where IF= fluorescence intensity and cheme = heme concentration. The Kheme constant is a complex term that is directly proportional to the excitation beam intensity, quantum yield, heme extinction coefficient (εheme), and path length (14). This relationship holds at absorbances below 0.05, which can be assumed with the 40-fold dilution of the standard solutions prior to analysis. The analytical sensitivity of the fluorescence assay was unexpectedly close to the sensitivity of the CLARiTY and less than half the sensitivity of the pyridine hemochromogen assay (Table 1). Analytical sensitivity is dependent on statistical variance and is thus indicative of changes in fluorescence signal with environmental conditions. Surprisingly, the CLARiTY demonstrated a lower normalized LOD (LODN) than the fluorescence assay. That is, given at least 1 mL of sample, the CLARiTY and pyridine hemochromogen assays are superior to the fluorescence assay at low concentrations. At higher concentrations, the greater linear range and analytical sensitivity of the pyridine hemochromogen assay make it the preferred approach. Indeed, the fluorescence assay protocol employed here is only unmatched in its ability to quantitate extremely small samples. Heme and hemoglobinization in mouse blood
As with the hemoglobin standards, the pyridine hemochromogen assay was used to determine heme levels in mouse blood dilutions (Figure 3A). Blood was collected from one male and one female C57BL/6 mouse on separate days and then analyzed (Table 2). With the exception of the female MCH (17.2 pg/cell), HGB and MCH values for both mice fell well within the normal ranges for these indices (25). Considering established inaccuracies in the hemocytometer relative to automated means (26), the elevated female MCH was likely due to an erroneously low cell count.
Plots of fluorescence (Figure 3B) and corrected CLARiTY absorbance at 410 nm (Figure 3C, Supplementary Figure S1C) against heme concentration generated slope values equivalent to Kheme and εheme, respectively (Table 2). While we expected εheme values for the CLARiTY to be the same for both mice, the lower value for the female mouse could also be explained by a lower than actual cell count. Regardless of the deviation, assuming that hemoglobin is saturated with heme and that it is the only hemoprotein present in the cells allows us to calculate εHb as 4 × εheme and directly convert absorbance/ cm (Beer-Lambert law) or fluorescence intensity (Equation 2) to hemoglobin concentration. In this way, the εHb for the CLARiTY was found to be 0.46 ± 0.01 µM-1 cm-1. Of further note, visually identified LOQR values for the CLARiTY and fluorescence calibration curves were similar to the values for hemoglobin standards. LOLR values, however, were significantly higher than the hemoglobin standards for both methods, suggesting that detection parameters are analyte-specific. Hemoglobin quantitation in differentiating K562 and MEL cells
K562 cells are known to differentiate irreversibly along the erythroid lineage, expressing markers such as glycophorin A and synthesizing unusually large quantities of embryonic and fetal globins when treated with inducing agents such as sodium butyrate (27). MEL cells induced to differentiate by DMSO also exhibit erythroid-specific membrane antigens (28), an increase in globin mRNA (29), and a decrease in cell size (30) not observed in K562 cells.
Applying the same approach used to analyze mouse blood, pyridine hemochromogen assays of MEL and K562 cells that were induced to differentiate with DMSO (5 days) and sodium butyrate (7 days), respectively, or left untreated (uninduced) were used to determine heme concentrations for each culture. Corrected absorbances from the CLARiTY plotted against the heme concentrations for MEL and K562 cells generated slope values that were not significantly different between the two cell types (Figure 4A). All samples were thus pooled to derive a single value for εHb of 0.462 ± 0.004 µM-1 cm-1 (Table 3), which is equivalent to εHb for pooled mouse blood (Table 2) to 2 decimal places. Incidentally, we decided to include cell sample data that fell outside of the defined linear range (LOQR–LOLR) for the hemoglobin standards. We felt this was justified for two reasons: (i) no sigmoidal trend was observed in the cell culture analysis, and (ii) the CV for the standards was less than 20% at 0.11 µM heme, well below the minimum observed sample concentration.