Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 143 (2015) 12–19

Contents lists available at ScienceDirect

Spectrochimica Acta Part A: Molecular andBiomolecular Spectroscopy

journal homepage: www.elsevier .com/locate /saa

The use of laser-induced fluorescence or ultraviolet detectorsfor sensitive and selective analysis of tobramycin or erythropoietinin complex samples

http://dx.doi.org/10.1016/j.saa.2015.02.0251386-1425/� 2015 Elsevier B.V. All rights reserved.

⇑ Corresponding author.E-mail address: hmaahmed@yahoo.co.uk (H.M. Ahmed).

Hytham M. Ahmed a,⇑, Wael B. Ebeid b

a Pharmaceutical Analysis Department, Faculty of Pharmacy, Damanhour University, Damanhour, Egyptb SEDICO Pharmaceuticals, Merck & Co External Partner, 6th of October City, Cairo, Egypt

h i g h l i g h t s

� LIF detector is used for tobramycinanalysis in human urine.� Urine samples were injected directly

without pretreatment.� Erythropoietin was analyzed in the

presence of albumin by CE-UV.� EK and discontinuous buffer used to

increase method sensitivity.

g r a p h i c a l a b s t r a c t

Chemical structure of tobramycin (upper) and primary structure of human erythropoietin (lower).

a r t i c l e i n f o

Article history:Received 22 October 2014Received in revised form 29 January 2015Accepted 4 February 2015Available online 14 February 2015

Keywords:Laser-induced fluorescenceCZEMEKCUrine direct injectionErythropoietinTobramycin

a b s t r a c t

Complex samples analysis is a challenge in pharmaceutical and biopharmaceutical analysis. In this work,tobramycin (TOB) analysis in human urine samples and recombinant human erythropoietin (rhEPO) ana-lysis in the presence of similar protein were selected as representative examples of such samples analysis.Assays of TOB in urine samples are difficult because of poor detectability. Therefore laser induced fluores-cence detector (LIF) was combined with a separation technique, micellar electrokinetic chromatography(MEKC), to determine TOB through derivatization with fluorescein isothiocyanate (FITC). Borate was usedas background electrolyte (BGE) with negative-charged mixed micelles as additive. The method was suc-cessively applied to urine samples. The LOD and LOQ for Tobramycin in urine were 90 and 200 ng/mlrespectively and recovery was >98% (n = 5). All urine samples were analyzed by direct injection withoutsample pre-treatment. Another use of hyphenated analytical technique, capillary zone electrophoresis(CZE) connected to ultraviolet (UV) detector was also used for sensitive analysis of rhEPO at low levels(2000 IU) in the presence of large amount of human serum albumin (HSA). Analysis of rhEPO wasachieved by the use of the electrokinetic injection (EI) with discontinuous buffers. Phosphate bufferwas used as BGE with metal ions as additive. The proposed method can be used for the estimation of largenumber of quality control rhEPO samples in a short period.

� 2015 Elsevier B.V. All rights reserved.

Introduction

Pharmaceutical and biopharmaceutical analysis are based onqualitative and quantitative analysis of traditional and biotech

H.M. Ahmed, W.B. Ebeid / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 143 (2015) 12–19 13

drugs. However one of the most important challenges in analysis isthe sensitivity of the analytical methods. This sensitivity is notneeded only for the analysis of low detection substances but alsofor low concentrations. Therefore, the use of sensitive hyphenatedanalytical techniques, such as capillary electrophoresis techniques(CE), are increasingly in a wide range of applications. In general,CE separates the components of a sample on the bases of differencesin their charge-to-size ratio, and then detects the separated compo-nents using UV or fluorescence based on their properties. Howevertheses detectors cannot be used for the analysis of low detectionsubstances. Also they are not sensitive enough for detection andquantitation of very minute concentrations. Therefore, derivatiza-tion reactions which give stable derivative are essential in the caseof low detection substances. On the other hand electrokinetic injec-tion (EK) coupled with discontinuous buffers are used for enhancesensitivity towards low analyte concentration. In this work, tobra-mycin (TOB) analysis in human urine samples and recombinanthuman erythropoietin (rhEPO) analysis in the presence of similarprotein were selected as representative examples of such samplesanalysis. TOB is a member of aminoglycosides antibiotics (Fig. 1).It exhibits bactericidal activity against a broad spectrum of bacteriaspecially Pseudomonas-aeruginosa [1]. However when determina-tion of the drug was required, particularly in biological fluids, itsdetection was complicated because of low detection sensitivitydue to the poor chromophore effects and when chemical derivatiza-tion was used, poor stability of the determination was found.

The literature showed a mass spectrometric [2], spectrofluori-metric [3,4], spectrophotometric and colorimetric methods [5–7]for TOB analysis. But each of these methods is not ideal to efficient-ly detect TOB at trace level. Regarding to chromatographic analysisof TOB, it was analyzed by paper chromatography [8] and thin lay-er chromatography [9]. Gas liquid chromatography [10]. However,

Fig. 1. Chemical Structure of TOB.

Fig. 2. Primary structure of human er

HPLC is the most common method of analysis of TOB [11,12] Butthe major drawback was the toxicity of the reagent and slownessof reaction. Also, the main disadvantages of the reported HPLCpre- and post-derivatizations were the instability of the derivativesor complicated procedures [13–18]. Few trials of separation of TOBby CE are reported [19–22]. However these methods were unlikelyto give low detection sensitivities. Therefore, derivatization wasdone with OPA with 3-mercaptopropionic acid (MPA) and thenseparation of the derivatives by capillary zone electrophoresis(CZE) [23,24] or by MEKC [25] then direct UV detection. However,instability of the produced derivative was a problem.

The other example used in this work is rhEPO (Fig. 2) which is aglycoprotein consisting of 165 amino acid residues. rhEPO is usedas erythropoiesis-stimulating agents for renal anemia during dialy-sis, anemia of prematurity, and cancer related anemia worldwide.rEPO innovator and biosimilar products have been marketed inthe USA, Japan, the EU and Egypt [26]. For clinical use, highly effi-cient methods are required to analyze recombinant proteins [27].CE has been established as an effective analytical separation toolfor a wide variety of analytes, ranging from small inorganic ionsto biological macromolecules [28–31]. Separation and detectionof erythropoietin by CE and CE–MS [32–36]. However, rhEPO eitherwas alone or formulated with polysorbate 80. Albumin is used asrhEPO stabilizer and both were good separated by CE howeverwithout good sensitivity [37]. A trial to increase sensitivity wasdone by immunochromatographic removal of albumin in erythro-poietin biopharmaceutical formulations for its analysis by CE [38].However, this method was complex, expensive and time consum-ing. The European Pharmacopoeia (Ph. Eur.) monograph for Ery-thropoietin Concentrated Solution [39] describes a CZE methodfor identification of rhEPO and separation of its glycoforms. How-ever, this method has shown poor reproducibility due to inade-quate capillary conditioning [33,40]. In CE, EK is a highlycontroversial sampling technique. It is a simple mode of sampleintroduction which is suitable for on-line preconcentration of theanalytes [41]. The main advantage of EK injection is that sensitivityof the methods can be by several orders of magnitude higher, andconsequently, the limit of detection (LOD) correspondingly lowerthan using conventional hydrodynamic (HD) injection [41]. EKsampling can be exploited primarily for the separation of compo-nents of low diffusion coefficient, e.g., proteins, where the numberof theoretical plates is in the order of millions [42]. The presence ofsalt, problematic to traditional CE methods and overly abundant inprotein samples. In this work, desalting of samples followed by the

ythropoietin (mature hormone).

Fig. 3. Representative electropherograms of TOB–FITC and its blank.

14 H.M. Ahmed, W.B. Ebeid / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 143 (2015) 12–19

use of discontinuous buffer in CE method were done to solve thisproblem.

Therefore, the aim of this work was to develop a new CE methodthat can be used for quantitative analysis of TOB in human urineafter derivatization with FITC. The stable fluorescent derivativeswas detected by LIF detector through direct urine injection. Also,this work aimed to develop a sensitive, selective and reproduciblemethod for the characterization and quantification of rhEPO glyco-forms in bulk and finished products.

Experimental

CE systems

LIF instrumentationA model P/ACE 5510 Beckman capillary electrophoresis instru-

ment (Fullerton, CA, USA) equipped with a 3 mW, 488 nm air-cooled argon-ion laser (Beckman Laser Module 488) was used.The fluorescence emission was 520 nm filtered by a band pass fil-ter, and a notch filter was used to attenuate background radiation.Uncoated fused silica capillaries (75 lm ID � 363 lm OD, totallength 75 cm and effective length 60 cm) were obtained fromSupelco (Bellefonte, PA, USA) were accommodated in a Beckmancartridge configured for LIF detection. The capillaries were keptat constant temperature using a thermostated liquid coolant. Alloperations of the P/ACE unit were controlled by a PC-Pentium75 MHz compatible computer running Beckman Gold Software.Nitrogen gas cylinder (BOC, Manchester, UK) was essential forthe sample injection and to flush the capillary.

UV instrumentationA Beckman Coulter P/ACE™ MDQ Capillary Electrophoresis Sys-

tem (Fullerton, CA) was used in this study. The instrument wasequipped with a UV detector module and all measurements weremade at 200 nm. eCAP Amine capillary with an i.d. of 75 lm wasused. The column temperature was controlled at 4 �C.

Chemicals and materials

TOB and FITC, boric acid, TX-100 and SDS were purchased fromsigma–aldrich, UK. Vials of formulated rhEPO (EPO 2000 IU, EPO4000 IU) were supplied by SEDICO manufacturer, Egypt. ReferenceEPO and high purity HSA was obtained from Miles (DiagnosticsDivision, Kankakee, IL, USA). All other chemicals were of the high-est purity analytical grade available from (BDH, UK). eCAP Aminecapillaries and reagents were purchased from Beckman Instru-ments (Fullerton, CA, USA). The wash and conditioning proceduresused were those recommended by the manufacturer.

Procedures

Analysis of TOB in bulk by CE-LIF after derivatization with FITCCapillary conditioning. Initially a new capillary was treated with1 M NaOH for 15 min, followed by water for 10 min, and then run-ning buffer for 10 min. Between runs, the capillary was flushedwith 0.1 M NaOH for 2 min followed by running buffer for 2 min.

Preparation of buffer solutions. When buffers were employed, thesalts and/or additives (like SDS) in question were weighed andtransferred to a suitable volumetric flask. The salts and/or additiveswere dissolved by addition of some double-distilled water (about80–90% v/v of total volume) before being made up to volume.The pH of the buffer was then corrected using an appropriate acidor alkali solution before filtration through a 0.45 lm membrane fil-ter. Care was taken to ensure that the pH meter was calibrated

twice daily using freshly prepared commercially available standardbuffers at pH 4.0 and 7.0. All buffers were freshly prepared on adaily basis.

Preparation of TOB solutions. A standard solution containing 1 mg/ml of TOB was prepared in deionized water. Further dilutions weremade with water to required concentration. From this stock stan-dard solutions, working standard solutions containing TOB in therange of (0.25–5 lg/ml) were prepared by dilution with distilledwater.

FITC derivatization procedure. 300 ll aliquoit of aqueous TOB solu-tion containing (0.25–5 lg/ml) was added to 300 ll FITC(0.35 mM) in ethanol and 200 ll buffer (5 mM boric acid adjustedto pH 7.8 with 2 M KOH) in 2 ml reaction vial. The mixture vial wascapped, homogenized, vortexed and allowed to react at 80 �C in anoven for 20 min. The derivatization mixtures were analyzed with-out dilution.

Optimization of derivatization reaction. The effect of the pH and theconcentration of the borate buffer were studied over the ranges6.5–9.5 and 5–100 mM, respectively. The effect of temperaturewas investigated by allowing the reaction to precede at differentincubation temperatures ranging from 40 to 85 �C. The stabilityof the derivatized TOB was monitored by measuring the peak areasof the derivative every 10 min up to 90 min.

CE operating parameters for separation of FITC–TOB derivative. TheBGB consisted of Boric acid 2.8 g, sodium borate 2.1 g dissolvedin 100 ml double-distilled water, 1.75 ml v/v TX-100 and 5.25 gw/v SDS in deionized distilled water in 100 ml volumetric flask.The pH was adjusted to 7.8 with 3 M NaOH solution using a mag-netic stirrer and pH meter and the volume was completed with dis-tilled water. Sample introduction was performed by hydrodynamicinjection at 50 mbar for 1–10 s. Separations were performed atroom temperature (25 �C) using a separation voltage of 10–30 kV

Table 1Intra-day and inter-day precision (n = 5) for TOB solutions injected twice after LIF-detection of FITC-derivatization.

TOB Concentration (lg/ml) Intra-day(n = 5)

Inter-day(n = 5)

RSD% Recovery% RSD% Recovery%

0.2 2.54 100.54 9.23 98.232 0.75 101.60 5.99 100.995 0.88 99.75 9.12 99.12

Scheme 1. Structure of the amino compounds and their derivatization reactionscheme with FITC.

H.M. Ahmed, W.B. Ebeid / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 143 (2015) 12–19 15

and on-line detection with the LIF detection system (Ex 488 nm,Em 520 nm).

MECK-LIF separation of TOB in spiked urine after derivatization withFITC.Preparation of TOB spiked urine (TSU) Samples. Human urine sam-ples were collected from five different volunteers (males and

OH

OH

CH3

CH3

Cn

.

.

B

OHOH

OH OH

+

B

OOH

OH O

CH3

CH3

Cn

.

. OH

OH

CH3

CH3

Cn

.

.

+

B

OH

OH

OH+ OH -

(B-) (L)

(B)

(BL-) (L)

Scheme 2. Equilibria between boric

females). The urine samples were mixed and a representative 2 Lsample was taken for the preparation of the standard solutions.Polymyxin B was used (0.4 g/2L) as an internal standard to giveconcentration of 0.2 mg/ml as standard solution. Urine TOB(0.1 g) was spiked into 100 ml urine to give concentration 1 mg/ml then 10 ml was taken from the last solution to a 100 volumetricflask and made up to the volume with the internal standard urinesolution to give a concentration of 100 lg/ml. Different drug con-centrations ranging from 0.25 to 5 lg/ml were prepared by serialdilution of TOB in the internal standard urine solution.

CE procedure for TSU samples. The derivatization of the spiked urinesamples was performed as above. CE parameters were, capillarylength 75 cm with 75 lm ID (66 cm effective length) and BGB con-sisted of Boric acid 2.8 g, sodium borate 2.1 g, 1.75% TX-100, 5.25 gSDS in 100 ml volumetric flask and the volume was completed withdeionized water. The pH was adjusted to 7.8 with 3 M NaOH. Theapplied voltage was 10 kV and the injection time was 8 s.

CZE-UV Separation of EPO and HASPreparation of rhEPO solution. Fresh HSA stock solution was pre-pared immediately prior to analysis in Milli-Q water at a nominalconcentration of 2.0 mg/ml from high purity HSA. 72 ll ‘‘1 mg/ml’’ rhEPO BP_Reference were added to 25 ll HSA. Then 1.9 ml distwater were added. The obtained solution was mixed very will byrotator for 30 s. For preparation of CE sample, 900 ll distilled waterwere pipetted to CE vial. Then rhEPO and Albumine (or EPO fin-ished product) were added. Finally, the volume was completed to2 ml with distilled water.

Desalting procedure. EPO vials (1 ml) contains HSA, Sodium citrate,Sodium chloride and Citric acid with the following amounts 2.5,5.8, 5.84 and 0.057 mg respectively. The volume was completedwith double distilled water. Due to this large amount of salt, adesalting procedure was done as follows: centrifugal filtration

B

OOH

OH O

CH3

CH3

Cn

.

.

+ OH2

.

.

CH3

CH3

Cn B

OO

O O

CH3

CH3

Cn

.

.

B

OHOH

OH OH

2

+ OH22

(B-)

(BL2-)

(BL-)

(1)

(2)

(3)

acid, borate and diols in water.

Fig. 4. Laser Induced Fluorescence detection of tobramycin FITC derivative inhuman urine.

Table 2Intra-day and inter-day precision and accuracy for TSU after FITC derivatizationmethod.

Spiked TOB concentration (lg/ml) Intra-day Inter-day

RSD% Recovery% RSD% Recovery%

0.5 0.86 99.30 2.45 98.51.0 2.01 99.67 8.30 99.72.0 1.77 99.50 2.20 98.44.0 4.89 99.20 1.37 99.2

Fig. 5. Sketch diagram for CE system for rhEPO analysis.

16 H.M. Ahmed, W.B. Ebeid / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 143 (2015) 12–19

device ‘‘Microcon YM-10’’ with membrane (Amicon, Beverly, MA,USA) was used as a means to concentrate the sample. The centrifu-gator must be cooled in the refrigerator before its use for at least1 h. All desalting steps were done with the use of cold distilledwater. By the use of micropipette, 4 filtration beds were condi-tioned with 300 ll cold distilled water for each at 11,500�g/12,500 rpm for 20 min. The filtrate was rejected. Then 200 ll EPOsample were pipette to the used beds followed by 250 ll cold dis-tilled water. The centrifugation was done at 2700�g for 50 min(repeated twice by using 300 ll cold distilled water each time).The retenate were collected at 3200�g for 5 min. All retenatedwere collected in a 2 ml volumetric centrifugation vial and the vol-ume were completed with cold distilled water.

Capillary electrophoresis procedure. Sodium phosphate buffers(200 mM) pH 4.0 and 9.0 containing 1 mM Nickel Chloride hexahy-drate. The prepared buffers were filtered before use. The appliedvoltage was 10 kV, with reversed polarity, at 20 �C. eCape Aminecapillary was used. Electrokinetic injection with 10 KV withreversed polarity, for 10 s.

Capillary regeneration. When the capillary shows signs of reducedperformance or when the percent relative standard deviation(%RSD) of the absolute migration time is greater than 10%. The fol-lowing procedure should be used to prevent any sample carry-overfrom previous separations; several high pressure (20 psi) rinses of1 N HCl were first done for 5 min, then water for 5 min, followed by1 N NaOH for 2 min, again water for 5 min, and finally with amineregenerator solution for 5 min.

Results and discussion

TOB analysis

Optimization of derivatization reactionEffect of pH and buffer concentration. Borate buffer solutions of dif-ferent pH values ranging from 6.5 to 9.5 were examined. Maximumpeak areas (which indicates fluorescence intensity) for derivatizedTOB were at a pH range of 7.5 to 7.9. In fact, at borate buffer con-centrations over 10 mM, the analytical signal strongly decreases,which can be ascribed to the formation of aminoglycoside–boratecomplexes. These complexes have a negative charge and therefore

an electrostatic repulsion can be expected in their reaction withthe label, with the subsequent decrease in the analytical signal.Based on these results, the pH in the derivatization medium wasadjusted to 7.8 with 10 mM borate buffer.

Effect of reaction time and temperature. The effect of temperaturewas investigated by allowing the reaction to precede at differentincubation temperatures ranging from 40 to 85 �C. The rate of reac-tion of TOB with FTIC was increased as temperature increased upto 70 �C and after that no increases were observed. However toensure rapid and full reaction, 80 �C was chosen as the best reac-tion temperature. In order to test the stability of the derivatizationwhich was suggested earlier as crucial, time studies were carriedout. The stability of the derivatized TOB was monitored by measur-ing the peak areas of the derivative every 10 min up to 90 min. Theresults indicate the maximum stability was one hour at room tem-perature reaction time of 10 min.

Optimization of CE conditions for FITC. Borate buffers were exam-ined and borate (boric acid from 1.4 g to 2.8 g and sodium tetrabo-rate 1.05 g–2.1 g) was found to be the best and neomycin wasexamined to be an internal standard. It is however known that,borate can form a complex with aminoglycosides [19,43]. Theinteraction of borate ions with TOB was therefore studied. Thiswas carried out by using different borate buffer concentrations.The optimum borate buffer concentration for the assay of TOBwas 2.8 g boric acid and 2.1 sodium tetraborate in 100 ml waterand amikacin was used as an internal standard.

Validation of FITC derivatization methodAmikacin sulfate was added (0.4 g/2L) as internal standard to

give concentration of 0.2 mg/ml as standard solution. 0.1 g TOB

Fig. 6. Electropherograms of rhEPO in bulk (upper black bold) and in the presence of albumin (lower black thin).

H.M. Ahmed, W.B. Ebeid / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 143 (2015) 12–19 17

was spiked into 100 ml internal standard solution to give concen-tration of 1.0 mg/ml then 10 ml was taken from the last solution toa 100 volumetric flask and completed to the volume with the inter-nal standard solution to give concentrations of 100 lg/ml then dif-ferent concentrations of TOB in the internal standard solution togive different concentrations in ranging from 0.25 to 5 lg/ml.The best CE conditions are shown in Fig. 3.

CE conditions: capillary length 75 cm � 75 lm ID and BGE:boric 2.8 g and sodium borate 2.1 g/100 ml water 20 kV injectiontime 20 s (Green) blank, (Blue) TOB–FITC (Red) Amikacin withtobramycin, detection was LIF Ex 488 nm; Em 520 nm.

The linearity of the CE pre-column FITC derivatization methodwas examined by using TOB at six different concentrations (injec-tions in duplicate), ranging from 0.25 to 5 lg/ml. The standardcalibration curve for TOB was linear, and described by the follow-ing equation:

y ¼ 0:7804��0:0764ðR2 ¼ 0:9991 n ¼ 6Þ:

The LOD and LOQ for TOB were calculated from the mean of theintercept of five calibration curves. The LOD and LOQ were 70 and160 ng/ml, respectively for LIF-detection.

The precision of assay was evaluated by determining the intra-day and inter-day %RSD for five replicates (n = 5) at three differentconcentrations of TOB solutions (low, medium and high) with LIF-detection. The results are summarized in Table 1.

Mechanism of TOB–FITC reactionIt is known that, FITC can react with hydroxyl-containing small

molecules [44,45]. However in the case of compounds containingboth hydroxyl and amino groups, the latter can be the only avail-able functional group for such derivatization reactions [46,47].Therefore, the five hydroxyl groups of tobramycin are not expectedto participate in these reactions. It is also unlikely to have two FITCmolecules reacting with one TOB molecule because of the large

size of the molecules and the close proximity of the amino groups.Therefore, it is likely that, only one amino group of the five distinctprimary amino groups of TOB can react with these reagents due tosteric effects (Scheme 1). At the same time, the presence of morethan one TOB peak in the electrogram could be attributed to theformation of more than one derivatization product due to differentamino sites on TOB structure. Therefore it is suggested that thesame ring reacts at different sites each time to give another pro-duct or the other peaks for degradation or impurities of TOB. Thisobservation follows a similar behavior as previously discussed forthese TOB–OPA derivatives when two derivative peaks have beengiven [24,48].

Mechanism of borate complexation and binding sitesThe use of borate buffer to alter selectivity in CE was first inves-

tigated in 1988 [49]. At higher borate concentration in the aqueousphase, tri- to pentameric polyanionic species such as [B3O3(OH)5]2�,[B4O5(OH)4]2� and [B5O6(OH)4]2� are present [50]. These species canreact with the single hydroxyl groups in molecules to form chargedand mobile complexes [51]. Borate complexation induces changes inthe charge to mass ratios of the ligands. Borate is known to inducethe formation of charged and mobile complexes with 1,3-diols six-membered hexoses [50]. Tetrahydroxyborate ion, rather than boricacid, is complexed by the polyols [52] and these complex formationproperties were employed for the characterization of polyols usingCZE and MECK [53]. According to the chemical structure of TOB, itcan be considered as polyol due to the presence of five hydroxylgroups. The complex formation can be described as shown inScheme 2.

Where [54] n = 0 or 1, L is polyol ligand and B- representstetrahydroxyborate, [B(OH)4 ]�. In a pH range from 8 to 12, aque-ous borate solutions contain not only tetrahydroxyborate ionsbut also more highly condensed polyanions such as triborate,[B3O3(OH)5]2�, tetraborate, [B4O5(OH)4]2�. The reported data

Minutes4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

AU

-0.002

-0.001

0.000

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0.008

0.009

AU

-0.002

-0.001

0.000

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0.008

0.009

UV - 200nmUV.10003 9-23-2010 12-17-33 PM.dat

NameMigration TimeArea

UV - 200nmuv.10003 9-23-2010 11-53-09 am.dat

Fig. 7. Electropherograms showing the difference of hydrodynamic (blue) and electrokinetic (red) injections of the same conc of rhEPO with the same duration. (Forinterpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

18 H.M. Ahmed, W.B. Ebeid / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 143 (2015) 12–19

[55] confirmed that: 1 – Polyols can form 1:1 and 1:2 complexeswith borate. 2 – Not only hydroxyl groups on adjuvant carbonatoms but also those on alternate carbon atoms are involved incomplexation. 3 – An increase in the number of hydroxyl groupsincreases the stabilization. So, in case of TOB, the decrease in itsmobility is not only due to the increase of the ionic strength ofthe borate buffer but also due to borate-TOB complexation becauseTOB acted as a typical polyol in its reaction with borate in BGE .

Determination of TOB in human urineTOB in a pooled human urine samples (0.25–5 lg/ml) was

determined by CE, using direct injection (without extraction) afterlabeling with FITC in the concentration range from 0.25 to 5 lg/ml.The derivatized sample was injected directly into the CE-LIF sys-tem for 8 s and separated at ambient temperature using a constantvoltage 10 kV. Endogenous components present in urine were alsoshown not to co-migrate with TOB which appeared as the first twopeaks in the electropherogram. The only suggested reason behindthis is an electrostatic repulsion between the anionic TOB deriva-tives and the TX100/SDS micelles because both the derivativeand the surfactant are negatively charged. The electrostatic repul-sion can prevent solubilization of TOB derivative in the mixedmicelle thus causing the optimum resolution between TOB andthe indigenous urine peaks and the internal standard peak. There-

fore the parameters in Fig. 4 are the optimum for this method. TheLOD and LOQ for TOB spiked in human urine were estimated prac-tically using a signal to noise ratio of 3 and 10 as 90 and 200 ng/ml,respectively.

CE conditions: capillary length 75 cm � 75 lm ID and BGE ((0.2boric acid + 1.75% TX100 + 5.25 g SDS/100 ml water in 100 mlvolumetric flask) the pH adjusted to 7.8 with 3 M NaOH, 10 kV,8 s inj time, detection was LIF Ex 488 nm; Em 520 nm.

The intra- and inter-day precision and accuracy of the assaywere established by calculating the mean concentration of fivereplicates (n = 5) at four different concentrations of TOB spikedurine (TSU) samples (0.5, 1.0, 2.0 and 4.0 lg/ml) (Table 2).

The recovery of TOB was determined by comparing replicate(n = 5) peak area ratios of urine spiked with known amounts ofthe used drug (0.5, 1.0, 2.0 and 4.0 lg/ml) vs. peak area ratios ofthe same concentrations calculated from the resultant regressionline. Five replicate analyses were carried out at each concentration.The precision, accuracy and recovery data obtained from TSU aresummarized in Table 2.

rhEPO analysis

CE has become a powerful analytical tool for resolving andquantifying complex protein mixtures as well as different forms

H.M. Ahmed, W.B. Ebeid / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 143 (2015) 12–19 19

of the same protein [56]. The main problem associated with pro-tein separation is protein adsorption onto the capillary wall [57].Therefore, eCAP Amine capillaries were used to prevent proteinadsorption. On the other hand, the high amount of salt which isavailable in EPO vials, causes problems in CE analysis due toimmoderate Joule heating and electrodispersion. Therefore, desalt-ing of proteins were done followed by CE with a discontinuous buf-fer of different pH. The discontinuous buffer was composed of anphosphate buffer pH 4 and 9. When these acidic and basic bufferswere placed at the cathode and anode, respectively, a sharp pHjunction was formed at the buffer boundary within the capillary.Proteins were trapped at the junction, resulting in the observedenrichment.

HSA is generally added to rhEPO formulations as a protein sta-bilizer. Separation of the two proteins by CZE may be problematicnot only because of their similarity but also due to the presence oflarge amounts of HAS. Metal ions are well known to selectivelyinteract with proteins and modify their electrophoretic mobilityCE separations [58]. The optimum conditions were sodium phos-phate buffers (200 mM) pH 4.0 and 9.0 containing 1 mM NickelChloride hexahydrate as shown in Fig. 5. The prepared buffers werefiltered before use. The applied voltage was 10 kV, with reversedpolarity, at 20 �C. eCape Amine capillary was used. Electrokineticinjection with 10KV with reversed polarity, for 10 s. Theses condi-tions for the separation of which allowed resolution of rhEPO intofour glycoforms as shown in Fig. 6.

These conditions also afforded complete separation of HSA andrhEPO without affecting the glycoform resolution pattern (Fig. 6).Typically, the migration time for the main rhEPO glycoform andHAS peaks were found to be 13 and 14 min respectively. The differ-ence between HD and EK injections of EPO can be shown in Fig. 7.

Conclusion

A MEKC-LIF using a pre-CE derivatization with FITC has beenused for the determination of TOB. A mixed micelle system com-posed of SDS and TX-100 as buffer additives improved resolution,selectivity and sensitivity of the method. The method was appliedsuccessfully for analysis of TOB in human urine samples by directinjection without sample pre-treatment procedures. The validationresults of this method indicate that it is accurate, precise and sen-sitive enough to be used for the analysis of TOB in urine samples.On the other hand, CZE-UV using nickel as metal ion additive tothe background electrolyte were used to overcome difficulties inassessing protein drug products due to the similarity of their phy-sicochemical properties particularly in the presence of largeamounts of excipients which interfere with the detection andseparation of the active ingredient leading to complete separationof the two proteins. Moreover, electrokinetic injection along withdiscontinuous buffer was used to increase the method sensitivity.The method was found to be useful for quantitative estimation ofrhEPO in both bulk and final drug preparations.

References

[1] R.D. Meyer, L.S. Young, D. Armstron, Appl. Microbiol. 22 (1971) 1147.[2] E. Kaale, C. Govaerts, J. Hoogmartens, A. Van Schepdael, Rapid Commun. Mass

Spectrom. 19 (2005) 2918–2922.[3] M. Omar, H. Ahmed, M. Hammad, S. Derayea, Spectrochim. Acta Part A Mol.

Biomol. Spectrosc. 135 (2014) 472–478.[4] A. Csiba, J Pharm. Pharmacol. 31 (1979) 115–116.[5] J.A. Ryan, J. Pharm. Sci. 73 (1984) 1301–1302.

[6] V. Dasgupta, J. Clin. Pharm. Ther. 13 (1988) 195–198.[7] S.S. Sampath, D.H. Robinson, J. Pharm. Sci. 79 (1990) 428–431.[8] R.L. Hussey, J. Chromatogr. 92 (1974) 457–460.[9] A.K. Dash, Tobramycin in Analytical profiles of drug substances and excipients,

Pub. Academic Press, 1996, 24, 579-613.[10] J.W. Mayhew, S.L. Gorbach, Antimicrob. Agents Chemother. 14 (1978) 851–

855.[11] A. Cabanes, Y. Cajal, I. Haro, J.M.G. Anton, M. Arboix, F. Reig, J. Liq. Chromatogr.

14 (1991) 1989–2010.[12] J. Marples, M.D.G. Oates, J. Antimicrob. Chemother. 10 (1982) 311–318.[13] B.G. Keevil, S.J. Lockhart, D.R. Cooper, Analytical technologies in the biomedical

and life science, J. Chromatogr. B 794 (2003) 329–335.[14] C.H. Feng, S.J. Lin, H.L. Wu, S.H. Chen, J. Chromatogr. B 780 (2002) 349–354.[15] M.H. Mashat, Pharmaceutical analysis, pharmacokinetic and in-vitro

aerodynamic characterisation of nebulised tobramycin. PhD Thesis, School ofPharmacy, University of Bradford, Bradford, 2006, p. 378.

[16] M. Mashat, H. Chrystyn, B.J. Clark, K.H. Assi, J. Chromatogr. B 869 (2008) 59–66.[17] D.M. Barends, C.L. Zwaan, A. Hulshoff, J. Chromatogr. 225 (1981) 417–426.[18] H. Fabre, M. Sekkat, M.D. Blanchin, B. Mandrou, J. Pharm. Biomed. Anal. 7

(1989) 1711–1718.[19] C.L. Flurer, J. Pharm. Biomed. Anal. 13 (1995) 809–816.[20] H.H. Yeh, S.J. Lin, J.Y. Ko, C.A. Chou, S.H. Chen, Electrophoresis 26 (2005) 947–

953.[21] P.D. Voegel, R.P. Baldwin, Electroanalysis 9 (1997) 1145–1151.[22] P. Srisom, B. Liawruangrath, S. Liawruangrath, J.M. Slater, S. Wangkarn, J.

Pharm. Biomed. Anal. 43 (2007) 1013–1018.[23] H. Fonge, E. Kaale, C. Govaerts, K. Desmet, A. Van Schepdael, J. Hoogmartens, J.

Chromatogr. B 810 (2004) 313–318.[24] E. Kaale, S.A. Van, E. Roets, J. Hoogmartens, Electrophoresis 23 (2002) 1695–

1701.[25] F. Wienen, U. Holzgrabe, Electrophoresis 24 (2003) 2948–2957.[26] W.M. Ebied, H.M. Ahmed, F.A. Elbarbry, Biosimilars 4 (2014) 11–22.[27] A. Pantazaki, M. Taverna, C. Vidal-Madjar, Anal. Chim. Acta 383 (1999) 137–

156.[28] F. Kvasnicka, Electrophoresis 28 (2007) 3581–3589.[29] M. Sniehotta, E. Schiffer, P. Zurbig, J. Novak, H. Mischak, Electrophoresis 28

(2007) 1407–1417.[30] Y. Li, X.B. Yin, X.P. Yan, Anal. Chim. Acta 615 (2008) 105–114.[31] V. Garcia-Canas, A. Cifuentes, Electrophoresis 29 (2008) 294–309.[32] B. Yu, H.L. Cong, H.W. Liu, Y.Z. Li, F. Liu, Trac – Trends Anal. Chem. 24 (2005)

350–357.[33] J. Zhang, U. Chakraborty, A.R. Villalobos, J.M. Brown, J.P. Foley, J. Pharm.

Biomed. Anal. 50 (2009) 538–543.[34] I. Lacunza, P. Lara-Quintanar, G. Moya, J. Sanz, J.C. Diez-Masa, M. de Frutos,

Electrophoresis 25 (2004) 1569–1579.[35] P. Dou, Z. Liu, J. He, J.-J. Xu, H.-Y. Chen, J. Chromatogr. A 1190 (2008) 372–376.[36] E. Gimenez, F. Benavente, C. de Bolos, E. Nicolas, J. Barbosa, V. Sanz-Nebot, J.

Chromatogr. A 1216 (2009) 2574–2582.[37] H.P. Bietlot, M. Girard, J. Chromatogr. A 759 (1997) 177–184.[38] P. Lara-Quintanar, I. Lacunza, J. Sanz, J.C. Diez-Masa, M. de Frutos, J.

Chromatogr. A 1153 (2007) 227–234.[39] European Pharmacopoeia 2 (2005) 1528–1529.[40] V. Sanz-Nebot, F. Benavente, A. Vallverdu, N.A. Guzman, J. Barbosa, Anal. Chem.

75 (2003) 5220–5229.[41] Z. Krivacsy, A. Gelencser, J. Hlavay, G. Kiss, Z. Sarvari, J. Chromatogr. A 834

(1999) 21–44.[42] H. Engelhardt, M.A. Cunatwalter, J. Chromatogr. A 717 (1995) 15–23.[43] C.L. Flurer, K.A. Wolnik, J. Chromatogr. A 663 (1994) 259–263.[44] Y. Kaneo, S. Hashihama, A. Kakinoki, T. Tanaka, T. Nakano, Y. Ikeda, Drug

Metab. Pharmacokinet. 20 (2005) 435–442.[45] D.A. Wicks, P.C.H. Li, Anal. Chim. Acta 507 (2004) 107–114.[46] D. Letourneur, D. Logeart, T. Avramoglou, J. Jozefonvicz, J. Biomater.

Sci. – Polym. Ed. 4 (1993) 431–444.[47] N.A. Stearns, S. PrigentRichard, D. Letourneur, J.J. Castellot, Anal. Biochem. 247

(1997) 348–356.[48] F. Lai, T. Sheehan, J. Chromatogr. 609 (1992) 173–179.[49] R.A. Wallingford, A.G. Ewing, J. Chromatogr. A 441 (1988) 299–309.[50] P. Schmitt-Kopplin, N. Hertkorn, A.W. Garrison, D. Freitag, A. Kettrup, Anal.

Chem. 70 (1998) 3798–3808.[51] A.L. Revilla, J. Havel, P. Jandik, J. Chromatogr. A 745 (1996) 225–232.[52] J.G. Dawber, G.E. Hardy, J. Chem. Soc. Faraday Trans. I 80 (1984) 2467–2478.[53] H.H. Thanh, J. Chromatogr. A 678 (1994) 343–350.[54] M. Makkee, A.P.G. Kieboom, H. van Bekkum, Recl. Trav. Chim. Pays-Bas 104

(1985) 230–235.[55] S. Hoffstetterkuhn, A. Paulus, E. Gassmann, H.M. Widmer, Anal. Chem. 63

(1991) 1541–1547.[56] Y. Chen, W. Lü, X. Chen, M. Teng, Cent. Eur. J. Chem 10 (2012) 611–638.[57] C.J. Booker, K.K.C. Yeung, Anal. Chem. 80 (2008) 8598–8604.[58] H. Kajiwara, J. Chromatogr. A 559 (1991) 345–356.