Methemoglobin determination by multi-component analysis in coho salmon (Oncorhynchus kisutch) possessing unstable hemoglobin

Abstract

Hemoglobin derivatives are often quantified in blood to establish cardio-respiratory status and possible causes of impaired oxygen transport. The derivative known as methemoglobin results from oxidation of hemoglobin and is pathologically relevant because it cannot transport oxygen. In species and individuals possessing unstable methemoglobin, methemoglobin formation leads to rapid hemichrome formation and precipitation. Oxidizing reagents in standard methemoglobin analysis techniques therefore prevent accurate quantification of hemoglobin oxidative degradation products in species possessing unstable hemoglobin. In this study, we demonstrated that individual coho salmon (Oncorhynchus kisutch) possess unstable methemoglobin. Because molar absorptivities of coho methemoglobin, hemichrome and carboxyhemoglobin were significantly different from humans, the use of previous standard methods leads to an overestimation of methemoglobin in coho. Spontaneous conversion of methemoglobin to hemichrome was also demonstrated in Chinook (O. tshawytscha), pink (O. gorbuscha) and chum salmon (O. keta), but not steelhead (O. mykiss), indicating there may be a frequent need to account for unstable hemoglobin when quantifying methemoglobin in salmonids.•Our method builds upon multi-component analysis (MCA) by using a multivariate modeling technique to derive the coho-specific molar absorptivities of major hemoglobin derivatives•This approach fills a current need for the accurate quantification of methemoglobin in fishes possessing unstable hemoglobin.

Keywords: Coho salmon; Fish hemoglobin; Hemichrome; Methemoglobin; Multi-component analysis; Unstable hemoglobin.

Graphical abstract

Fig. 1

Absorbance spectral time series of coho hemoglobin solution after oxidation with a slight excess of potassium ferricyanide (solid lines). Arrows indicate the direction of peak changes. Modeled pure coho methemoglobin spectra (dashed lines) and pure hemichrome spectra (dashed-dotted lines) are overlaid. The spectral time series shows a shift in absorbance pattern from that most similar to pure methemoglobin, towards that of hemichrome over a period of around 18 min (inset) before the sample begins to precipitate (dotted lines), which is recognized as a rise in the baseline of the spectra and loss of isosbestic points.

Fig. 2

Chemical regression goodness-of-fit test results for one modeled spectral series. (A) The predicted relative decreasing concentrations of unstable coho methemoglobin (dashed line) and increasing concentrations of hemichrome (solid line) based on the first-order reaction model show close agreement with calculated concentrations (circles and x's) using the determined rotation matrix and time series absorption spectra. (B) Residuals between actual and predicted absorbance measurements throughout the conversion of coho methemoglobin to hemichrome are slight, affirming goodness-of-fit.

Fig. 3

The millimolar absorptivities of major adult coho hemoglobin derivatives. Coho absorptivities are shown in the spectral range of 450–700 nm for oxyhemoglobin (OHb), deoxyhemoglobin (HHb), carboxyhemoglobin (COHb), methemoglobin (MetHb) and hemichrome (HemiHb) prepared in phosphate buffered saline at pH 8.0. Prespawn coho blood samples collected (n = 4) from the Quilcene hatchery (Quilcene, WA, USA).

Fig. 4

Comparison of percent coho methemoglobin results using the current coho multi-component analysis (MCA) method and a human multi-wavelength (MW) technique against a control method based on mixtures of 0% and 100% coho methemoglobin stock solutions, which were analyzed for iron content using ICP-MS. The coho MCA method (crosses) performed better than the human MW method (circles) following Benesch et al. using human molar absorptivities published by van Assendelft and Zijlstra to match assay pH conditions of 8.0. The solid line shows a 1:1 relationship between the control and spectroscopic methods. Relative percent differences (RPD) are shown for the Benesch et al. MW method above the line, and for the coho MCA method below the line. The adult coho blood sample was collected from the Skookumchuck hatchery (Tenino, WA, USA).