See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/7741550

Heterogeneity and oxidation status of

commercial human albumin preparations in
clinical use

Article in Critical Care Medicine · August 2005

DOI: 10.1097/01.CCM.0000169876.14858.91 · Source: PubMed

CITATIONS READS

92 191

6 authors, including:

Leonard T Rael David Gardner

Ampio Pharmaceuticals, Inc. University of Melbourne
49 PUBLICATIONS 1,174 CITATIONS 347 PUBLICATIONS 14,412 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Oxidative stress as an indicator of acute outcome in patients with trauma View project

De novo transcriptome assembly for the spiny mouse (Acomys cahirinus) View project

All content following this page was uploaded by Leonard T Rael on 17 February 2015.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
Brief Report

Heterogeneity and oxidation status of commercial human albumin

preparations in clinical use*
David Bar-Or, MD; Raphael Bar-Or, BS; Leonard T. Rael, MS; David K. Gardner, PhD;
Denetta S. Slone, MD; Michael L. Craun, MD

Objective: Human serum albumin is indicated for the treatment
of shock, acute restoration of blood volume, and in hypoalbumin-
emia. Conflicting reports are found in the literature for the clinical
safety and efficacy of human serum albumin administration to
critically ill patients. We sought to analyze various commercially
available albumin preparations for common, posttranslational
modifications.
Design: Analysis of six commercially available albumin prep-
arations for clinical use.
Setting: Trauma research laboratory.
Subjects: Commercially available human serum albumin prep-
arations and healthy volunteers.
Interventions: Six commercially available human serum albu-
min preparations were analyzed by high-performance liquid chro-
matography. The presence of various posttranslational modifica-
tions was identified by positive electrospray ionization, time-of-
flight mass spectrometry. Three different lots from three
preparations were also analyzed to assess variability within lots
from the same manufacturer. Also, for the purpose of comparison,
human serum albumin was analyzed in the plasma of healthy
volunteers.
Measurements and Main Results: The six human serum albu-
min preparations analyzed contained a high percentage (57.2 ⴞ
3.3%) of bound Cys34 (oxidation of cysteine in position 34 on the
human serum albumin molecule) in comparison to the plasma
human serum albumin from healthy volunteers (22.9 ⴞ 4.8%).
Lot-to-lot variability in native human serum albumin ranged be-
tween 4.8% and 11.2% in three separate commercial albumins.
Significant differences existed among the various commercial
preparations in other posttranslational modifications of albumin.
Conclusions: Human serum albumin species with a bound Cys34
account for a large percentage of the composition of human serum
albumin preparations used for the treatment of critically ill patients.
Also, the variability within lots from the same manufacturer is sig-
nificant. Consequences of the administration of these oxidized forms
of human serum albumin to critically ill patients warrants further
investigation. (Crit Care Med 2005; 33:1638 –1641)
KEY WORDS: human serum albumin; cysteinylation; nitrosylation;
oxidation

I n a recent patient study, Quinlan
and colleagues (1) demonstrated
the beneficial effect of albumin ad-
ministration in patients with
acute lung injury by improving total an-
tioxidant capacity in plasma. In addition
to treating acute lung injury, possible
clinical indications for administering hu-
man serum albumin (HSA) include the
emergency treatment of shock, the ur-
*See also p. 1667.
From Swedish Medical Center, Trauma Research
Laboratory (DB-O, RB-O, LTR), Center for Reproductive
Medicine (DG), and Trauma Services (DSS, MLC),
Englewood, CO; and DMI BioSciences (DB-O, RB-O,
LTR), Englewood, CO.
Supported by Swedish Medical Center, Englewood,
CO.
Address requests for reprints to: David Bar-Or, MD,
Swedish Medical Center, Trauma Research Lab, 501 E.
Hampden Ave. Rm. 4-454, Englewood, CO 80113.
Copyright © 2005 by the Society of Critical Care
Medicine and Lippincott Williams & Wilkins
DOI: 10.1097/01.CCM.0000169876.14858.91
gent restoration of blood volume, the
acute management of burns, and clinical
situations associated with hypoproteine-
mia (2). Multiple research protocols and
publications assessing the utility of HSA
in other clinical conditions and in chem-
ical assays used HSA from various
sources. Despite numerous publications
reporting conflicting results on the ef-
fects of albumin administration in criti-
cally ill patients (3–7), no investigation
has examined the composition of the ad-
ministered HSA or the clinical efficacy of
the various albumin species.
Therefore, we analyzed the albumin
species in six commercially available HSA
products. Special attention was directed
toward oxidation species of HSA (i.e., cys-
teinylation, nitrosylation, truncation,
etc.) and lot variability within three of the
HSA formulations. Also, HSA from
healthy volunteers was analyzed and
compared with commercial HSA. The ef-
fect of administered, commercial albu-
min (which is mostly oxidized) to criti-
cally ill patients is discussed.
MATERIALS AND METHODS
This study received approval by the Heal-
thOne Alliance Institutional Review Board ac-
cording to guidelines published by the HHS Of-
fice for Protection from Research Risk. HSA
from healthy volunteers and six commercially
available HSA products (ZLB Bioplasma, Berne,
Switzerland; Aventis Behring, Kankakee, IL;
Bayer Corporation, Elkhart, IN; Grifols Biologi-
cals, Barcelona, Spain; Baxter Healthcare, West-
lake Village, CA; and Alpha Therapeutic, Los An-
geles, CA) were analyzed by high-performance
liquid chromatography (Waters, Milford, MA)
coupled to positive electrospray ionization time
of flight mass spectrometry (⫹ESI-TOF MS, Mi-
cromass, UK). Each HSA formulation (25%
weight/volume) and the separated plasma from
each normal volunteer were diluted 1:10 with
dH2O. Ten microliters of each sample was in-
jected in triplicate onto a YMC-Pack Protein-RP,
150 mm ⫻ 4.6 mm, 5␮, high-performance liq-

1638 Crit Care Med 2005 Vol. 33, No. 7

uid chromatography column heated to 50°C acid and acetonitrile/0.1% TFA. The chromato- The relative proportion of each species was
(Waters, Milford, MA) using a 20-min linear gra- graphic conditions are described in Table 1. analyzed separately using a random effects
dient method using water/0.1% trifluoroacetic The output of the high-performance liquid analysis of variance model based on restricted
chromatography was split 1:20 (volume/ maximum likelihood methods. The model
volume) and injected into the mass spectrom- contained a fixed effect for source (healthy
Table 1. Chromatographic conditions eter with a scan range of 500 –3500 m/z, cone subjects or commercial brand), random effects
voltage of 30 eV, source temperature of 100°C, for lots (or subjects), and the combined with-
Time, Flowrate,
mins %A %B mL/min
and gas temperature of 250°C. Albumin elutes in-run and between-vial error term. Pairwise
at 8.15 mins and is visualized as a charge contrasts between the healthy subject mean
0.0 90 10 1 envelope from 1100 to 2500 m/z representing and each commercial preparation mean were
1.0 90 10 1 ⫹44 to ⫹26 charges. The spectrum was then then made. No adjustments were made for
11.0 30 70 1 deconvolved to the uncharged parent mass multiple comparisons. Attempts to add a sep-
13.0 30 70 1
using MaxEnt 1 (Micromass). The parent mass arate variability for subjects and lots did not
13.1 90 10 1
20.0 90 10 1 spectrum was then integrated, and relative substantially improve the ⫺2 ln likelihood sta-
proportions of each species were calculated. tistics, a measure of relative model fit. All tests

Figure 1. A, normal healthy volunteer human serum albumin (HSA) profile. B, typical commercial albumin preparation. HSA identification: DA,
aspartate-alanine absent from N-terminus; L, leucine absent from C-terminus; ⫹Cys, Cysteine 34 is bound by a free cysteine; NO, cysteine 34 is bound by
nitric oxide; ⫹Glyc, glycated albumin.

Table 2. Differences in proportions of posttranslational modifications of all preparations, %

Unbound Bound
⫺DA (N-term) ⫺L (C-term) Native ⫹NO ⫹Cys ⫹Cys ⫹Glyc Cys34 Cys34

Normal Patients, mean, n ⫽ 27 2.9 3.8 53.7 7.7 15.2 0.0 77.1 22.9
SD 0.5 1.5 4.4 2.4 3.4 0.0 4.8 4.8
Alpha, mean, n ⫽ 6 5.9 4.1 28.0 13.1 25.9 3.1 42.9 57.1
SD 0.1 0.2 0.5 0.5 0.4 0.1 0.4 0.4
Aventis, mean, n ⫽ 27 8.2 5.6 26.7 12.0 21.9 2.8 46.3 53.7
SD 0.3 1.3 1.6 0.7 0.9 0.6 2.0 2.0
Baxter, mean, n ⫽ 27 3.6 4.0 27.1 13.6 26.8 4.0 40.0 60.0
SD 0.5 0.9 1.3 0.9 1.4 0.6 1.7 1.7
Bayer, mean, n ⫽ 9 4.0 4.0 26.3 13.8 26.4 3.5 39.7 60.3
SD 0.1 0.7 0.6 0.3 0.4 0.3 0.7 0.7
Grifols, mean, n ⫽ 9 4.8 3.0 29.6 11.4 29.3 4.2 42.1 57.9
SD 0.2 0.5 0.6 0.4 0.3 0.2 0.9 0.9
ZLB, mean, n ⫽ 9 5.9 3.7 29.8 11.7 25.4 3.7 44.6 55.4
SD 0.3 0.5 3.3 0.3 1.4 0.5 2.9 2.9

Crit Care Med 2005 Vol. 33, No. 7 1639

were two-sided and conducted at the 5% level
of significance. The precision of the device and
method was estimated with coefficient of vari-
ance, which was calculated using multiple
measurements. Data were analyzed using JMP
version 4.0 (SAS, Cary, NC).
RESULTS
Figure 1 demonstrates the major post-
translational species of HSA detected in
one of the preparations as well as a sam-
ple from a healthy volunteer. Native, non-
modified HSA has a theoretical molecular
weight of 66,438 Da. All HSA products
tested had a high proportion (⬎25%) of
cysteinylated albumin (a disulfide cys-
teine addition on the only free thiol
group of albumin in position 34), and all
contained an S-nitroso derivative (addi-
tion of nitric oxide to the free thiol group
in position cysteine 34). There were also
several other species: glycosylated albu-
min, albumin that lost the first two
amino acids (aspartate-alanine) from the
N-terminus, albumin that lost leucine
from the C-terminus, and other combi-
nations. The commercial preparation
tended to have a high proportion of
bound cysteine in position 34 (Cys34)
when compared with normal patients
(57.2% vs. 22.9% p ⬍ .02). The Aventis
preparation had a significantly higher
level of albumin missing the first 2 N-
terminal amino acids, aspartate-alanine,
than any of the other preparations (p ⬍
.001), and all of the preparations had a
significantly higher level than the normal
patients (p ⬍ .001). All of the prepara-
tions had a significantly higher level of
nitrosylated albumin than normal pa-
tients (p ⬍ .001). Table 2 shows the dif-
ferences in the proportions of posttrans-
lational modifications of all the
preparations. We assessed the lot-to-lot
variations in three of the preparations
(Aventis, Baxter, ZLB) and found ZLB to
be the most variable, with the native form
having a variance of 11.2% vs. Aventis
and Baxter having variances of 6.1% and
4.8%, respectively (p ⬍ .05). The coeffi-
cient of variation of the measurement
technique was estimated to be 5.6%.
DISCUSSION
HSA, which is the most abundant
plasma protein, is subjected to many
posttranslational modifications in vivo as
a result of oxidative stress and in vitro
during the preparation procedure from
pooled plasma (8). We are now able to
1640
detect these species accurately using
high-resolution mass spectrometry and
to quantify relative amounts of each spe-
cies including cysteinylated albumin.
Cysteinylation of albumin occurs at cys-
teine 34 (Cys34) and has important phys-
iologic implications. Cys34 accounts for
the majority of plasma free sulfhydryl
groups (9) and is the major extracellular
antioxidant in plasma. Cysteinylation or
nitrosylation of albumin could lead to a
significant loss of its oxidant-buffering
capacity. Indeed, elevated levels of cystei-
nylated albumin are seen during placen-
tal ischemia in humans (10) and bowel
ischemia in rats (11). The relatively high
percentage of S-nitrosoalbumin in all
commercial albumin tested has the po-
tential for release of nitric oxide in vivo
and might have additional biological ef-
fects (12–14). Additionally, the loss of as-
partate-alanine from the N-terminus of
albumin completely abolishes the ability
of albumin to chelate free copper that is
released during acidosis (15–17). This
truncation of the N-terminus is associ-
ated with the loss of the free radical scav-
enging ability of albumin.
Using our method, about 23% of albu-
min in healthy volunteers was found to
contain oxidized Cys34, which is in
agreement with another published report
(8). In contrast, commercial albumins
analyzed in this study contained 57% ox-
idized Cys34. The oxidative state of albu-
min administered to patients with vari-
ous ailments reported in the literature
does not detail the source of commercial
albumin used, the posttranslational mod-
ifications present in that preparation, and
the lot numbers used in the study (3–7).
Therefore, we are unable to analyze the
beneficial or detrimental effects observed
in these studies relative to albumin post-
translational modification species admin-
istered. An investigation into the mecha-
nism responsible for the enrichment of
these posttranslational modifications and
truncated albumin in the manufacturing
process is necessary. The extent of oxida-
tion of plasma albumin in critically ill
patients is currently unknown and is un-
der investigation.
Based on these results, the contro-
versy surrounding the administration of
albumin is not only attributable to vary-
ing mortality rates. Factors such as albu-
min manufacturer, lot number variabil-
ity, and posttranslational modifications
were never taken into account in explain-
ing the results of conflicting clinical tri-
als assessing the beneficial effects of al-
A
systematic review
of the composition
of human serum
albumin, with particular at-
tention to the oxidized
forms, used in critically ill
patients and of correlation to
morbidity/mortality rates
with its uses in randomized
trials is urgently needed.
bumin administration. It is possible that
the administration of commercially avail-
able HSA that is largely oxidized might
augment the oxidative stress burden on
an already oxidatively challenged, criti-
cally ill patient. Clearly, large, prospec-
tive clinical trials are needed to validate
this theory. This could help to explain the
conflicting results and increased mortal-
ity rate observed with the clinical use of
HSA. A systematic review of the compo-
sition of HSA, with particular attention to
the oxidized forms, used in critically ill
patients and of correlation to morbidity/
mortality rates with its uses in random-
ized trials is urgently needed. Addition-
ally, all clinical studies using commercial
HSA without attention to its oxidation
state and other posttranslational modifi-
cations should be viewed with caution.
REFERENCES
1. Quinlan GJ, Mumby S, Martin GS, et al:
Albumin influences total plasma antioxidant
capacity favorably in patients with acute lung
injury. Crit Care Med 2004; 32:755–759
2. ABPI Compedium of Data Sheets and Sum-
maries of Products Characteristics 1998 –99.
London, Association of the British Pharma-
ceutical Industries, 1998
3. Wilkes MM, Navickis RJ: Patient survival af-
ter human albumin administration. A meta-
analysis of randomized, controlled trials. Ann
Intern Med 2001; 135:149 –164
4. Cochrane Injuries Group Albumin Review-
ers: Human albumin administration in crit-
ically ill patients: Systematic review of ran-
domized controlled trials. BMJ 1998;317:
235–240
5. Sort P, Navasa M, Arroyo V, et al: Effect of
Crit Care Med 2005 Vol. 33, No. 7
 intravenous albumin on renal impairment
and mortality in patients with cirrhosis and
spontaneous bacterial peritonitis. N Engl
J Med 1999; 341:403– 409
6. von Hoegen I, Waller C: Safety of human
albumin based on spontaneously reported se-
rious adverse events. Crit Care Med 2001;
29:994 –996
7. Finfer S, Bellomo R, Boyce N, et al: A com-
parison of albumin and saline for fluid resus-
citation in the intensive care unit. N Engl
J Med 2004; 350:2247–2256
8. Peters T: All About Albumin: Biochemistry,
Genetics, and Medical Applications. First
Edition. San Diego, Academic Press, 1996
9. Carballal S, Radi R, Kirk MC, et al: Sulfenic
acid formation in human serum albumin by
hydrogen peroxide and peroxynitrite. Bio-
chemistry 2003; 42:9906 –9914
10. Bar-Or D, Heyborne KD, Bar-Or R, et al:
Crit Care Med 2005 Vol. 33, No. 7
View publication stats
Cysteinylation of maternal plasma albumin
and its association with intrauterine growth
restriction. Prenat Diagn 2005; 25:245–249
11. Bar-Or D, Curtis CG, Sullivan A, et al: Plasma
albumin cysteinylation is regulated by cysta-
thionine beta-synthase. Biochem Biophys
Res Commun 2004; 325:1449 –1453
12. Gandley RE, Tyurin VA, Huang W, et al:
S-nitrosoalbumin-mediated relaxation is en-
hanced by ascorbate and copper: Effects in
pregnancy and preeclampsia plasma. Hyper-
tension 2005; 45:21–27
13. Ng ES, Jourd’heuil D, McCord JM, et al:
Enhanced S-nitroso-albumin formation from
inhaled NO during ischemia/reperfusion.
Circ Res 2004; 94:559 –565
14. Bayir H, Kochanek PM, Liu SX, et al: In-
creased S-nitrosothiols and S-nitrosoalbu-
min in cerebrospinal fluid after severe trau-
matic brain injury in infants and children:
Indirect association with intracranial pres-
sure. J Cereb Blood Flow Metab 2003; 23:
51– 61
15. Bar-Or D, Rael LT, Lau EP, et al: An analog of
the human albumin N-terminus (Asp-Ala-
His-Lys) prevents formation of copper-
induced reactive oxygen species. Biochem
Biophys Res Commun 2001; 284:856 – 862
16. Bar-Or D, Curtis G, Rao N, et al: Character-
ization of the Co(2⫹) and Ni(2⫹) binding
amino-acid residues of the N-terminus of hu-
man albumin. An insight into the mecha-
nism of a new assay for myocardial ischemia.
Eur J Biochem 2001; 268:42– 47
17. Chevion M, Jiang Y, Har-El R, et al: Copper
and iron are mobilized following myocardial
ischemia: Possible predictive criteria for tis-
sue injury. Proc Natl Acad Sci U S A 1993;
90:1102–1106
1641