Anda di halaman 1dari 5

ANALYTICAL

BIOCHEMISTRY
Analytical Biochemistry 341 (2005) 199–203
www.elsevier.com/locate/yabio

A spectrophotometric method for estimating hemin


in biological systems
María Elisa Lombardo, Lidia Susana Araujo, Alejandra Beatriz Ciccarelli, Alcira Batlle ¤
Centro de Investigaciones sobre PorWrinas y PorWrias–CIPYP (CONICET-FCEN, UBA), Ciudad Universitaria Pabellón II 2do.Piso,
1428 Buenos Aires, Argentina

Received 10 June 2004


Available online 2 December 2004

Abstract

Hemin chlorides exhibit two absorption maxima in the Soret region, one at about 360–380 nm (S⬘ band) and the other between
400 and 430 nm (S band). We present here a simple and fast spectrophotometric assay to determine concentration of hemin
between 1.15 and 9.20 M employing the Soret region (S⬘ band) as a reference. In this method the hemin is quantitatively
extracted from biological materials by acidiWed chloroform. By recording the absorbance of the chloroform extract at its
maximum peak at 388, 450, and 330 nm and applying the correction formula Ac D 2A388 ¡ (A450 + A330), a very good linear corre-
lation between the Ac and the concentration of hemin is attained. The method can be used to estimate hemin in the presence of
protein (0.06–5.00 mg/ml) and porphyrin (0.19–2.97 M). Compared with the pyridine hemochromogen method, the assay
reported here is highly reproducible, with 15- to 30-fold more sensitivity, and it allows the quantiWcation of four times lower
hemin concentrations.
 2004 Elsevier Inc. All rights reserved.

Keywords: Hemin quantiWcation; Pyridine hemochromogen method; S⬘ Soret band; Hemin extraction

Heme is a chelate of protoporphyrin with a ferrous ther, the requirement of heme for protein synthesis in
ion. It is very readily autoxidized to ferriprotoporphyrin. mammalian reticulocytes and in a number of noneryth-
This ferric complex has one residual positive charge roid cells is well established [3,4]. On the other hand
which may be neutralized with chloride ion, forming a hemin is an obligatory requirement for growth of heme-
pentacoordinate square pyramidal complex often called deWcient organisms such as trypanosomatids [5]. In
hemin chloride or simply hemin. Hemin is practically Leishmania donovani, heme is an important regulator in
insoluble in both water and organic solvents. It does, both proliferation and cellular transformation [2]. Our
however, freely dissolve in aqueous alkali due to the aim was to develop a method for quantiWcation of hemin
formation of what is traditionally referred to as hydrox- in diVerent biological systems, such as cell-free extracts,
yhemin or hematin [1]. growing medium of heme-deWcient organisms, and bio-
Heme in the form of hemin has been reported to act logical Xuids. So far hemin levels can be estimated by
as an intracellular regulator of systems utilizing oxygen high-performance liquid chromatography [6], by radio-
[2]. Thus, heme appears to be controlling the synthesis, activity emission if hemin has been labeled with 59Fe,
transport, and assembly of mitochondrial proteins of and by spectrophotometry with the absorbance being
nuclear origin, which are involved in respiration. Fur- recorded at 400 nm [7] or after its conversion to pyridine
hemochromogen [8].
*
Corresponding author. Present address: Viamonte 1881, 10 “A”,
Spectroscopic studies have shown that all tetrapyr-
CP-1056 Buenos Aires, Argentina. Fax: +54 11 4811 7447. roles in which the porphyrinic nucleus is fully conju-
E-mail address: batlle@mail.retina.ar (A. Batlle). gated, including porphyrins, metalloporphyrins, and

0003-2697/$ - see front matter  2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.ab.2004.11.002
200 Hemin quantiWcation in biological systems / M.E. Lombardo et al. / Anal. Biochem. 341 (2005) 199–203

hemeproteins, have a Soret (S)1 band and two bands in hemin concentration (M) and Wtted with a straight line
the visible spectrum usually identiWed as the  and  determined by a linear regression (Sigma Plot 2001 soft-
bands. In the pyridine hemochromogen method [8], ware). DiVerences between methods and graphical repre-
hemin concentration is calculated considering the diVer- sentations were analyzed using the Student t test, taking
ence in absorbance between the maximum of the  band p < 0.05 as signiWcant.
(557 nm) and the minimum occurring between the  and
 bands (541 nm). Hemins that are not covalently bound
to protein may be extracted from biological materials by Results
acidiWed organic solvents. When examining the spectro-
scopic properties of a variety of hemin chlorides in low- Optimal conditions for hemin quantiWcation
polarity solvents, the presence of two absorption peaks
in the Soret region, one at about 360–380 nm (S⬘ band) Preliminary experiments led us to establish the condi-
and the other between 400 and 430 nm (S band, as men- tions described under Materials and methods. By anal-
tioned above) is found [9]. On these grounds, we have ogy with the pyridine hemochromogen method a sample
developed a new method to quantify hemin based on the volume of 4 ml was utilized. Extraction with acidiWed
recording of the S⬘ Soret band, following extraction of chloroform was tested using variable volumes of chloro-
hemin in acidiWed chloroform. form (between 2 and 4 ml) acidiWed to pHs between 1
and 3.5; the optimum conditions for extracting hemin
were then using 2 ml of solvent, within the pH range of
Materials and methods 1–2. We have also found that a single extraction with
2 ml of acidiWed chloroform was as good as extracting
Hemin chloride (equine) and albumin (bovine) were twice with 1 ml each time. Actual extraction was carried
from Sigma Chemical (London, UK). Protoporphyrin out employing 2 ml chloroform and shaking for three
IX (PPIX) was from Porphyrin Products (Logan, UT, consecutive times for 5 s, waiting for the two phases to
USA). All other chemicals were of the highest purity separate before proceeding with the next shaking. The
commercially available. ratio of chloroform:sample of 2:4 (v:v) led to both
Standard solutions of hemin were prepared by dis- eYcient extraction and concentration of the sample.
solving the solid reagent either in the growing medium of The S⬘ Soret band was characterized after extraction
a heme-deWcient organism such as Trypanosoma cruzi of hemin under the optimum conditions already deter-
[10] or in 50 mM sodium phosphate buVer, pH 7.4. PPIX mined of a 23 M hemin solution dissolved in the grow-
was stored in 0.5 M KOH:ethanol (1:1, v:v). Standard ing medium. Fig. 1 shows the absorption spectra (A) and
solutions of PPIX were prepared in 5% HCl. Standard the proWle of the derivative order one of the absorption
solution of bovine albumin was prepared in 50 mM spectra (B). From the graph in Fig. 1B we can accurately
sodium phosphate buVer, pH 7.4. determine the wavelength of the absorption maximum
and minimum at the points where the derivative order
Assay conditions one is 0. At the same time the inXection points will now
be the maximum and minimum in the Wrst derivative
Four milliliters of hemin solution, 2 ml of 50 mM gly- spectrum. Consequently we can mark out the interval
cine–HCl buVer, pH 2.0, 0.1 ml of 4 N HCl (to adjust corresponding to the spectral trace due to the extracted
Wnal pH to around 2.0), 0.2 ml of 5 M NaCl, and 2 ml of hemin. In conclusion the absorption spectrum curve
chloroform were vigorously shaken in a test tube and starts at 330 nm; its maximum peak is at 388 nm and it
then centrifuged to separate the organic phase. ends at 450 nm with inXection points at 362 and 418 nm.
Extracted hemin was spectrophotometrically quantiWed Employing these absorbance data we have studied the
(Hewlett–Packard HP 8452A diode-array UV/visible relationship between the concentration of hemin
spectrophotometer) by measuring the absorbance at the (referred to a sample volume of 4 ml) and the four diVer-
wavelengths indicated below. ent manners of relating these absorbance values. In
Fig. 2 we demonstrate a very good linear correlation
Statistical analysis (r D 0.99–1.00) between the concentration of hemin and
the four absorbance relationships indicated in the legend
All data are expressed as means § standard deviations to this Wgure. DiVerent representations were compared
of three or four separate experiments, running in dupli- by the Student t test (Table 1). To illustrate an optimum
cate. The absorbance relations were plotted against graph showing a slope signiWcantly higher than others
(p < 0.05), we have selected the representation 2A380–
A330–A450 vs hemin concentration which showed the
1
Abbreviations used: PPIX, protoporphyrin IX; S, Soret; DMSO, di- highest sensitivity. Employing both the absorbance rela-
methyl sulfoxide. tion and the methodology here described, the recovery
Hemin quantiWcation in biological systems / M.E. Lombardo et al. / Anal. Biochem. 341 (2005) 199–203 201

Fig. 1. Absorption spectra (A) and derivative order one of the absorption spectra (B) corresponding to hemin extracted in chloroform. Absorbance
value was obtained in a Hewlett–Packard HP 8452A diode-array UV/visible spectrophotometer employing spectral bandwidth of 2 nm and integra-
tion time of 1 s. Derivative order one was recorded using HP 89531A MS-DOS (R) UV/VIS operating software (Hewlett–Packard).

Comparison with other spectrophotometric methods

So far, there are two spectrophotometric methods to


estimate hemin, recording absorbance at 400 nm [7] or
after its conversion to pyridine hemochromogen [8]. The
former can be used only to measure hemin dissolved in a
solvent not interfering with absorbance read at 400 nm;
so, hemin dissolved in growing medium cannot be quan-
tiWed with this procedure. Using a standard solution of
200 M hemin dissolved in 50 mM sodium phosphate
buVer, pH 7.4, both spectrophotometric methods obey
Beer’s law for hemin concentration ranging from 4.76 to
23.80 M, with the absorption at 400 nm method being
three times more sensitive (slope: 0.0331 § 0.0004 M¡1)
than the pyridine hemochromogen method (slope:
0.0108 § 0.0002 M¡1).
We have compared the method here developed with
Fig. 2. QuantiWcation of hemin extracted in chloroform. The extrac- the pyridine hemochromogen method using either
tion and data analysis were performed as described under Materials
growing medium or 50 mM sodium phosphate buVer,
and methods. DiVerent absorbance vs hemin concentration curves are
plotted: (䊊) 2A388–A330–A450, (䊉) A388, (䉮) 2A388–A362–A418, and (䉲) pH 7.4, as solvents. For the pyridine hemochromogen
A388–A330. method, depending on the solvents used the slope values
obtained were diVerent. The sensitivity of the assay in
Table 1 sodium phosphate buVer (slope 0.0108 § 0.0002 M¡1)
Comparison between diVerent absorbance relations and hemin con- was twice (p < 0.05) that obtained in culture medium
centrations
(slope 0.0058 § 0.0002 M¡1).
Absorbance relations Slope (M¡1) § SD Interception § SD Instead nearly superimposed straight lines were
2A388–A330–A450¤ 0.179 § 0.003 ¡0.018 § 0.019 obtained in both solvents for our chloroform extrac-
A388 0.108 § 0.002 ¡0.013 § 0.013 tion method. The medium slope value was 14 and 30
A388–A330 0.081 § 0.002 ¡0.020 § 0.010
times higher in sodium phosphate buVer and culture
2A288–A362–A418 0.089 § 0.003 ¡0.013 § 0.015
medium, respectively, than that recorded for the pyri-
Note. The slope and intercept of regression lines corresponding to Fig.
dine hemochromogen method. A very good linear cor-
2 are summarized.
¤
Statistically diVerent (p < 0.05) from the other plotted relations. relation (r D 0.98–1.00) was found in all cases.
Reproducibility for the chloroform extraction method
was 96.50 § 3.80%, the limit of detection was 0.072 § was better than that obtained for the pyridine hemo-
0.011, the range of quantiWcation was 1.15–9.20 M, and chromogen method.
there were high intraday and interday precision values
and excellent accuracy (r » 1). This absorbance relation- Interference of albumin in the estimation of hemin
ship is analogous to that employed for spectrophotomet-
ric determination of porphyrins, steroids, and glycine Taking into account that hemin determinations
[11]. would be performed often in the presence of consider-
202 Hemin quantiWcation in biological systems / M.E. Lombardo et al. / Anal. Biochem. 341 (2005) 199–203

able amounts of protein and that some hemin can be Interference of protoporphyrin in the estimation of hemin
withheld by the proteins precipitated after chloroform
addition, we have examined our method further and When hemin is in solution along with other porphy-
accordingly modiWed the experimental protocol to rins, addition of chloroform can extract hemin together
allow us to measure hemin in the presence of signiWcant with less carboxylic porphyrins (such as PPIX). Consid-
amounts of albumin. To evaluate the eVect of albumin ering that both Soret bands (S and S⬘) appear between
on the hemin quantiWcation, 1 ml of sample volume 330 and 450 nm (Fig. 3), the eVect of diVerent levels of
was replaced by 1 ml of dimethyl sulfoxide (DMSO) PPIX on hemin quantiWcation was investigated.
and extraction was then carried out with 3 ml of In Fig. 3 we can clearly appreciate the spectral behav-
chloroform. ior of chloroformic extracts obtained from hemin alone
In Table 2, the eVect of varying concentrations of (curve C) and of diVerent PPIX amounts in the presence
albumin between 6.25 £ 10¡2 and 5.00 mg/ml (equivalent (curves 1–5) and absence (curves 6–10) of hemin. For
to 0.95–75.76 M, considering a MW of 66 kDa for albu- this assay 0.5 ml 10 N NaOH were added before 2 ml
min) on the quantiWcation of a Wxed hemin concentra- HCl–glycine buVer, pH 2.0. When a mixture of hemin
tion of 5.75 M is shown. Independently on the bovine and PPIX was extracted with chloroform the resulting
albumin concentration, DMSO addition would beneWt absorption curve was the sum of the control and the
hemin extraction. individual curves for PPIX. Therefore the system can be
For bovine albumin concentrations of 0.19–5.00 mg/ analyzed as though it were a mixture of the separate por-
ml, with extraction in DMSO the presence of 80% of phyrins. Similar behavior was observed for the relation
total hemin can be quantiWed. In this case the hemin cal- 2A388–A330–A450 when the PPIX/hemin rate was varied
ibration curve in the presence of 1 mg/ml bovine albumin between 0.13 and 1.30 (data not shown). Therefore the
had a slope of 0.1308 § 0.005 M¡1 (20% lower than that interference of PPIX can be abolished by arithmetic sub-
obtained in the absence of bovine albumin). traction of the absorbance value corresponding to PPIX
alone.
Table 2
EVect of varying concentrations of albumin on hemin quantiWcation
Validity of the method
Albumin (mg/ml) 2A388–A330–A450
Without DMSO With DMSO In most biological systems hemin synthesis is Wnely
0 0.991 § 0.035 1.088 § 0.052 regulated, so that there is not any accumulation of this
6.25 £ 10¡2 0.776 § 0.032 1.001 § 0.040 metabolite whatsoever [1]. The method here described
12.50 £ 10¡2 0.696 § 0.025 0.939 § 0.045
has been tested in a system whereby hemin was exoge-
18.75 £ 10¡2 0.622 § 0.043 0.826 § 0.031
0.25 0.465 § 0.011 0.836 § 0.015 nously added, such as to a culture medium of heme-deW-
0.38 0.428 § 0.015 0.833 § 0.020 cient microorganisms in concentrations varying from 5
0.50 0.423 § 0.012 0.827 § 0.046 to 30 g/ml. The stability of a 15-g/ml hemin solution
1.00 0.439 § 0.023 0.783 § 0.025 dissolved in the growing medium and kept at 30 °C was
2.00 0.448 § 0.039 0.798 § 0.049
evaluated every 24 h for Wve days; in this instance hemin
5.00 0.422 § 0.020 0.794 § 0.035
concentrations from 14.90 § 1.83 to 2.82 § 0.40 g/ml
Note. A Wxed hemin amount (23 nmol/tube) was quantiWed in the pres- were quantiWed. The method was used with biological
ence of diVerent concentrations of albumin. The extraction with acidi-
Wed chloroform was carried out in the absence and the presence of
systems in which hemin was being synthesized, such as in
DMSO. Experimental conditions were as described under Materials the in vitro determination of ferrochelatase activity (EC
and methods. 4.99.1.1; ferrochelatase or protoheme ferrolyase, is the

Fig. 3. Absorption spectra of hemin (23.00 nmol/tube) extracted in the absence (C) or the presence of diVerent PPIX amounts (1, 0.76 nmol; 2,
2.29 nmol; 3, 3.82 nmol; 4, 5.73 nmol, and 5, 7.64 nmol). Curves of 6–10 represent the absorption spectra corresponding to the diVerent amounts of
PPIX extracted without hemin.
Hemin quantiWcation in biological systems / M.E. Lombardo et al. / Anal. Biochem. 341 (2005) 199–203 203

enzyme catalyzing the formation of heme from PPIX method here described could be very useful for control-
and Fe2+). Levels of ferrochelatase activity of about ling these processes.
6.23 § 0.75 and 19.36 § 1.89 nmol of hemin/mg of protein
were measured in free cell extracts of T. cruzi and Sac-
charomyces cerevisiae, respectively. The performance of Acknowledgments
the assay in quantifying the intracellular content of
hemin during growth of T. cruzi in medium containing María Elisa Lombardo and Alcira Batlle hold the
variable concentrations of hemin has also been tested post of ScientiWc Researchers at the Argentine National
and it has been found that after six days of growth, with Research Council (CONICET). Lidia Susana Araujo is a
0, 5, 15, and 30 g hemin/ml of medium, the content of Research Fellow of the University of Buenos Aires
hemin in a cell-free supernatant was 0.35 § 0.22, (UBA). Alejandra Beatriz Ciccarelli is a Research Fel-
4.79 § 1.38, 45.84 § 4.50, and 72.34 § 6.87 nmol/ml of low of the Agencia de Promoción CientíWca y Tecnológ-
supernatant, respectively. ica, Argentina. This work was supported by grants from
CONICET, UBA, and the Agencia de Promoción
CientíWca y Tecnológica.
Discussion
References
We have developed a simple and fast assay to deter-
mine concentrations of hemin between 1.15 and 9.20 M [1] A.M. del, C. Batlle, Biosíntesis del Hemo, in: PorWrias y PorWrinas.
employing the Soret region S⬘ band. Following the pro- Aspectos Clínicos, bioquímicos y biologí a molecular, Acta Bio-
cedure here described, recording the absorbance of quím. Clín. Latinoamericana Sup. 3, Argentina, 1997.
hemin extracted by acidiWed chloroform at its maximum [2] J.K. Pal, M. Joshi-Purandare, Dose-dependent diVerential eVect of
peak at 388, 450, and 330 nm and applying the correction hemin on protein synthesis and cell proliferation in Leishmania
donovani promastigotes cultured in vitro, J. Biosci. 26 (2001) 225–
formula Ac D 2A388 ¡ (A450 + A330), a very good linear 231.
correlation between the Ac and the concentration of [3] J.-J. Chen, I.M. London, Regulation of protein synthesis by heme-
hemin is attained even in the presence of protein (0.06– regulated eIF-2 kinase, Trends Biochem. Sci. 20 (1995) 105–108.
5.00 mg/ml) and porphyrin (PPIX 0.19–2.97 M). Com- [4] J.K. Pal, S. Anand, J. Joseph, Association of Hsp90 with the heme-
pared with the pyridine hemochromogen method, the regulated eukaryotic initiation factor 2 kinase-a collaboration
for regulating protein synthesis, J. Biosci. 21 (1996) 191–205.
assay here reported showed higher intraday precision, a [5] C.S. Chang, K.P. Chang, Heme requirement and acquisition by
range of quantiWcation of hemin 4 times lower, and more extracellular and intracellular stages of Leishmania mexicana
sensitivity (15–30 times). Comparison between our amazoneusis, Mol. Biochem. Parasitol. 16 (1985) 267–276.
method and that of recording absorbance at 400 nm can [6] J. Bauer, J. Fornnarino, High-performance liquid chromato-
not be performed because the latter produces interfer- graphic analysis of “available” hemin in hematin solutions, J.
Chromatogr. 283 (1984) 378–382.
ences in most of the systems used. Parallel experiments [7] C.A. Genco, B.M. Odusanya, G. Brown, Binding and accumula-
using HPLC analyses [6] have not been carried out tion of hemin in Porphyromonas gingivalis are induced by hemin,
because with the HPLC system reported the hemin Infect. Inmmun. 62 (1994) 2885–2892.
response was linear for a concentration range (0.10– [8] R.J. Porra, O.T.G. Jones, Studies on ferrochelatase. 1. Assay and
15.34 mM) higher than that utilized in our method. properties of ferrochelatase from a pig-liver mitochondrial
extract, Biochem. J. 87 (1963) 181–185.
The method here described was employed for the [9] P.S. Clezy, D.B. Morell, A spectroscopic study of haematin com-
determination of hemin that has been added artiWcially pounds in the Soret region, Biochim. Biophys. Acta 71 (1963) 150–
to a solution and for the direct determination of hemin 156.
present in biological extracts. [10] M.E. Lombardo, L.S. Araujo, A. Batlle, 5-Aminolevulinic acid
Considering that hemin plays crucial roles in very synthesis in epimastigotes of Trypanosoma cruzi, Int. J. Biochem.
Cell Biol. 35 (2003) 1263–1271.
important physiological processes such as protein syn- [11] M.E. Lombardo, L.S. Araujo, A. Branca, A.M. del, C. Batlle, A
thesis, respiration, and proliferation and diVerentiation method for estimating glycine in the presence of excess glutamate
of heme-Xagellate organisms, the hemin quantiWcation with o-phthaldialdehyde, Z. Naturforsch. C 45 (1990) 911–914.