
1
2
4
5
7
9
10
11
C
—
143.63
142.81
147.79
144.40
163.51
—
1
2
4
5
7
9
10
11
C
—
143.59
144.30
147.73
142.79
163.58
—
CH
CH
CH
C
NH
CH
CH2
9.28(s)
8.44(d)
8.69(d)
—
8.32(d)
4.95(m)
3.17–3.26(m)
CH
CH
CH
C
NH
CH
CH2
8.52(s)
8.78(d)
9.33(d)
—
8.37(d)
4.83(m)
f1-3.26(m),
f2-3.22(m)
—
CH
CH
CH
C
NH
CH
CH2
9.27(s)
8.74(d)
8.51(d)
—
8.40(d)
4.77(m)
f1-3.23(m),
f2-3.21(m)
—
52.62
38.14
55.44
38.26
55.71
38.24
12
C
—
135.72 12
C
136.18 12
C
136.00
13, 17
14, 16
15
18
20
21
22
CH
CH
CH
C
NH
CH
CH2
7.19–7.29(m)
7.19–7.29(m)
7.19–7.29(m)
—
6.87(bs)
2.93(m)
c1: 1.27(m),
c2: 1.42–1.49(m)
1.42-1.49(m)
0.82(d)
128.82 13, 17
129.41 14, 16
127.28 15
CH
CH
CH
C
NH
CH
CH2
7.32–7.38(m) 128.87 13, 17
7.32–7.38(m) 129.35 14, 16
7.32–7.38(m) 127.45 15
CH
CH
CH
C
NH
CH
CH2
7.22–7.30(m)
7.22–7.30(m)
7.22–7.30(m)
—
6.54(d)
5.61(m)
c1: 1.17(m),
c2: 1.43–1.49(m)
1.43–1.49 (m)
0.85(m)
128.81
129.27
127.25
171.09
—
84.08
40.37
172.71 18
—
6.27(s)
5.66(m)
c1: 1.28(m), 40.46
c2: 1.54(m)
170.85 18
—
43.51
39.89
20
21
22
—
20
21
22
84.17
23
24, 25
27, 28
CH
CH3
OH
25.89
22.69
—
23
24, 25
26
CH
CH3
OH
1.57(m)
0.86(d)
8.52(s)
24.55
22.42
—
23
24, 25
26
CH
CH3
OH
24.39
22.45
—
1.81(bs)
9.67(bs)
(Fig. S2 and S3, ESI†). Furthermore, the assigned chemical shift replaced with an electron withdrawing group (–OH). The proton
values (1H and 13C) were confirmed with COSY and HSQC. chemical shift value at the C-5 atom in DP-9 was shifted
1H–1H COSY showed that the protons couple to each other, downfield about B0.64 ppm, as compared with BTZ, revealing
which are shown clearly in Fig. S4 (ESI†). The HSQC spectrum that the oxygen atom must be present at the N-6 position in the
(Fig. S5, ESI†) shows that it contains 11 CH groups (C-2, C-4, form of N-oxide in the pyrazine ring. However, the HRMS and
C-5, C-10, C-13, C-14, C-15, C-16, C-17, C-21, and C-23), 2 CH2 MS/MS data suggested that presence of an oxygen atom in the
groups (C-11 and C-22) and 2 CH3 groups (C-24 and C-25).
pyrazine ring (DP-9), but did not reveal its exact position.
NMR studies of DP-9 and DP-10 are similar to that of BTZ, Similarly, a down field shift (B0.75 ppm) and up field
except at C-2, C-5 and C-21 positions (Fig. S6, S7, S12 and S13, shift (B0.82 ppm) of proton resonances at C-2 and C-5 atoms
ESI†). The chemical shift values of the proton and carbon at the in DP-10, respectively, as compared with DP-9, indicate the
C-21 position in DP-9 and DP-10 were shifted downfield about presence of an oxygen atom at the N-3 position. A down
B2.73 ppm and B40.66 ppm respectively, as compared with field shift of proton resonance by B1.15 ppm at O-26 atom
BTZ, suggesting that the boron group (B(OH)2) might be (or the OH group) in DP-10, as compared with DP-9, and the
Table 3 NMR data of DP-11 and DP-12
DP-11
DP-12
Atom number Type of atom 1H ppm (d)
13C ppm (d) Atom number Type of atom 1H ppm (d)
13C ppm (d)
1
2
4
5
7
9
C
—
8.76(s)
8.18(d)
8.40(d)
—
8.37(d)
4.75(q)
f1-3.26(dd), f2-3.19(dd)
—
7.27–7.37(m)
7.27–7.37(m)
7.27–7.37(m)
—
6.08(d)
5.61(q)
c1: 1.26(m), c2: 1.50–1.54(m) 40.51
1.59–1.66(m)
0.88(d)
148.78
133.87
136.04
145.92
161.33
—
55.55
38.44
135.81
129.00
129.43
127.55
170.43
—
1
2
4
5
7
9
C
—
8.73(s)
8.18(d)
8.41(d)
—
8.41(d)
4.76(q)
f1-3.21(m), f2-3.15(m)
—
7.24–7.31(m)
7.24–7.31(m)
7.24–7.31(m)
—
6.43(d)
5.61(q)
c1: 1.16(m), c2: 1.43–1.51(m) 40.35
1.43–1.51(m)
0.87(d)
148.69
133.83
135.86
146.02
161.28
CH
CH
CH
C
NH
CH
CH2
C
CH
CH
CH
C
NH
CH
CH2
CH
CH3
OH
CH
CH
CH
C
NH
CH
CH2
C
CH
CH
CH
C
NH
CH
CH2
CH
CH3
OH
10
11
12
13, 17
14, 16
15
18
20
21
22
23
24, 25
26
10
11
12
13, 17
14, 16
15
18
20
21
22
23
24, 25
26
55.67
38.42
135.78
128.86
129.25
127.34
170.77
—
84.13
83.96
24.54
22.41
—
24.40
22.46
—
8.29(s)
9.66(s)
8188
|
New J. Chem., 2021, 45, 8178–8191
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
Paper
NJC
Table 4 In vitro toxicity results on normal cell lines
IC50 values (mM)
configuration of R or S, vice versa. Furthermore, DEPT 135 has
shown the presence of two –CH2– groups at C-11 and C-22, and
four quaternary (C) carbons at C-1, C-7, C-12 and C-18 in DP-11
and DP-12 (Fig. S20 and S26, ESI†). The proton connectivity in
COSY, HSQC and HMBC spectra of DP-11 and DP-12 is also the
same as that in BTZ (Fig. S21–S23 and S27–S29, ESI†). These
NMR data also support the presence of two oxygen atoms in the
pyrazine ring at the N-3 and N-6 atoms of DP-11 and DP-12.
(However, HRMS and MS/MS data also support the presence of
two oxygen atoms in the pyrazine ring of DP-11 and DP-12.)
In addition to these DPs, another two isolated DPs, DP-3
and DP-6, were also confirmed with 1H-NMR experiments
(Fig. S30 and S31, ESI†) and chemical shift values (d in ppm)
S. no. Test compound CHO-K1
HEK
NRK 49F
1
2
3
4
5
6
7
BTZ
9.60 Æ 0.10
49.88 Æ 1.8
10.96 Æ 0.44
50.50 Æ 2.04
9.495 Æ 0.05
72.88 Æ 0.40
32.86 Æ 0.03
29.06 Æ 0.79
18.20 Æ 0.17
32.96 Æ 0.61
24.34 Æ 0.80
—
DP-3
DP-6
DP-9
DP-10
DP-11
DP-12
53.67 Æ 0.35 39.89 Æ 1.95
44.23 Æ 1.29 28.91 Æ 1.4
32.80 Æ 0.47 29.62 Æ 2.1
20.50 Æ 0.43 54.81 Æ 0.07
42.19 Æ 0.97 44.29 Æ 0.81
Mitomycin C
Doxorubicin
13.1 Æ 0.68
—
—
0.010 Æ 0.001
—
presence of a chiral center at the C-21 atom in DP-9 and DP-10, are as follows:
indicates that the OH group in DP-10 might be present in the
DP-3 (d in ppm, CDCl3): 9.32 (s, 1H), 8.69 (s, 1H), 8.45
paramagnetic region compared to the OH group in DP-9, and (s, 1H), 8.21 (s, 1H), 7.20–7.25 (m, 5H), 5.01 (s, 1H), 3.22–3.35
also suggests that the OH group might be placed above or (m, 2H).
below the plane of proposed structures. Hence, DP-9 and DP-10
DP-6 (d in ppm, CDCl3): 9.35 (s, 1H), 8.75 (d, 1H), 8.54
might have a configuration of R or S, vice versa. The proton (d, 1H), 8.34 (d, 1H), 7.29 (m, 5H), 5.75 (s, 1H), 5.43 (s, 1H),
couplings in DP-9 and DP-10 are also similar to that in standard 4.83–4.88 (m, 1H), 3.17–3.28 (m, 2H).
drug BTZ, which were supported by 1H–1H COSY (Fig. S9
and S15, ESI†). DEPT135, HSQC and HMBC experiments of
DP-9 and DP-10 revealed the skeletal information of primary
3.5 In vitro toxicity assay
(CH3), secondary (CH2), tertiary (CH) and quaternary (C) car- The in vitro cytotoxicity results showed that BTZ was found to
bons, which are assigned in Fig. S8, S10, S11, S14, S16 and be more toxic than the degradation products (DP-3, DP-6, DP-9,
S17 (ESI†). These NMR data confirm the predicted structures of DP-10, DP-11 and DP-12). BTZ induced 50% cell death in CHO-
DP-9 and DP-10, and also differentiate them structurally and K1, HEK and NRK-49F at 9.60 mM, 10.96 mM and 9.49 mM,
configurationally.
respectively, whereas, the all test compounds were found to
Similarly, 1H and 13C chemical shift values of DP-11 and be with around 2–6 times lesser toxicity. Among all the DPs,
DP-12 (Fig. S18, S19, S24 and S25, ESI†) are very similar to DP-10 showed relatively more toxicity against rat kidney cells at
each other, except at the O-26 atom (or OH group). A down field 18.20 mM compared to other test compounds. Hence, this study
shift of proton resonance by B1.37 ppm at the O-26 atom revealed that all the test compounds are safer toward normal
(or OH group) in DP-12, as compared with DP-11, revealed that cell lines, which will be further useful for pharmacological
DP-11 and DP-12 are configurationally different and have a studies in detail (Table 4).
Table 5 Toxicity prediction of bortezomib and its degradation products
Bortezomib and its degradation products
Classification
Toxicity class
Target
Bortezomib DP-1
III III
DP-2
DP-3
DP-4
DP-5
DP-6
DP-7
III
III
III
III
III
III
Predicted LD50 Acute oral toxicity
2.507 mol 0.965 mol 3.366 mol 1.478 mol 2.389 mol 2.389 mol 1.929 mol 1.794 mol
kgÀ1
+(0.7250)
+(0.8130)
kgÀ1
—
—
kgÀ1
—
—
kgÀ1
kgÀ1
kgÀ1
kgÀ1
kgÀ1
Organ toxicity
Organ toxicity
Hepatotoxicity
+(0.7750) +(0.7000) +(0.7000) +(0.7750) +(0.8000)
—
Human either-a-go-go-ralted
gene (hERG) inhibition
Eye irritation
+(0.7970) +(0.7970)
—
—
Organ toxicity
Organ toxicity
Genomic toxicity Ames mutagenesis
Genomic toxicity Carcinogenesis
Genomic toxicity Micronucleus assay
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
Eye corrosion
+(0.8600) +(0.6874) +(0.8000) +(0.7500) +(0.7300) +(0.7300) +(0.7300) +(0.8800)
Toxicological
pathway
Toxicological
pathway
Toxicological
pathway
Toxicological
pathway
Androgen receptor (AR)
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
Aromatase
+(0.7535)
+(0.8084) +(0.7360)
Estrogen receptor (ER)
—
—
—
—
—
—
—
—
Peroxisome proliferator
activated receptor Gamma
(PPAR-Gamma)
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 8178–8191 | 8189

View Article Online
NJC
Paper
Table 6 Toxicity prediction of bortezomib and its degradation products
Bortezomib and its degradation products
Classification
Toxicity class
Target
DP-8
DP-9
DP-10
DP-11
DP12
DP-13
DP-14
DP-15
III
DP-16
III
III
III
III
III
III
III
III
Predicted LD50 Acute oral toxicity
2.308 mol 2.503 mol 2.766 mol 2.761 mol 2.761 mol 1.847 mol 1.811 mol 2.316 mol 2.316 mol
kgÀ1 kgÀ1 kgÀ1 kgÀ1 kgÀ1 kgÀ1 kgÀ1 kgÀ1 kgÀ1
+(0.7250) +(0.6750) +(0.6000) +(0.5750) +(0.5750) +(0.7750) +(0.7500) +(0.7500) +(0.7500)
Organ toxicity
Organ toxicity
Hepatotoxicity
Human either-a-go-go-ralted —
gene (hERG) inhibition
—
—
—
—
—
+(0.7302) +(0.9167) +(0.9167)
Organ toxicity
Organ toxicity
Genomic toxicity Ames mutagenesis
Genomic toxicity Carcinogenesis
Genomic toxicity Micronucleus assay
Eye irritation
Eye corrosion
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
+(0.8400) +(0.8700) +(0.8400) +(0.8900) +(0.8900) +(0.7200) +(0.7300) +(0.7100) +(0.7100)
Toxicological
pathway
Toxicological
pathway
Toxicological
pathway
Toxicological
pathway
Androgen Receptor (AR)
—
+(0.5286)
—
—
—
—
+(0.5836) +(0.5836)
—
—
—
—
—
—
—
—
—
—
—
—
—
Aromatase
+(0.8012) +(0.5609)
+(0.5827) +(0.5827)
+(0.7003)
Estrogen Receptor (ER)
+(0.7020)
—
—
—
—
—
—
Peroxisome Proliferator
Activated Receptor Gamma
(PPAR-Gamma)
+(0.5304)
+(0.5711) +(0.5711)
3.6 In silico toxicity studies
products were found using methanol as a co-solvent. A total of
sixteen degradation products were identified and characterized
by LC-MS/MS in combination with accurate mass measure-
ments. Major degradation products (DP-3, DP-6, DP-9, DP-10,
DP-11 and DP-12) were isolated by preparative HPLC and
subjected to advanced NMR experiments to confirm their
structures. Furthermore, in vitro cytotoxicity of the drug and
isolated DPs was assessed on normal cell lines (CHO-K1, HEK-
293 and NRK-49F) using MTT assays and showed that they were
less toxic as compared with standard drugs. In silico toxicity
studies of BTZ and its DPs were evaluated by the admetSAR
prediction tool.
The different levels of potential toxicity of BTZ and its stress
DPs were predicted by the admetSAR tool. The organ toxicity,
acute oral toxicity, toxicity targets and toxicological endpoints
were determined. This information provides the possible
underlying mechanisms for toxic response. The predicted out-
come for each target is given by a probability value. The
predicted toxicities for BTZ and its DPs are shown in the
Tables 5 and 6. All the DPs of BTZ fall under toxicity class III,
and the predicted acute oral toxicity is within the range of 500–
5000 mg kgÀ1. The hepatotoxicity of BTZ and it DPs (DP-3 to
DP-16) is predicted to be active. BTZ and all DPs were predicted
to possess potential genotoxicity. BTZ and its DP-4, DP-5, DP-
14, DP-15, and DP-16 are predicted to show cardiotoxicity.
Moreover, DP-3, DP-6 to DP-9, DP-11, DP-12 and DP-14 are
shown to have a probable binding towards aromatase. DP-9,
DP-11, and DP-12 are predicted to have probable binding to the
androgen receptor (AR) and peroxisome proliferator activated
receptor gamma (PPAR-Gamma). Therefore, the admetSAR
prediction tool indicated that there is a probability that BTZ
and its DPs can be hepatotoxic and genotoxic resulting in
severe toxicity.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
The authors thank Dr S. Chandrasekhar, Director, IICT,
Hyderabad, for facilities and their cooperation. R. Srinivas
thanks CSIR emeritus project, ES1617Y5030, for partial finan-
cial support. S. U. is thankful to CSIR, New Delhi, for awarding
a Senior Research Fellowship and a Junior Research Fellowship,
respectively. G. S. is thankful to U.G. C., New Delhi, for award-
ing a Junior Research Fellowship. R. B. and A. J. thank Dr USN
Murty, Director, NIPER-Guwahati, for cooperation.
4. Conclusions
Forced degradation study of bortezomib was carried out
according to ICH recommended guidelines to understand its
degradation behavior, characterize DPs and to evaluate their
toxicities. Stress experiments were performed under hydrolytic,
oxidative, thermal and photolytic conditions using acetonitrile
and methanol as co-solvents. A gradient LC-MS method was
utilized to separate the drug and its degradation products.
Thirteen degradation products were identified using aceto-
nitrile as a co-solvent. In addition, another three degradation
References
1 ICH guidelines, Q1A (R2): Stability Testing of New Drug
Substances and Products (revision2), International Confer-
ence on Harmonization.
8190
|
New J. Chem., 2021, 45, 8178–8191
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
Paper
NJC
2 Committee for proprietary medicinal products (CPMP) 17 A. Bolognese, A. Esposito, M. Manfra, L. Catalano,
guideline on stability testing: stability testing of existing
active substances and related finished products date for
coming into operation, 2003.
F. Petruzziello, M. C. Martorelli, R. Pagliuca, V. Mazzarelli,
M. Ottiero, M. Scalfaro and B. Rotoli, Adv. Hematol., 2009,
704928.
3 M. Blessy, R. D. Patel, P. N. Prajapati and Y. K. Agrawal, 18 ICH, Impurities in New Drug Products Q3B (R2). Interna-
J. Pharm. Anal., 2014, 4(3), 159–165.
4 S. Singh and M. Bakshi, Guidance on conduct of stress tests
tional Council for Harmonisation, IFPMA: Geneva (Switzer-
land), 2006.
to determine inherent stability of drugs, Pharm. Tech. On- 19 ICH, Impurities in New Drug Substances Q3A (R2). Interna-
line, 2000, 24, 1–14.
5 S. Ventkataraman and M. Manasa, Drug Invent. Today, 2018,
10(2), 137146.
6 P. G. Richardson, T. Hideshima and K. C. Anderson, Cancer
Control, 2003, 10, 361–369.
7 M. Cavo, Leukemia, 2006, 20, 1341–1352.
tional Council for Harmonisation. IFPMA: Geneva (Switzer-
land), 2006.
20 A. Chatzitakis, C. Berberidou, I. Paspaltsis, G. Kyriakou,
T. Sklaviadis and I. Poulios, Water Res., 2008, 42, 386–394.
21 A. Mendez, P. Chagastelles, E. Palma, N. Nardi and
E. Schapoval, Int. J. Pharm., 2008, 350, 95–102.
8 M. D. Jones, J. C. Liu, T. K. Barthel, S. Hussain, E. Lovria, 22 J. M. Gross, Ann. Intern. Med., 1963, 58, 523.
D. Cheng, J. A. Schoonmaker, S. Mulay, D. C. Ayers, 23 S. Udutha, G. Shankar, R. M. Borkar, K. Kumar, G. Srinivasulu,
M. L. Bouxsein, G. S. Stein, S. Mukherjee and J. B. Lian,
Clin. Cancer Res., 2010, 16, 4978–4989.
L. Guntuku, V. G. M. Naidu and R. Srinivas, J. Mass Spectrom.,
2018, 53, 963–975.
9 S. Venkat Rao, M. Srinivasarao, G. Ramu and C. Rambabu, 24 G. Shankar, R. M. Borkar, U. Suresh, L. Guntuku,
Der Pharm. Lett., 2012, 4(3), 720–727.
10 C. Rambabu, S. Venkatrao, G. Ramu and M. Ganesh, Int,
J. Pharm. Pharm. Sci., 2011, 3, 303–305.
11 N. Sanni Babu and B. Hari Babu, Future J. Pharm. Sci., 2021,
7(8), 1–11.
V. G. M. Naidu, N. Nagesh and R. Srinivas, J. Mass Spectrom.,
2017, 52, 459–471.
25 G. Shankar, R. M. Borkar, S. Udutha, M. Kanakaraju,
G. Sai Charan, S. Misra and R. Srinivas, New J. Chem.,
2019, 43, 7294–7306.
12 B. Chandramowli and B. B. Rajkamal, J. Appl. Pharm. Sci., 26 M. M. Bhandi, R. M. Borkar, G. Shankar, S. Raut, N. Nagesh
2017, 7(01), 035–041. and R. Srinivas, RSC Adv., 2016, 6, 10719–10735.
13 P. Ravipratap, K. Vanitha Prakash, U. B. Abhini, B. Naga 27 ICH guideline, Q1B Photo Stability Testing on New Drug
Maheswari, V. Divya, M. V. Shushmitha and V. Ravali, Br.
J. Pharm. Res., 2016, 14(4), 1–17.
Substances and Products, International Council for Harmo-
nization, IFPMA, Geneva, Switzerland, 2000.
14 T. Pekol, J. S. Daniels, J. Labutti, I. Parsons, D. Nix, 28 K. Javvaji, G. Begum, S. S. Deshpande, R. K. Rana and
E. Baronas, F. Hsieh, L.-S. Gan and G. Miwa, Drug Metab.
Dispos., 2005, 33, 771–777.
15 F. Y. Hsieh, E. Tengstrand, T. M. Pekol, R. Guerciolini and
G. Miwa, J. Pharm. Biomed. Anal., 2009, 49, 115–122.
S. Misra, Chem. Res. Toxicol., 2018, 31, 629–636.
29 P. N. Patel, P. D. Kalariya, M. Sharma, P. Garg, M. V. N. Kumar
Talluri, S. Gananadhamu and R. Srinivas, J. Mass Spectrom.,
2015, 50, 918–928.
16 S. R. Byrn, P. A. Tishmack, M. J. Milton and H. van de Velde, 30 G. W. Kabalka, T. M. Shoup and N. M. Goudgaon, J. Org.
AAPS PharmSciTech, 2011, 12, 461–467. Chem., 1989, 5930–5933.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 8178–8191 | 8191