Determination Of The
Iron Content Of Hemoglobin
1998 BE 210 Project
W2
Marc Chodock
Benjamin Fleischer
Ava Segal
Kimberly Kirby
April 30, 1998
Dr. Mitchell Litt
Table of Contents
Section / Page NumberAbstract / 2
Background / 3
Methods and Materials / 7
Results / 10
Discussion / 14
References / 17
Abstract
The iron fraction in hemoglobin (Hb) was calculated using atomic absorption spectrophotometry (AAS) to be 0.32% (p = 0.04) by mass. The number of Fe atoms/molecule Hb was calculated from the literature molecular weight of Hb (mw = 64.5 kd[1]) to be 3.65. The literature values are 0.34%[2] and 4.00[3] respectively. Assuming the literature value of 4.00 atoms of iron per molecule Hb, the MHb = 70.5 kd compared to the literature value of MHb = 64.5 kd. The 95% confidence intervals allow this deviance.
Hemoglobin was found to affect the absorbance curve for iron upon nebulization (p = 0.16). The percent of iron lost per ppm of hemoglobin was calculated using linear regression. The function was used to correct the [Fe] calculated in hemoglobin and yield a new mass fraction of 0.32%. It was hypothesized that the decrease in concentration at higher values of Hb was due to a decrease in flow rate and subsequent decrease in the effective [Hb] measured.
Alternative methods of removing the hemoglobin from solution were explored. Hemoglobin solutions were made by adding dilute tricholoroacetic acid (TCA) in one trial and nitric acid in another. In the TCA solutions ranging from 2.5 to 10% of 6 M, an average of 85.10% of the iron was removed as precipitate. The nitric acid, ranging in concentration from 2.5 to 10% of 6 M, caused 77.83% of the iron was removed as precipitate. The benefit of acid treatment is therefore in the precipitation of protein for removal rather than in releasing iron from the heme group. As an application, the experimental mass fraction was used to calculate the hemoglobin content of coagulated horse serum to be 1628 ppm.
Background
Hemoglobin, Iron, and Atomic Absorption Spectrophotometry
Hemoglobin (Hb) is a versatile protein. Its main functions include regulating the blood pH[4], maintaining blood pressure by nitric oxide transport[5], and bringing oxygen from the lungs to tissues while removing carbon dioxide[6]. The iron, in the ferrous (+2) state, is responsible for this latter task. Therefore, cell metabolism and organism health is directly correlated to the hemoglobin content of serum and therefore also to the iron content.
Figure 1: Structure of heme[7].
Iron in Hb is bound to a heme group, a protoporphyrin IX ring (Figure 1). There are four heme groups per Hb and thus four iron molecules.[8] The iron may be removed from the heme group when the Hb (even whole blood) is diluted in water and burned in an acetylene flame.[9] An atomic absorption spectrophotometer (AAS) can then be used to measure absorbance. The AAS emits light at a specific wavelength (=248.3 nm for iron) from an iron lamp which is specifically absorbed by iron. A calibration curve may be constructed for a known standard solution to obtain the functional relationship between absorbance and concentration from which future concentrations may be evaluated.[10]
The Beer-Lambert Law describes the ideally linear relationship between concentration and absorbance of a monochromatic wavelength. However, pure monochromatic energy is not produced and the curve is not linear. Only in the linear region of the curve is Beer’s Law upheld, and therefore only this region should be considered.[11] The linear region occurs at dilute ion concentrations. This region usually occurs at less than 20 ppm for the AAS.
AAS is a specific, accurate, and fast method for the determination of iron content of a solution.[12] Zettner demonstrated that whole blood dilutant yields the same values as toluene hemolysed samples. Further, samples showed no matrix interference effects at less than 1% aqueous concentration. There were also no apparent interference effects from other inorganic constituents or chelators. AAS values are well correlated to the standard cyanmethemoglobin spectroscopy method[13]. Therefore, the simple measurement of iron in diluted Hb as proposed in this lab proves sufficient.
Hemoglobin Statistics
Hemoglobin accounts for more than 65% of body iron (total body iron being 3.5 g) and more than 10% of blood mass. Serum storage iron (serum ferritin) and serum iron (mostly transferrin-bound) account for about 25% and 0.1% of total body iron respectively. These three sources account for more than 90% of body iron[14]. Therefore, hemoglobin iron is 72% of all blood iron. By determining the iron content of serum, the amount of hemoglobin can also be determined. Table 1 shows the normal ranges of iron content in humans.
Table 1: Ranges of Iron Content in Humans[15].
Hb iron/kg body weight (mg) / Body iron /kg body weight (mg) / Hb concentration(g/dL blood)
Men / 31 / 50 / 14-18
Women / 28 / 40 / 12-15
Recovery Spiking
Recovery spiking tests either for interference by the protein upon nebulization of the sample or interference within the flame[16]. Zettner concluded that whole blood has no interference effects upon absorbance. But, since Zettner’s AAS was different from the one we used, the Hb may still have an effect upon the flow rate. If the flow rate for the iron standard (used in the calibration) is faster than for the Hb solution, the measured concentration will be less than the actual. To test for this effect, a known concentration of the iron standard (Ferrous Nitrate) was added (spiked) to a known concentration of Hb. If any correction is necessary, it will be expressed as the difference between the expected iron recovered from the spike and the amount actually recovered.
Acid Analysis
In the event that the recovery spiking demonstrates Hb interference, it would be beneficial to find a less complicated method of measuring the actual concentrations. Therefore, on the advice of Al Giandomenico (lab coordinator), dilute acid solutions could be used to release the iron from Hb. The Hb would precipitate and no longer interfere in the reading. The dilute acid would not have appreciable hazards.
Biomedical Relevance
People are susceptible to anemia, a deficiency of hemoglobin (Hb) in the blood. An iron deficiency caused either by lack of iron or inability to absorb iron causes a decrease in the amount of hemoglobin produced and a reduction of Hb in erythrocytes. Otherwise, a disability to produce erythrocytes or erythrocyte loss, for example, through menstruation, causes a low count of erythrocytes in the blood[17].
In a 1986 review, Dallman reported that a 95% level of the reference Hb range for age and sex is indicative of iron-deficiency anemia.Once iron-deficiency anemia has been identified, the iron treatment increases hemoglobin synthesis to almost normal levels within a month[18]. Therefore, a method that quickly and easily identifies iron-deficiency anemia is imperative.
According to a 1997 paper which Dallman co-authored, iron deficiency anemia consists of having iron deficiency in addition to a low Hb count. Iron deficiency is diagnosed as having an abnormal value for at least 2 of 3 laboratory tests of iron status (erythrocyte protoporphyrin, transferrin saturation, or serum ferritin) as determined by the National Health and Nutrition Examination Survey (NHANES II). Iron deficiency anemia is relatively common in the United States among women of childbearing age and toddlers. Extrapolated from the US Bureau of the Census data, of the 7.8 million adolescent girls and women of childbearing age with iron deficiency, approximately 3.3 million have iron deficiency anemia. Approximately 700,000 iron deficient toddlers have iron deficiency anemia[19]. Knowing the iron content of Hb allows quantification of the amount of Hb in serum if the concentration of iron in solution is determined. Thus, a person can easily be diagnosed as having iron-deficiency anemia.
Material and Apparatus
- Perkin Elmer Model AA 4000 Atomic Absorption Spectrophotometer
- Acetylene
- Air
- Bovine Hemoglobin, Sigma#H2500
- Iron standard, 1000 ppm
- Coagulated Horse Blood
- Trichloroacetic Acid (TCA)
- Nitric Acid (HNO3)
- Deionized H2O purified to 18 M resistance with four-cartridge purification system (Millipore)
- Parafilm
- Volumetric flask
- Pipettes
Methods And Procedure
Making the Iron Standard and Calibrating the AAS
The iron solution was made by dissolving 1.000 0.0005 gram of iron powder in nitric acid, and then diluting to 1.000 0.003 liter with deionized H2O. The Perkin-Elmer AA (PE-AA) wavelength source, a dual-beam iron lamp, was set to =248.3 nm. The current was 30 mA and the slit was 0.2 L[20]. A calibration curve for the PE-AA was made using varying dilutions of the iron standard and plotting concentration versus absorbance.
[Fe]/[Hb] Determination and Recovery Spiking
Adding 22 mg and 33 mg each to 25 mL of deionized water made hemoglobin concentrations of 880 ppm and 1330 ppm, respectively. Initial Fe absorbance readings were taken of the solutions and their concentrations determined.
The recovery spiking solutions (1,2,3, and 4) were prepared according to Table 1. The columns specify the concentration of the 5 mL spike and the rows specify what 5 mL solution is being spiked. The iron samples were prepared by diluting the iron standard with deionized water. Iron absorbances were measured of the four samples.
Table 2: Determination of Spiking Concentrations.
2.4 ppm Fe / 4.2 ppm Fe / 7.2 ppm Fe / 12.6 ppm Fe880 ppm Hb / (1) / (2)
1330 ppm Hb / (3) / (4)
[Fe]/[Hb] Determination with Trichloroacetic acid (TCA) and Nitric Acid
A 2368 ppm Hb solution was made by adding 59 mg Hb to 25 mL of deionized water. Using a 6.2 M TCA, 5 mL dilutions of 5%, 10%, 15%, and 20% were prepared along with another 5 mL 20% solution to be used as a blank. The same concentrations were prepared from 20 % 6 M nitric acid. The final solutions and their preparation are shown in table 2. The top row acid was mixed with the 2368 ppm Hb to yield the indicated solutions.
Table 3: Method of Acid Preparation.
5 mL @ / 0% TCA (HNO3) / 5% TCA (HNO3) / 10% TCA (HNO3) / 15% TCA (HNO3) / 20% TCA (HNO3)1184 ppm Hb / 1184 ppm Hb / 1184 ppm Hb / 1184 ppm Hb / 1184 ppm Hb
2368 ppm Hb
0% TCA (HNO3) / 2.5% TCA (HNO3) / 5% TCA (HNO3) / 7.5% TCA (HNO3) / 10% TCA (HNO3)
The solutions were then thoroughly mixed and placed in the centrifuge. The centrifuge was set at 8oC, 2500 rpm for 10 minutes. The solution was decanted from the precipitates that had formed at the top and bottom and analyzed. Iron absorbancies were measured for each solution.
Hb Determination in Horse Serum
Coagulated horse serum[21] was obtained. A 10 mL aliquot was centrifuged at 8°C, 2500 rpm for 15 minutes. The supernatant was then analyzed in the AAS. The reading was within the linear range on the calibration curve and therefore did not need to be diluted.
Results
Table 4:Calibration of Iron Standard Concentration vs. Absorbance.
Slope of A vs. [Fe] / R Square /P-value
/ Lower 95% / Upper 95%0.031 / 0.978 / 6.67E-05 / 0.028 / 0.034
Ferrous Nitrate solutions below 12 ppm were found to remain in this linear region.
Table 5: Determination of Percentage of Iron in Hemoglobin.
%Fe by mass / R Square / P-value / Lower 95% / Upper 95%0.30% / 0.969 / 0.027 / 0.14% / 0.46%
The percent iron by mass was determined using the relationship of the concentrations of iron vs. hemoglobin. The data (880 and 1330 ppm Hb without additional iron) is displayed in Table 3.
Table 6: Recovery Spiking of Hemoglobin Solutions.
Hb Solution / [Fe] Found (ppm) / [Fe] Expected (ppm) / % Recovery Error880 ppm Hb / 2.485
+ 2.4 ppm Fe / 2.39 / 2.443 / -2.27%
+ 7.2 ppm Fe / 4.91 / 4.843 / 1.30%
1330 ppm Hb / 4.088
+ 4.2 ppm Fe / 3.92 / 4.144 / -5.30%
+ 12.4 ppm Fe / 7.78 / 8.344 / -6.72%
The iron concentrations were calculated using the previously determined calibration curve. The experimental values were generally less than what was expected
Table 7: Calibration of Percent Recovery Error vs. Hb Concentration.
Coefficients (y=m*x+b) / R Square / P-value / Lower 95% /Upper 95%
Intercept(b) / -0.112 / 0.816 / 0.155 / -0.329 / 0.104X-variable(m) / 0.134 / 0.096 / -0.059 / 0.326
The data from Table 3 shows a negative relationship between Hb concentration and percent error, however, the x-variable is positive, because the correction is based on the amount found not the amount expected.
Table 8:Recalculation ofDetermination of Percentage of Iron in Hemoglobin.
Corrected % Fe by mass / R Square / P-value / Lower 95% / Upper 95%0.32% / 0.935 / 0.042 / 0.05% / 0.58%
Using the regression line from the data in Table 4, the percent recovery error was used along with the initially calculated iron concentration at the two hemoglobin solutions to calculate the adjusted iron concentration. Using the new iron concentrations a new relationship between the masses of iron and hemoglobin was calculated with a difference of 0.02%. The new mass fraction of iron in hemoglobin is 0.32%
Table 9:Determination of Iron released from Hemoglobin Using HNO3.
[HNO3] added / % Iron Removed With Precipitate2.50% / 76.54%
5.00% / 77.41%
7.50% / 79.08%
10.0% / 78.30%
Average / 77.83%
A 20% HNO3 blank was used to detect any iron in the HNO3 solution. The percentage of iron removed with precipitate in each solution was calculated by subtracting one from the percentage of iron remaining in solution. The percentage of iron remaining was found by subtracting the corrected iron concentration found in the blank from the corrected iron concentration of the solution and dividing the difference by the corrected iron concentration from the 0% HNO3 solution.
Table 10: Regression of % Iron Removed in Precipitate vs. % HNO3 of solution
Coefficients (y=m*x+b) / R Square / P-value / Lower 95% /Upper 95%
Intercept(b) / 0.761 / 0.664 / 1.6E-04 / 0.720 / 0.802X-variable(m) / 0.278 / 0.185 / -0.323 / 0.879
This data shows that there is a small positive relationship (p = 0.185) between the nitric acid concentration and the amount of iron removed in precipitate. The R square value (.664) shows that it is a non-linear relationship.
Table 11: Determination of Iron released from Hemoglobin Using TCA.
[TCA ] added / % Fe released from Hb2.5% / 85.52%
5.0% / 85.23%
7.5% / 84.96%
10% / 84.72%
Average / 85.10%
The same procedure as in Table 7 was followed.
Table 12: Regression of % Iron Recovered vs. % TCA of solution
Coefficients (y=m*x+b) / R Square / P-value / Lower 95% /Upper 95%
Intercept(b) / 0.858 / 0.999 / 5.E-08 / 0.857 / 0.859X-variable(m) / -0.107 / 0.001 / -0.119 / -0.095
This data shows that there is a small negative relationship (p = 0.001) between the TCA concentration and the amount of iron removed in precipitate. The R square value (.999) shows that it is a highly linear relationship.
Table 13: Comparison of Literature Values to Calculated Values
Atoms Fe per molecule Hb / Molecular weight Fe (dalton) / Molecular weight Hb (dalton) / Mass Fraction of Fe in HbLiterature values: / 4.00 / 55.6 / 64,500 / 0.34%
Calculated values: / 3.71 / [Not Calculated] / 69,500 / 0.32%
The literature mass fraction was calculated by dividing the mass of iron in Hb by the literature mass of Hb. The experimental value was 0.32% (Table 8). The experimental value (0.32% Fe/Hb) was used to solve for the calculated values. For example, 3.71=(0.32%)(64,500)/(55.6). These values are not linearly independent. They are merely indicative of ways one could solve an unknown.
Table 14:Determination of Hemoglobin Concentration in Horse Serum.
Assume Iron is from Hb / Assume iron is not from Hb[Hb] Horse (ppm) / 334 / 1628
After using atomic absorption to find the iron concentration of the solution, the previously determined [Fe]/[Hb] relationship (0.32%) was used to determined the concentration of Hb in the serum along with the approximation that 72% of blood iron is from Hb[22]. With the possibility that the centrifugation removed the hemoglobin from solution and left only the other sources of iron, another calculation was made. In the second calculation, the iron found was 28% of the total, and the Hb content was determined to be (1.46 ppm Fe)/(28% Fe not in Hb)/(.32% Fe/Hb) = 1628 ppm.
Discussion
[Fe]/[Hb] Determination and Recovery Spiking
In this lab we first discovered that the mass fraction of iron in hemoglobin was 0.30%. This is an 12% deviation from the literature 0.34%[23] value. All data used to arrive at 0.30% Fe/Hb was significant (p<0.027) and highly linear (R2>0.96). To assess whether the deviation is due to a change in flow rate as described in the background (Recovery Spiking), a recovery spike was performed. The percent difference between the amount of iron expected and the amount recovered (Table 6) demonstrates that hemoglobin has a diminishing effect upon iron recovery. Zettner determined that hemoglobin and other chelators do not interfere with iron absorbance in an acetylene-air flame[24]. At a constant concentration of Hb, an increasing spike concentration increased the recovery loss. An increase in Hb concentration also increased[25] the amount of recovery loss. Therefore, it was determined that the Hb is responsible for the recovery loss by slowing down the flow rate of the AAS. A linear regression was done to correlate percent recovery loss of [Hb] (Table 7). The correlation produced was used to correct the iron concentrations reported in Table 6 and a new regression of [Fe] vs. [Hb] yielded the improved iron mass fraction of 0.32% (Table 8). This value only has a 6% deviation from the 0.34% calculated value based on assumptions from the literature. Furthermore, though all data used in its calculation was significant (p<0.05) the confidence intervals were quite large. The 95% confidence interval for the mass fraction is.0.01% to 0.58%, which is due to the small number of trials, rather than imprecise data, as shown by the p-values and R2. The confidence interval is only indicative that there is some “breathing space” for the data. The primary conclusion to be made is that as few as four trials can give highly accurate values using the AAS method.
[Fe]/[Hb] Determination with Trichloroacetic acid (TCA) and Nitric Acid
The low percentages of iron recovered by HNO3 and TCA (Tables 9 and 11) indicate that most of the iron remained in the precipitated Hb. Visual analysis of the solutions determined that not all the precipitate had been removed and some had become suspended in the solution. Based on this analysis, the iron that was recovered is thought to represent either the iron in suspended hemoglobin or the iron freed from the protein. Assuming that most of the Hb did precipitate and that the iron measured was from the suspended Hb, it was determined that acid-precipitation of Hb is not an effective means of releasing iron. Therefore, this test could not be used to calculate the mass percentage of iron in hemoglobin. This defeats the original purpose of this experiment, but demonstrates an easy method for removing protein from solution. A longer time centrifuging would likely have increased the amount of Hb removed from solution.
While small relationships exist between the concentration of the acids and the amount of iron released (p < 0.001 for TCA, p < 0.19 for HNO3), it appears that the amount of iron released is nearly independent of the concentration of the dilute acid. While the nitric acid regression has a positive slope, the range of its 95% confidence values for the slope (-0.323, 0.879) is both above and below 0. The 95% confidence limits for TCA’s regression do not fall below 0, but the slope is very close to zero (-0.107).
The nearly horizontal nature of the slopes illustrates that there is at most a very small relationship between the concentration of a dilute acid and the amount of iron released. With this conclusion, the average amount of iron released seems to be a much more relevant number for the dilute acid. Comparing the two averages proves that HNO3 precipitates less well than TCA.
An application of this method could be the determination of the concentration of hemoglobin in serum. For this application, it must be assumed that when an acid solution is mixed with a hemoglobin solution it can be centrifuged to a point where none of the precipitate is suspended. In this case the precipitate should consist of hemoglobin and other blood proteins. A purification assay to separate the molecules in the precipitate would allow the hemoglobin to be collected, dried, and massed. The concentration of hemoglobin in the serum can be easily determined.