Jensen Et Al. (Supplemental Material)

Jensen et al. (Supplemental Material)

Results

GC-MS matrix effect, assay coefficient of variation, and method validation.

Prior to executing the experiments on brain tissue, we conducted a variety of control studies to validate the accuracy of our GC-MS approach to quantify neurosteroids in dissected brain tissues. As with our recent report, we used deuterated estradiol as the ITSD, since the retention time for the ITSD was in the mid-range of retention times for the 10 neurosteroids, with an easily resolved and separate peak from other steroids (see Table 1 in Snelling et al., 2014 and justification on choice of ITSD). Solid phase extraction of the dextran-coated charcoal stripped brain samples effectively removed any discernable matrix effects (no deviation > 10% of the nominal value interpolated from solvent only standards) from the analyzed samples. The matrix effect did not interfere with the quantification of the 10 analytes of interest, and the limit of quantification of our analytes were comparable to our initial assay in plasma per 2L aliquot (2 pg for the 3,5/3,5-androstanediols and 3,5/3,5-androsterones, 10pg for DHEA, ALLO, pregnanolone, pregnenolone, 3,5-THDOC, and 20 pg for 3,5-THDOC). Additionally, there was no interference or co-elution in the gross scale chromatograms or in the retention period groups at the m/z (mass to charge) monitored for the brain samples. Intra-assay and inter-assay coefficients of variation ranged from 3.00 – 7.41% and from 3.05 – 8.53%, respectively (Supplemental Table 1), thereby confirming the accuracy and reliability of our methodology. Notably, the coefficients of variation were lower than our initial assessments in plasma (Snelling et al., 2014), indicating that the refinements to the GC-MS assay improved precision.

For method validation, calibration curves were formed by spiking 300 L of ddH2O or stripped brain homogenate with a subset of reference standards (0, 70, 300, and 700 pg) of all 10 analytes and 10 ng ISTD. The concentrations of reference standards were similar to Snelling et al. (2014). The spiked samples were then extracted, derivatized, and analyzed by GC-MS. Calibration curves for each neurosteroid were formed by plotting the ratio of the response of the analyte divided by the response of the ISTD versus the concentration of the analyte. Linear regression analysis of each calibration curve confirmed excellent linear response for all ten analytes. The r2 values ranged from 0.9987 to 0.9996, with the exception of DHEA, where the r2 was 0.9982 (Supplemental Figure 1). Then, the subsequent abundances for the spiked ddH2O and stripped brain homogenates were integrated and interpolated against the concurrently full calibration curve (see Methods), and the resulting interpolated values were plotted against the spiked amount (Supplemental Figure 2). Linear regression analysis of the interpolated values from the four spiked concentrations in the brain homogenates also confirmed excellent linear response for all ten analytes. The range in r2 values was identical to those in the calibration curves depicted in Supplemental Figure 1. Notably, the interpolated values of all neurosteroid analytes were statistically similar or identical to the spiked amount. The precision and accuracy of the interpolated values compared to the calibration curve nominal values confirm the high level of confidence in the optimized assay.

Figure Legends

Supplemental Figure 1. Calibration curves for each neurosteroid. A subset of reference standards (0, 70, 300, and 700 pg) were plotted against the ratio of the response of the analyte over the response of the internal standard (ISTD) per 2 L sample. Depicted are results from a representative experiment that has been repeatedly performed with similar values. The r2 values from the linear regression analysis are shown for each neurosteroid. Due to differences in ratios, neurosteroids are depicted in two panels. The top panel contains the 3,5/3,5-androstanediols and 3,5/3,5-androsterones. The bottom panel contains dehydroepiandrosterone (DHEA), allopregnanolone (ALLO), pregnanolone, pregnenolone, and the 3,5/3,5-THDOCs (tetrahydrodeoxycorticosterones).

Supplemental Figure 2.Validation of brain assay by determination of spiked samples.Supernatant from stripped brain homogenate (300 L) was spiked with internal standard (ISTD) and a subset of reference standards (0, 70, 300, and 700 pg) of each neurosteroid and extracted in the same manner as the reference standards. Depicted are the interpolated values (obtained amount, solid triangle with solid line, from a standard curve of each analyte), which are plotted with the value for the amount added to the sample (spiked amount, open square and dashed line). Linear regression analysis was conducted on the interpolated values of the obtained amounts, and the r2 values (> 0.99 for all neurosteroids) confirmed the excellent linear response for all neurosteroids. DHEA = dehydroepiandrosterone; THDOC = tetrahydrodeoxycorticosterone

Supplemental Figure 1

Supplemental Figure 2

Supplemental Table 1. Intra- and inter-assay precision of GC–MS analysis of neurosteroids

Intra-assay (%CV) / Inter-assay (%CV)
70 pg / 300 pg / 700 pg / 70 pg / 300 pg / 700 pg
Neurosteroid
Pregnenolone / 4.22 / 6.50 / 4.71 / 4.93 / 5.58 / 4.24
DHEA / 5.76 / 6.48 / 3.24 / 7.27 / 5.76 / 6.87
Allopregnanolone / 3.55 / 5.07 / 7.23 / 6.42 / 7.89 / 5.42
Pregnanolone / 6.44 / 3.00 / 3.01 / 6.92 / 7.52 / 6.28
3α,5α-Androstanediol / 5.71 / 6.00 / 7.38 / 8.53 / 5.26 / 7.79
3α,5β-Androstanediol / 6.23 / 4.94 / 4.09 / 5.16 / 6.05 / 6.06
3α,5α-Androsterone / 4.95 / 3.68 / 5.22 / 3.86 / 8.36 / 6.32
3α,5β-Androsterone / 6.25 / 4.68 / 6.32 / 6.54 / 8.38 / 7.96
3α,5α-THDOC / 3.06 / 7.41 / 5.33 / 7.46 / 6.82 / 8.36
3α,5β-THDOC / 5.21 / 5.58 / 5.36 / 3.05 / 5.71 / 6.62

For intra-assay variability, three replicates of samples (dextran-coated charcoal stripped samples spiked with 70, 300, or 700pg) were assayed along with the calibration standards in the same run. Intra-assay variability was calculated by comparing the interpolated values for the standards when they were used as unknowns with the actual value. Inter-assay variability was obtained by comparing three different analytical runs of samples (dextran-coated charcoal stripped samples spiked with 70, 300, or 700pg). Coefficients of variation (CV) were calculated as standard deviation×100, divided by the mean.

DHEA = dehydroepiandrosterone; THDOC = tetrahydrodeoxycorticosterone

Supplemental Table 2. ANOVA results for cortical neurosteroids at 8 h in male WSP-1 and WSR-1 mice

Steroid / Main effect: Line / Main effect: Treatment / Interaction
Pregnenolone / F(1,47)=2.80, p=0.10; WSP>WSR / F(1,47)=0.17, p=0.68 / F(1,47)=2.78, p=0.10
DHEA / F(1,46)=5.87, p<0.05; WSP>WSR / F(1,46)=0.23, p=0.64 / F(1,46)=2.09, p=0.15
Allopregnanolone (3,5-THP) / F(1,48)=3.19, p=0.08; WSP>WSR / F(1,48)=0.05, p=0.82 / F(1,48)=5.94, p<0.05
Pregnanolone (3,5-THP) / F(1,47)=15.07, p<0.001; WSP>WSR / F(1,47)=0.30, p=0.59 / F(1,47)=1.83, p=0.18
3,5-Androstanediol / F(1,48)=3.40, p=0.56 / F(1,48)=4.76, p<0.05; air<ethanol / F(1,48)=0.38, p=0.54
3,5-Androstanediol / F(1,47)=0.69, p=0.41 / F(1,47)=1.23, p=0.27 / F(1,47)=3.36, p=0.07
3,5-Androsterone / F(1,46)=3.04, p<0.09; WSP>WSR / F(1,46)=1.07, p=0.31 / F(1,46)=13.75, p=0.001
3,5-Androsterone / F(1,47)=6.40, p<0.05; WSP>WSR / F(1,47)=1.13, p=0.29 / F(1,47)=9.75, p<0.01
3,5-THDOC / F(1,48)=0.57, p=0.45 / F(1,48)=1.25, p=0.27 / F(1,48)=4.73, p<0.05
3,5-THDOC / F(1,48)=1.12, p=0.30 / F(1,48)=5.39, p<0.05; air>ethanol / F(1,48)=6.41, p<0.05

Two way ANOVA results at the 8 h time point are shown for the 10 neurosteroids (and in the same order) that are depicted in Figure 1.

DHEA = dehydroepiandrosterone; THDOC = tetrahydrodeoxycorticosterone

Supplemental Table 3. ANOVA results for hippocampal neurosteroids at 8 h in male WSP-1 and WSR-1 mice

Steroid / Main effect: Line / Main effect: Treatment / Interaction
Pregnenolone / F(1,42)=3.24, p<0.08; WSP>WSR / F(1,42)=0.82, p=0.37 / F(1,42)=0.18, p=0.68
DHEA / F(1,47)=2.47, p<0.12; WSP>WSR / F(1,47)=0.36, p=0.55 / F(1,47)=0.92, p=0.34
Allopregnanolone (3,5-THP) / F(1,43)=0.12, p=0.73 / F(1,43)=0.89, p=0.35 / F(1,43)=0.24, p=0.63
Pregnanolone (3,5-THP) / F(1,41)=0.10, p=0.75 / F(1,41)=0.58, p=0.45 / F(1,41)=2.59, p=0.11
3,5-Androstanediol / F(1,47)=6.59, p<0.05; WSP>WSR / F(1,47)=0.90, p=0.35 / F(1,47)=0.98, p=0.33
3,5-Androstanediol / F(1,44)=4.37, p<0.05; WSP>WSR / F(1,44)=0.14, p=0.71 / F(1,44)=2.74, p=0.10
3,5-Androsterone / F(1,46)=0.90, p=0.35 / F(1,46)=1.76, p=0.19 / F(1,46)=0.06, p=0.81
3,5-Androsterone / F(1,43)=4.23, p<0.05; WSP>WSR / F(1,43)=0.59, p=0.45 / F(1,43)=1.44, p=0.24
3,5-THDOC / F(1,43)=0.002, p=0.96 / F(1,43)=0.06, p=0.81 / F(1,43)=0.58, p=0.45
3,5-THDOC / F(1,46)=1.67, p=0.20 / F(1,46)=2.98, p=0.09; air>ethanol / F(1,46)=0.008, p=0.93

Two way ANOVA results at the 8 h time point are shown for the 10 neurosteroids (and in the same order) that are depicted in Figure 2.

DHEA = dehydroepiandrosterone; THDOC = tetrahydrodeoxycorticosterone

Supplemental Table 4. ANOVA results for cortical neurosteroids at 48 h in male WSP-1 and WSR-1 mice

Steroid / Main effect: Line / Main effect: Treatment / Interaction
Pregnenolone / F(1,32)=0.02, p=0.90 / F(1,32)=0.02, p=0.88 / F(1,32)=0.05, p=0.82
DHEA / F(1,32)=0.17, p=0.68 / F(1,32)=2.40, p=0.13 / F(1,32)=0.90, p=0.35
Allopregnanolone (3,5-THP) / F(1,30)=3.01, p=0.09; WSR>WSP / F(1,30)=0.82, p=0.37 / F(1,30)=0.23, p=0.63
Pregnanolone (3,5-THP) / F(1,33)=1.45, p=0.24 / F(1,33)=0.04, p=0.84 / F(1,33)=0.32, p=0.58
3,5-Androstanediol / F(1,32)=0.15, p=0.70 / F(1,32)=1.13, p=0.30 / F(1,32)=0.27, p=0.61
3,5-Androstanediol / F(1,32)=0.22, p=0.64 / F(1,32)=0.15, p=0.70 / F(1,32)=6.67, p<0.05
3,5-Androsterone / F(1,33)=1.80, p=0.19 / F(1,33)=3.33, p<0.08 / F(1,33)=0.02, p=0.90
3,5-Androsterone / F(1,32)=5.03, p<0.05; WSP>WSR / F(1,32)=0.63, p=0.43 / F(1,32)=4.41, p<0.05
3,5-THDOC / F(1,32)=1.32, p=0.26 / F(1,32)=2.83, p=0.10 / F(1,32)=1.92, p=0.18
3,5-THDOC / F(1,34)=0.38, p=0.54 / F(1,34)=0.41, p=0.53 / F(1,34)=1.07, p=0.31

Two way ANOVA results at the 48 h time point are shown for the 10 neurosteroids (and in the same order) that are depicted in Figure 3.

DHEA = dehydroepiandrosterone; THDOC = tetrahydrodeoxycorticosterone

Supplemental Table 5. ANOVA results for hippocampal neurosteroids at 48 h in male WSP-1 and WSR-1 mice

Steroid / Main effect: Line / Main effect: Treatment / Interaction
Pregnenolone / F(1,33)=36.44, p<0.001; WSR>WSP / F(1,33)=6.03, p<0.05; air>ethanol / F(1,33)=3.18, p=0.08
DHEA / F(1,34)=0.32, p=0.58 / F(1,34)=7.62, p<0.01; air<ethanol / F(1,34)=8.76, p<0.01
Allopregnanolone (3,5-THP) / F(1,34)=8.98, p<0.01; WSR>WSP / F(1,34)=0.66, p=0.42 / F(1,34)=0.41, p=0.53
Pregnanolone (3,5-THP) / F(1,34)=2.29, p=0.14 / F(1,34)=11.22, p<0.01; air<ethanol / F(1,34)=0.94, p=0.34
3,5-Androstanediol / F(1,33)=3.42, p=0.07; WSR>WSP / F(1,33)=1.01, p=0.32 / F(1,33)=7.28, p=0.01
3,5-Androstanediol / F(1,33)=5.31, p<0.05; WSP>WSR / F(1,33)=4.24, p<0.05; air>ethanol / F(1,33)=1.13, p=0.30
3,5-Androsterone / F(1,33)=0.39, p=0.54 / F(1,33)=3.26, p=0.08; air>ethanol / F(1,33)=0.11, p=0.75
3,5-Androsterone / F(1,34)=4.04, p=0.05; WSP>WSR / F(1,34)=4.98, p<0.05; air<ethanol / F(1,34)=0.01, p=0.95
3,5-THDOC / F(1,34)=1.36, p=0.25 / F(1,34)=0.07, p=0.79 / F(1,34)=0.02, p=0.88
3,5-THDOC / F(1,34)=4.63, p<0.05; WSP>WSR / F(1,34)=0.47, p=0.50 / F(1,34)=1.15, p=0.29

Two way ANOVA results at the 48 h time point are shown for the 10 neurosteroids (and in the same order) that are depicted in Figure 4.

DHEA = dehydroepiandrosterone; THDOC = tetrahydrodeoxycorticosterone

Supplemental Table 6. ANOVA results for plasma neurosteroids at 48 h in male DBA mice

Steroid / Main effect: Treatment
Pregnenolone / F(1,20)=6.98, p<0.05; air<ethanol
DHEA / F(1,20)=223.13, p<0.001; air>ethanol
Allopregnanolone (3,5-THP) / F(1,20)=16.71, p=0.001; air>ethanol
Pregnanolone (3,5-THP) / F(1,20)=27.42, p<0.001; air<ethanol
3,5-Androstanediol / F(1,20)=10.06, p<0.01; air<ethanol
3,5-Androstanediol / F(1,20)=18.78, p<0.001; air>ethanol
3,5-Androsterone / F(1,20)=7.88, p=0.01; air<ethanol
3,5-Androsterone / F(1,20)=84.00, p<0.001; air>ethanol
3,5-THDOC / F(1,20)=4.04, p=0.058; air>ethanol
3,5-THDOC / F(1,20)=1.14, p=0.30

ANOVA results at the 48 h time point are shown for the 10 neurosteroids (and in the same order) that are depicted in Figure 8.

DHEA = dehydroepiandrosterone; THDOC = tetrahydrodeoxycorticosterone