Trans-Cyclooctene Isomerization Catalyzed by Thiamine Degradation Products in Cell Culture Media

Abstract

PET imaging of intrathecally dosed ASO neurology drugs is challenging due to the long time needed to achieve steady-state brain distribution, making the use of a simple ¹⁸F tag impossible with its short radioactive half-life. To overcome this challenge, a pretargeted imaging solution has been developed in which an ASO tagged with a tetrazine is dosed intrathecally, and after 24 h a reactive trans-cyclooctene (TCO) tagged with ¹⁸F is dosed intravenously. The two molecules form a click-chemistry adduct, allowing for PET imaging scans immediately following ¹⁸F-TCO administration. Although it has been demonstrated that TCOs can be relatively stable in vivo, they rapidly isomerize to cis-cyclooctenes (CCOs) in cell culture media and "aged" plasma, making many DMPK experiments challenging to interpret and not representative of the in vivo stability. The predominant cause of isomerization was determined to be thiamine degradation product(s) in media such as DMEM. Several techniques to overcome the challenges of in vitro and ex vivo isomerization during analytical experiments are herein proposed, such as the use of custom media and/or fresh plasma, adding antioxidants, using surrogate molecules, and using TCO trapping agents. These findings and techniques may also be relevant to other applications in which TCOs are incubated in thiamine-containing cell culture media.

1

PET imaging click-chemistry strategy for CNS-targeting ASOs.

2

Structures of TCOs tested in this study.

3

Isomerization of (A) compound B TCO (blue trace) to CCO (red trace) in (B) DMEM (using UV peak area, with chromatogram traces provided in the inset) and (C) aged plasma (using MS-XIC peak area).

4

Four-hour incubation of Compound B in DMEM containing 50 mM ascorbic acid (AA). The TCO half-life was equivalent to that in media which did not contain ascorbic acid. The low level of CCO was present in the standard material and did not increase during incubation. Equivalent results were observed for Compound A.

5

Isomerization of Compound B TCO to CCO in phosphate buffer containing (A) cysteine, (B) hydrogen peroxide, and (C) cysteine and hydrogen peroxide (with UV chromatogram traces provided in the inset). The thiyl radical formed from the oxidation of cysteine/cystine (D) is the presumed cause of isomerization in this experiment, but not likely significant in the isomerization of TCO in media.

6

Degradation products of thiamine, as summarized by Schnellbaecher et al. (displayed with permission from the author, copyright 2019 by Wiley). Product 38 (5-hydroxy-3-mercapto-2-pentanone) is shown to be responsible for TCO isomerization using an analogue compound in the present study. Products 39 and 42 may also contribute to the TCO isomerization.

7

Isomerization of Compound B in phosphate buffer containing 0.12 μM 3-mercapto-2-pentanone (structure shown, equivalent to 1% of the thiamine concentration in DMEM), which is a mercapto-ketone analog compound to the thiamine degradation product 38. UV chromatograms are shown on the right, color coded by 6 different incubation times (blue→purple→red→yellow→green→cyan).

8

LC-MS extracted ion chromatograms (XIC) of human hepatocyte incubations of the CCO analog of Compound A in (A) DMEM and (B) custom DMEM without vitamins; and (C) the proposed metabolic scheme.

9

Quenching of TCO after incubation (to prevent further chemical isomerization) and subsequent aromatization in red blood cells, as observed in Compounds A and B.