Direct Analysis of Morphine, M3G and M6G Metabolites, and Related Compounds in Urine by LC-MS/MS

Abstract

The analysis of total morphine is typically conducted by first subjecting urine samples to acid or enzymatic hydrolysis in order to cleave the glucuronide conjugates from morphine’s primary metabolites [morphine-3β-D-glucuronide (M3G) and morphine-6β-D-glucuronide (M6G)] prior to analysis. With the glucuronide moiety removed, the resulting morphine molecule is much less polar and, therefore, more amenable to traditional reversed-phase chromatography. However, both hydrolysis procedures cost an analyst time and result in sample variability due to incomplete hydrolysis or analyte conversion. By utilizing the retention characteristics of the Restek Force Biphenyl column, hydrolysis was not required and direct analysis of morphine, its M3G and M6G metabolites, and several related compounds using a simple “dilute-and-shoot” sample preparation was performed.

Introduction

Accurate analysis of morphine is critical to monitoring patient adherence to prescribed pain management drug regimens, and it also plays an important role in forensic investigations that target the illicit use of morphine, heroin, and codeine. Heroin and codeine can both be metabolized into morphine, so detection of morphine levels that are outside of the normal therapeutic range can indicate abuse of these related drugs. The analysis of total morphine is typically conducted by first treating urine samples with an acid or enzymatic hydrolysis step in order to remove the glucuronide conjugates from the primary morphine metabolites [morphine-3β-D-glucuronide (M3G) and morphine-6β-D-glucuronide (M6G)]. With the glucuronide conjugate moiety removed, the resulting morphine molecule is much less polar and, therefore, more amenable to traditional reversed-phase chromatography.

While either acid or enzymatic hydrolysis can improve analyte retention for reversed-phase LC, it is a time-consuming sample preparation procedure and has several drawbacks. Acid hydrolysis provides the most efficient conversion of glucuronide conjugates into morphine, but the resulting samples can corrode metal components in the analytical instrumentation if the samples are not appropriately neutralized following hydrolysis. Enzymatic hydrolysis with β-glucuronidase requires that the enzyme be removed prior to analysis in order to avoid on-column protein precipitation and also to prevent any additional hydrolysis from occurring during analysis. The time and temperature at which samples are incubated with β-glucuronidase are highly source and substrate dependent, which can result in incomplete hydrolysis and contribute significantly to sample variability. Additionally, both types of hydrolysis have been shown to convert the unique metabolite of heroin, 6-monoacetylmorphine (6-MAM), into morphine, which is a critical loss in forensic testing for the presence of heroin.

Hydrolysis procedures require a significant amount of analyst time and introduce a large source of error into the analysis of morphine. In this article, we demonstrate that the retention of the Force Biphenyl column allows the direct analysis of morphine, its M3G and M6G metabolites, and related compounds in urine following a simple and robust “dilute-and-shoot” sample preparation method. Direct analysis of M3G, M6G, and morphine allows these analytes to be reported separately, which means time-consuming and variable hydrolysis steps can be avoided.

Experimental

Sample Preparation Method 1: Enzyme Hydrolysis

Calibration standards and quality control (QC) samples were prepared by fortifying human urine with morphine, M3G, M6G, morphine-N-oxide, 6-MAM, and hydrocodone. 6-MAM and hydrocodone were included in the assay as they are metabolites of heroin and codeine, respectively. Eight calibration standards were prepared across a range of 25–2,500 ng/mL and QC samples were prepared at 25, 75, 750, and 2,000 ng/mL. Experimental samples were prepared at 750 ng/mL for all analytes and were subjected to hydrolysis by 20 μL of β-glucuronidase from abalone (≥100,000 units/mL) or to hydrolysis by 20 μL of purified β-glucuronidase from abalone (150,000-250,000 units/mL). Six replicate samples were prepared using each enzyme solution. A pH 4.5 acetate buffer was added to the calibration and QC samples as an enzyme solution surrogate. Aliquots of 150 μL of calibration standards, QC samples, and enzyme-treated samples were incubated with 10 μL of a working internal standard (10,000 ng/mL of stable isotope labeled deuterated analogs in 0.1% formic acid in water) for 16 hours at 60 °C. The quantitation of morphine-N-oxide was performed using morphine-d3 due to the lack of availability of labeled morphine-N-oxide. Following incubation, 600 μL of acetonitrile was added to precipitate the enzyme (for consistent treatment, acetonitrile was also added to the enzyme-free calibration standards and QC samples). Samples were then vortexed and centrifuged at 4,500 rpm for 10 min. The supernatant was then dried down under nitrogen and reconstituted in 1 mL of 0.1% formic acid in water prior to analysis.

Sample Preparation Method 2: Dilute-and-Shoot

Calibration standards and quality control samples were prepared by fortifying human urine with morphine, M3G, M6G, morphine-N-oxide, 6-MAM, and hydrocodone at the same levels as in the hydrolysis experiment. The samples were diluted 10-fold in a working internal standard solution (50,000 ng/mL of stable isotope labeled deuterated analogs prepared in 0.1% formic acid in water) and were vortexed and centrifuged prior to analysis. The quantitation of morphine-N-oxide was performed using morphine-d3 due to the lack of availability of labeled morphine-N-oxide.

LC-MS/MS Analysis

Analysis of M3G, M6G, morphine, and related compounds was conducted on a Shimadzu Nexera UHPLC equipped with a SCIEX API 4500 MS/MS. Detection was performed using electrospray ionization in positive ion mode with multiple reaction monitoring (MRM). A diverter valve was used prior to the MS/MS in order to direct the initial 0.5 minutes of the matrix front to waste. Reversed-phase separations were performed using water and methanol mobile phases modified with 0.1% formic acid under gradient conditions with a Restek Force Biphenyl 1.8 μm, 50 mm x 2.1 mm column equipped with an UltraShield UHPLC precolumn filter with a 0.2 μm frit. Instrument conditions are detailed below and analyte transitions are shown in Table I.

Table I: Analyte Transitions for Analysis of M3G, M6G, Morphine, and Related Compounds

 

Results and Discussion

As shown in Figure 2, the amount of M3G and M6G detected in fortified urine samples varied between the purified and unpurified enzymatic hydrolysis treatment groups. After 16 hours of incubation at 60 °C, it was observed that >99% of M3G had been converted into morphine for both enzymes (Table II). Purified β-glucuronidase was able to convert >99% of M6G into morphine while only 80% was converted by the less pure enzyme. For both enzymes, only approximately 400 ng/mL of 6-MAM was detected indicating that a significant 6-MAM to morphine conversion had occurred. Interestingly, both enzymes resulted in the detection of elevated levels of hydrocodone (940 ng/mL). Total morphine concentration was only coincidentally close to the nominal value of 2,250 ng/mL (assumes 100% conversion of M3G and M6G to morphine) using the less pure enzyme due to the incomplete hydrolysis of M6G coupled with the conversion of a significant amount of 6-MAM to morphine and total morphine was above nominal value using the purified enzyme due to conversion of 6-MAM to morphine. Morphine-N-oxide levels were not significantly impacted by the hydrolysis procedures. Because linearity and QC accuracy/precision testing passed acceptance criteria, the results presented here reflect the impact of the hydrolysis procedure, not the performance of the analytical method.

Table II: Enzyme Hydrolysis Results

Dilute and Shoot

The chromatogram in Figure 3 demonstrates that good chromatographic results were obtained on the Force Biphenyl column for all compounds in terms of retention, resolution, peak shape, and peak response. It is notable that attempting a similar analysis on a C18 or phenyl-hexyl column would typically require high aqueous conditions, which reduces MS sensitivity. In contrast, the Force Biphenyl column provides enough retention to avoid matrix interference under a gradient starting at 15% organic mobile phase, which provides much more sensitivity than isocratic analysis at 5% organic or even less.

As shown in Table III, the dilute-and-shoot method provided excellent performance for the direct analysis of morphine, its M3G and M6G metabolites, and related compounds in urine. Linearity was established in the range of 25-2,500 ng/mL for all analytes. Using 1/x² weighted linear regression for morphine-N-oxide and 1/x weighted linear regression for the remaining analytes, all compounds showed good linearity with r² values of 0.996 or greater with relative standard deviations of <10%. The lowest signal-to-noise ratio observed at the LLOQ across all analytes was 49:1, indicating that this method could be modified to increase the dilution factor or decrease the LLOQ as necessary. Analysis of blank urine showed no observed matrix interferences.

Table III: Dilute and Shoot Linearity, Accuracy, and Precision Results

 

Conclusion

The hydrolysis of M3G and M6G back into morphine is a time-consuming sample preparation step that is necessary if the analytical column used does not provide adequate retention of these compounds in urine. In this study, we demonstrated that the retention characteristics of the Restek Force Biphenyl column allow for the direct analysis of morphine, its M3G and M6G metabolites, and several related compounds without hydrolysis, which significantly simplifies sample preparation and provides more accurate results compared to traditional hydrolysis methods.