Advertisement

Monatshefte für Chemie - Chemical Monthly

, Volume 149, Issue 9, pp 1555–1560 | Cite as

Degradation of ouabain in 80-year-old injection solution studied by HILIC–MS

  • Karel Kudláček
  • Karel Nesměrák
  • Martin Štícha
  • Petr Kozlík
  • Jan Babica
Original Paper
  • 96 Downloads

Abstract

An 80-year-old injection solution of ouabain was analyzed using HPLC–MS. A new separation method involving hydrophilic interaction liquid chromatography (HILIC) using an amide-based stationary phase with mass spectrometric detection was developed for the analysis. The optimized mobile phase composition was acetonitrile:water 85:15 (v/v) at the flow rate of 0.2 cm3/min under the isocratic elution. The ouabain content was found reduced to 56% of declared content and probable degradation product with oxidized hydroxy group was established. The ESI-MS2 fragmentation mechanism of ouabain was proposed.

Graphical abstract

Keywords

High-performance liquid chromatography Decomposition Long-term stability Mass spectrometry g-Strophanthin 

Introduction

Injection solutions have become widely used in medicine since the second half of the nineteenth century [1]. These solutions are usually packed, supplied, and preserved in small hermetically sealed ampoules. The most used material for ampoule fabrication is glass, as it easily makes an impermeable container which hardly interacts with its contents. An injection solution stored in a glass ampoule is virtually completely separated from the outside world and it is not desirable to be decomposed for several years [2, 3]. The chemical changes of historical injection solutions, rarely preserved in museum collections, are interesting in pharmaceutical as well as historical context [4]. Although such analyses provide a unique opportunity to study composition of solutions that have been—as is supposable—prepared before many decades, they are relatively rare yet (Table 1). From the published studies, it could be argued that the stability of relatively complex organic molecules is very different, depending especially on the reactivity of the molecule [5]. It may be noted that, based on these findings, a discussion on the possible extension of shelf life of medicinal preparations has been recently opened [5, 6, 7].
Table 1

Published results of analyses of active substances in historical injection solutions, which were over 50 years in the time of analysis

Active substance(s)

Age/years

Observation

References

Cocaine

Approx. 70

Decrease to 27%; found decomposition products: benzoylecgonine, ecgonine, and ecgonine methyl ester

[8, 9]

Epinephrine

83

Decrease to 70%; found impurities: norepinephrine, adrenochrome and unknown

[10]

Etilefrine

55

No degradation, found unknown impurity

[10]

Furosemide

53

No degradation, found small amount of saluamine

[10]

Heroine

77

Completely deacetylated to morphine

[9]

Metamizole

53

No degradation

[10]

Procaine + caffeine

68

Decrease to 79%, found 4-aminobenzoic acid

[10]

Quinine

79

Decrease to 87%; decomposition product quinotoxine

[11]

Sodium salicylate + caffeine

72

No degradation, found small amount of theobromine

[10]

Synephrine

55

No degradation, found unknown impurity

[10]

Here, we communicate the analysis of 80-year-old injection solution of ouabain (Fig. 1), which has been preserved in the collections of the Czech Pharmaceutical Museum, and the comparison of analysis results with the information described in the literature. For that aim, we have developed a new hydrophilic interaction liquid chromatography (HILIC) method for separation of ouabain and its degradation product. Furthermore, ESI-MS2 fragmentation of ouabain is discussed.
Fig. 1

Packaging and ampoule of analyzed historical injection solution of ouabain named “Strophosan” manufactured by the Dr. Robert Heisler company (formerly Czechoslovakia) in the 1938

Ouabain (CASRN 630-60-4, also known as g-strophanthin) is a cardiac glycoside that non-selectively inhibits the sodium–potassium ion pump in the heart muscle [12, 13]. In lower doses, it can be used medically to treat hypotension, atrial fibrillation, and flutter or heart failure. Ouabain can be found in the roots, stems, leaves, and seeds of the Strophanthus gratus or Acokanthera schimperi plants, which are native to eastern Africa [14, 15]. Ouabain is a highly toxic compound [16], which has been traditionally used as an arrow poison (just for interest, it was also used by Agatha Christie in the novel Triangle at Rhodes). Ouabain was introduced to European medicine in the second half of the nineteenth century [17]. In 1882, ouabain was first isolated from the plant; the total synthesis of ouabain was achieved in 2008 [18]. In 1991, the so-called endogenous ouabain, which regulates contractility of heart muscle, was identified [19].

The stability of ouabain substance or its solutions has been first studied by Haskell and Doeppers [20]. They found that both the pure substance of ouabain and its aqueous solution undergo no change in 2 years under normal climatic conditions, even being exposed to the air and light. Other authors suggested that the stability of ouabain aqueous injection solution is increased by dissolution of the substance in a 0.02 M standard phosphate solution at pH 7.0 and using hard-glass ampoules [21, 22, 23]. Berry [24] investigated the effect of pH on the stability of ouabain aqueous solution and found no influence in the pH range 3.0–9.0. Even autoclaving for 5 h does not influence ouabain concentration when the pH of the solution is set to 7.0, which is naturally convenient for its intravenous application. Tajiri et al. [25] studied the effect of direct illumination on the stability of ouabain injections and found no impact. Endeavoring to achieve total synthesis of ouabain, Overman and Rucker [26] studied the degradation of ouabain by strong acidic hydrolysis to several products.

Results and discussion

Published HPLC methods for the determination of ouabain or structurally related k-strophanthin are based on reverse phase and gradient elution with (1) a mixture of aqueous ammonium formate or acetate and methanol [27, 28] or (2) a mixture of formic acid or ammonium formate in water and acetonitrile [29, 30]. Due to the polar character of ouabain, we developed HILIC on an amide-based stationary phase. The mobile phase consisted of acetonitrile and water. The ratio of both mobile phase constituents was investigated in the range of acetonitrile:water (v/v) = 90:10, 85:15, and 80:20. The best resolution was achieved at mobile phase of acetonitrile:water = 85:15 (v/v). The higher amount of acetonitrile in the mobile phase in comparison with RP-HPLC is more suitable for connection with mass spectrometric detection, because it provides a better ionization efficiency of our analytes in MS. The addition of a formic buffer (0.02 M ammonium formate and formic acid of pH = 4.00) into the mobile phase did not improve resolution of peaks. The influence of the flow rate of mobile phase on resolution of ouabain and other constituents of the analyzed sample was tested in the range 0.2–0.8 cm3/min. The flow rate of 0.2 cm3/min enables satisfactory resolution of peaks in chromatogram.

Figure 2 shows the HILIC–MS chromatograms of the analyzed injection solutions under the optimized conditions. It is obvious that retention times slightly shifted; however, relative standard deviation of the retention time was up to 7% which is common value for HILIC methods. The identification of ouabain was performed by comparison with standard; other constituents were identified using mass spectrometry (Table 2).
Fig. 2

HILIC–MS chromatograms of the analyzed “Strophosan” injection solution in ESI a positive and b negative modes. For peak identification, see Table 2. S denotes peak of the mass calibration standard (sodium formate). XBridge® BEH Amide column (150 × 3.0 mm i.d., particle size 2.5 μm; Waters), mobile phase acetonitrile:water = 85:15 (v/v), and flow rate 0.2 cm3/min

Table 2

Identification of the peaks in the HILIC–MS chromatograms of the analyzed “Strophosan” injection solution

Peak number

tr/min

ESI mode

Ion type

m/z

Molecular formula

Compound

Experimental

Theoretical

Δ/ppm

1

2.98

Neg.

[M − H]

110.9774

110.9758

− 14.5

CH3O4S

Not identified

2

3.77

Neg.

[M − H]

179.0718

179.0714

− 2.6

C10H12O3

Propyl salicylate

3

3.89

Neg.

[M − H]

151.0406

151.0401

− 3.4

C8H8O-3

Methylsalicylic acid

4

4.21

Neg.

[M − H]

137.0250

137.0244

− 4.0

C7H6O3

Salicylic acid

5

8.60

Neg.

[M + Cl]

601.2408

601.2421

− 2.2

C29H42O11

Ouabain oxidation product

6

8.63

Pos.

[M + H]+

258.2787

258.2791

1.9

C16H35NO

Not identified

7

8.79

Pos.

[M + H]+

230.2474

230.2478

1.7

C14H32NO

Not identified

8

9.29

Pos.

[M + H]+

130.1571

130.1590

14.6

C8H19N

Not identified

9

10.25

Pos.

[M + H]+

396.041

Not identified

10

11.06

Pos.

[M + 2H]2+

455.517

Not identified

11

11.67

Pos.

[M + 2H]2+

241.2040

241.2056

7.0

C33H52O2

Cholesterol derivative

12

12.43

Pos.

[M + H]+

585.2903

585.2901

0.4

C29H44O12

Ouabain

12

13.00

Neg.

[M + Cl]

619.2541

619.2527

− 2.3

C29H44O12

Ouabain

Quantification of ouabain in the analyzed “Strophosan” injection solution was based on the calibration curve. It was found 56 ± 2% (sr = 1.75%) of declared ouabain content, i.e., 0.14 mg/cm3 of ouabain. Its degradation product is probably compound found as peak number 5 in the chromatogram. It is derivative of ouabain with oxidized hydroxy group, probably on position 10 in the steroid ring. Unfortunately, its concentration is so low that the structure could not be adequately confirmed by MS2 and it is only hypothesis supported by the comparison with literature [31, 32]. Some of the other found compounds (salicylates and sodium chloride) come from manufacture of preparation and were added to maintain osmotic concentration of the injection solution. The chemical constitution of some found compounds was not determined due to their trace concentration.

The ESI-MS2 spectrum of ouabain in negative mode has been studied in a detail as the mass spectra of ouabain are inadequately published in the literature. Light et al. [33] presented the characterization of cardiac glycosides by the desorption/ionization mass spectrometric technique potassium ion ionization; for ouabain, they give only ions: m/z = 623 ([M + K]+) and m/z = 605 ([M–H2O + K]+). Tracqui et al. [30] published only main representative ion-spray MS ions for characterization of ouabain: m/z = 585 ([M + H]+), m/z = 607 ([M + Na]+), and m/z = 623 ([M + K]+). Seeking to supplement the literature data, both ESI modes for acquisition of MS2 spectra of ouabain were tested, from which negative mode gives much more abundant spectrum. Thus, the ion m/z = 583.2760 ([M − H]) was selected as the precursor ion of ouabain, its MS2 spectrum is given in Fig. 3. The individual product ions are formed mainly by elimination rhamnosyl group and molecule(s) of water. The mechanism of ouabain ESI fragmentation is proposed in Fig. 4. The proposal of rhamnosyl group fragmentation to ions of m/z = 119.0357 or 103.0400 is based on the analogy with literature [34]. The location of double bonds in the product ion of m/z = 419.2080 is only probable hypothesis, as the double bond could also be formed between C2 and C3.
Fig. 3

ESI-MS2 spectra of the [M − H] ion of ouabain

Fig. 4

Proposed ESI-MS2 fragmentation of [M − H] ion of ouabain

Conclusion

A new HILIC method was developed for determination of ouabain and applied to the 80-year-old injection solution of this glycoside. Ouabain was found to be half degraded to product with oxidized hydroxy group. To the best of our knowledge, the analysis of such old pharmaceutical preparation containing glycoside has not yet been described in the literature. In addition, the ESI-MS2 fragmentation mechanism of ouabain was proposed.

Experimental

Analyzed historical preparation sample Strophosan is an injection solution manufactured by the Dr. Robert Heisler company (formerly Czechoslovakia) in 1938. The declared content of ouabain is 0.25 mg/cm3. The analyzed preparation was preserved in the Czech Pharmaceutical Museum in Kuks, and it has been assumed that it was stored in dark, at room temperature for all time. The colorless solution was stored in a sealed glass ampoule. The ampoule was opened directly prior to the analysis. pH of the solution was measured after opening the ampoule as 6.9. An aliquot of the solution was appropriately diluted with the mobile phase.

Ouabain octahydrate standard was purchased from Sigma-Aldrich. The other chemicals used were acetonitrile, ammonium hydroxide solution (≥ 25%), ammonium formate, formic acid, and sodium formate; they were of HPLC or p.a. grade, and purchased from Sigma-Aldrich.

An UHPLC chromatograph Nexera XR (Shimadzu, Japan) connected with a Compact QTOF Bruker mass spectrometer (Bruker, Germany) with ESI ionization was used. The separation was performed on an XBridge® BEH Amide column (150 × 3.0 mm i.d., particle size 2.5 μm; Waters); the column temperature was maintained at 40 °C. The mobile phase was composed of acetonitrile and water and pumped with the flow rate of 0.2 cm3/min. The volume of injected sample was 2 mm3. The compounds were detected using a mass spectrometer; ionization of the analytes was performed in the positive and negative ion mode at the capillary voltage 2.8 kV. The pressure of the nitrogen (nebulizing gas) was set to 0.40 bar. Nitrogen (4.0 dm3/min) also served as drying gas at 250 °C. The scan range was set to 50–1000 m/z. Ouabain was quantified using six level calibration curve of the standard in the concentration range 5–200 mg dm−3, as the analytical signal served SIM of deprotonated molecule of ouabain of m/z = 583.2760.

Notes

Acknowledgements

The financial support by the projects SVV and Progress Q46 of Charles University is gratefully acknowledged.

References

  1. 1.
    Kravetz RE (2005) Am J Gastroenterol 100:2614CrossRefPubMedGoogle Scholar
  2. 2.
    Carstensen JT, Rhodes C (2000) Drug stability. Principles and practices, 3rd edn. Marcel Dekker, New YorkGoogle Scholar
  3. 3.
    Yoshioka S, Stella VJ (2002) Stability of drugs and dosage forms. Kluwer, New YorkGoogle Scholar
  4. 4.
    Nesměrák K, Kudláček K, Babica J (2017) Monatsh Chem 148:1557CrossRefGoogle Scholar
  5. 5.
    Lyon RC, Taylor JS, Porter DA, Prasanna HR, Hussain AS (2006) J Pharm Sci 95:1549CrossRefPubMedGoogle Scholar
  6. 6.
    Cantrell L, Suchard JR, Wu A, Gerona RR (2012) Arch Intern Med 172:1685CrossRefPubMedGoogle Scholar
  7. 7.
    Greene J (2018) Ann Emerg Med 71:13ACrossRefGoogle Scholar
  8. 8.
    Jordi Gonzalez R (1989) Farm Clin 6:451Google Scholar
  9. 9.
    Nesmerak K, Sticha M, Cvancarova M (2010) Anal Lett 43:2572CrossRefGoogle Scholar
  10. 10.
    Zilker M, Sörgel F, Holzgrabe U (2018) J Pharm Biomed Anal 150:318CrossRefPubMedGoogle Scholar
  11. 11.
    Kudláček K, Nesměrák K, Štícha M, Kozlík P, Babica J (2017) Monatsh Chem 148:1613CrossRefGoogle Scholar
  12. 12.
    Doggett NS, Spencer PSJ (1971) Br J Pharmacol 42:242CrossRefPubMedPubMedCentralGoogle Scholar
  13. 13.
    Erdmann E, Greef K, Skou JC (1986) Cardiac glycosides 1785–1985: biochemistry, pharmacology, clinical relevance. Springer, New YorkCrossRefGoogle Scholar
  14. 14.
    Reichstein T (1965) Planta Med 13:382CrossRefGoogle Scholar
  15. 15.
    Geiger UP, Weiss E, Reichstein T (1967) Helv Chim Acta 50:179CrossRefPubMedGoogle Scholar
  16. 16.
    Klaassen CD (1973) Toxicol Appl Pharmacol 24:37CrossRefPubMedGoogle Scholar
  17. 17.
    Osseo-Asare AD (2008) Soc Hist Med 21:269CrossRefGoogle Scholar
  18. 18.
    Reddy MS, Zhang H, Phoenix S, Deslongchamps P (2009) Chem Asian J 4:725CrossRefPubMedGoogle Scholar
  19. 19.
    Hamlyn JM, Blaustein MP (2016) Hypertension 68:526CrossRefPubMedPubMedCentralGoogle Scholar
  20. 20.
    Haskell CC, Doeppers WA (1914) J Am Pharm Assoc 3:646Google Scholar
  21. 21.
    Pomeroy CA, Heyl FW (1922) Am J Pharm 94:479Google Scholar
  22. 22.
    Levy RL, Cullen GE (1919) J Exp Med 31:267CrossRefGoogle Scholar
  23. 23.
    Haag HB, Hatcher RA (1929) J Am Med Assoc 93:26CrossRefGoogle Scholar
  24. 24.
    Berry H (1935) Q J Pharm Pharmacol 8:464Google Scholar
  25. 25.
    Tajiri Y, Dosako K, Matsumoto Y, Goya K, Kurata M, Shinohara C (1978) Hiroshima-ken Byoin Yakuzaishikai Gakujutsu Nenpo 13:54Google Scholar
  26. 26.
    Overman LE, Rucker PV (2000) Heterocycles 52:1297CrossRefGoogle Scholar
  27. 27.
    Vaclavik L, Krynitsky AJ, Rader JI (2014) Anal Chim Acta 810:45CrossRefPubMedGoogle Scholar
  28. 28.
    Yamaguchi H, Miyamori K, Sato T, Ogura J, Kobayashi M, Yamada T, Mano N, Iseki K (2014) J Chromatogr B 972:73CrossRefGoogle Scholar
  29. 29.
    Grosa G, Allegrone G, Del Grosso E (2005) J Pharm Biomed Anal 38:79CrossRefPubMedGoogle Scholar
  30. 30.
    Tracqui A, Kintz P, Ludes B, Mangin P (1997) J Chromatogr B 692:101CrossRefGoogle Scholar
  31. 31.
    Templeton JF, Hamlyn JM, Hamilton BP, Ayotte J, Majgier-Baranowska H, Lester A, Marat K, Perreault H (2000) Steroids 65:379CrossRefPubMedGoogle Scholar
  32. 32.
    Strobach H, Wirth KE, Rojsathaporn K (1986) Naunyn Schmiedebergs Arch Pharmacol 334:496CrossRefPubMedGoogle Scholar
  33. 33.
    Light KJ, Kassel DB, Allison J (1989) Biomed Environ Mass Spectrom 18:177CrossRefPubMedGoogle Scholar
  34. 34.
    Carroll JA, Willard D, Lebrilla CB (1995) Anal Chim Acta 307:431CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Austria, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Department of Analytical Chemistry, Faculty of ScienceCharles UniversityPragueCzech Republic
  2. 2.Department of Chemistry, Faculty of ScienceCharles UniversityPragueCzech Republic
  3. 3.Czech Pharmaceutical Museum, Faculty of Pharmacy in Hradec KrálovéCharles UniversityKuksCzech Republic

Personalised recommendations