Methods 27 (2002) 77–86

www.academicpress.com

Natural rubber latex protein reduction with an emphasis on

enzyme treatmentq
Frank W. Perrellaa,* and Anthony A. Gasparib
a
Tillotson Healthcare Corporation, 360 Route 101, Bedford, NH 03110, USA
b
University of Maryland School of Medicine, Baltimore, MD 21201, USA
Accepted 18 March 2002

Abstract

Natural rubber latex (NRL), derived from the Hevea brasiliensis tree, is a material used to manufacture products in health care,
including medical gloves. Proteins are a naturally occurring component of NRL. These proteins, which can be present on the surface
of NRL gloves, have been related to hypersensitivity reactions in some humans who come into contact with them. These same
proteins also help to maintain the latex colloidal stability during collection and transport prior to manufacture. Consequently, when
measures are taken to remove or degrade these proteins, other problems can be introduced, such as destabilization of the latex and
changes in its coagulation properties. Practical methods are available to reduce the extractable antigenic protein content of NRL
products. We describe here methods of reducing proteins in commercial-grade NRL and finished products. NRL gloves manu-
factured with adequate leaching can produce products with lower levels of extractable antigenic proteins. Emphasis is given here to
enzyme treatment of NRL, as this process is very effective in reducing antigenic proteins in NRL. While this technology adds
marginally to the production cost of standard grades of NRL, it is still quite cost-effective when compared with postwashing NRL
products or the use of synthetic latex. Moreover, enzyme-treated NRL maintains the excellent physical properties and performance
of NRL. Ó 2002 Elsevier Science (USA). All rights reserved.

Keywords: Natural rubber; Hevea brasiliensis; Latex; Protein; Allergen; Enzyme-treated; Latex allergy; Interleukin-4; T cell; Biocompatibility; Glove

1. Introduction ammonia is the increase in stability of the NRL due to

1.1. Natural rubber
Natural rubber latex (NRL) is an emulsion of rubber
polymer in an aqueous dispersion [1–4]. NRL is har-
vested from the rubber tree by cutting grooves in and
excising the bark at an angle from the tree. Latex then
flows down along the cut segment of the tree and into a
small collection cup. The latex is collected from the cups
and processed into commercial rubber latex. Rubber
latex from field collection contains many natural sub-
stances including proteins and 33% polyisoprene rubber
polymer. Ammonia is usually added to the latex as a
preservative to increase the alkalinity (pH) and retard
microbial growth. The additional benefit from adding
q
This work was supported, in part, by Tillotson Healthcare.
*
Corresponding author.
E-mail address: perrella@thcnet.com (F.W. Perrella).
the increase in negative surface charge of the rubber
particles. In addition, the alkaline conditions of am-
moniated latex can denature some proteins and may
even partially hydrolyze others to smaller polypeptides.
The resulting field NRL is further processed to a con-
centrate by creaming or centrifugation. These processes
concentrate the latex to about 60% rubber content.
Significant quantities of nonrubber material such as
proteins are removed during this processing step. This
concentrated form of ammoniated latex is what is typi-
cally used for dipping rubber products such as medical
gloves.
The proteins of NRL constitute about 2% by weight
of the field latex [3–6]. The average rubber tree may
exude 100 mL of latex (33% rubber content) at each
tapping of the tree about every other day. This is ap-
proximately 2 g of protein per tapping, or about 330 g of
protein per year (165 tapping days per year). Most of
these proteins are water soluble and present as aqueous

1046-2023/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved.
PII: S 1 0 4 6 - 2 0 2 3 ( 0 2 ) 0 0 0 5 5 - 5
78 F.W. Perrella, A.A. Gaspari / Methods 27 (2002) 77–86
extractable proteins in latex products such as medical
gloves. In the manufacture of medical gloves, NRL
proteins migrate to the surface of the glove during
drying and heat vulcanization. The traditional water
leaching process removes many of these surface pro-
teins, but measurable amounts can still remain. Al-
though the leaching process is an effective approach to
remove residuals, it does not remove all the proteins,
which can be a source of potential allergens.
NRL proteins are organic substances that contain
carbon, hydrogen, nitrogen, oxygen, and sulfur. They
are naturally occurring polymers that may contain
hundreds of individual amino acid residues linked to-
gether by peptide bonds. The smaller degraded natural
rubber proteins of 10 amino acids or less are typically
called peptides. NRL undergoes several changes during
ammonia addition and storage. Typically some of the
proteins are partially degraded by the alkaline condi-
tions imparted by the ammonia. The change in the
composition of proteins in NRL can have an effect on
the physical properties of the latex [7]. Some of the
proteins lose their secondary structure and are partially
hydrolyzed to small peptides [7–10]. Freshly prepared
centrifuged NRL typically ages over a 4-week matura-
tion period before it is used for dipping purposes.
There may be hundreds of different proteins con-
tained in the latex of the rubber tree. Several of these
proteins have been well characterized and their amino
acid sequences determined [11]. Some of these proteins
are considered potential allergens based on their
capacity to bind to IgE antibodies from the sera of
NRL-allergic people. It is generally believed that the
potentially allergenic protein components are those
associated with the aqueous soluble serum fraction of
NRL. These are the proteins that are dissolved in the
latex serum and tend to concentrate during glove
manufacturing on the rubber film surface. Therefore,
most of the extractable protein on a glove is typically
located on the inner surface in contact with the wearer
(gloves are typically inverted during manufacturing).
We describe here some methods used to reduce the
extractable protein levels of NRL with an emphasis on
enzyme-treated NRL.
2. Procedures affecting protein content in NRL
2.1. Centrifugation of NRL
Most of the protein (75%) in field NRL is freely
dissolved in the serum fraction while a minor fraction
(25%) is bound to the surface of the rubber particles
[2,3]. NRL is typically concentrated to reduce the water
content. When latex is centrifuged, the field latex is
concentrated from about 33% rubber solids to 60%
solids. About one-half of the soluble proteins are re-
Fig. 1. Centrifugation of raw field natural rubber latex concentrates the
rubber and reduces the total protein content by approximately 50%.
moved from the NRL by this process. Since most of
the reduction is from the natural rubber serum frac-
tion, the actual percentage of protein bound to rubber
particles is increased from 25% to about 50% of the
total (Fig. 1).
If the NRL is processed further by another round of
centrifugation, the latex protein content can be reduced
even further [6,12–14]. This is usually achieved by di-
luting the centrifuged NRL concentrate with water
back to about 30% solids, and then recentrifuging it
back to 60% solids. This additional processing can
further reduce the serum proteins by another 50% or
more. Therefore, double centrifuging of latex can re-
duce the overall protein content by 63% or greater.
Although this approach is a simple way of reducing the
protein content of NRL, it may not be the most cost
effective. Each time the latex is concentrated by cen-
trifugation, about 10% of the rubber is lost to skim
latex, and fresh field latex cannot be processed since
the centrifuges are being used. In addition, the double-
centrifuged latex may have to be supplemented with
chemical stabilizers to maintain the NRL stability and
to prevent any unacceptable coagulation from occur-
ring. Unless the NRL processing facilities are
expanded, this approach would produce only about
one-half the current volume of latex in the same pe-
riod, costing more per pound for the finished latex.
The extractable protein content attributable to serum
proteins in finished unleached gloves made from dou-
ble-centrifuged NRL can be reduced by 60% or more
when compared with gloves made from single-centri-
fuged latex. However, the extractable protein attribut-
able to bound proteins is not affected by this process
and therefore the reduction in total antigenic protein
content in finished gloves may be somewhat less than
60%. Another 50% reduction in serum-extractable
protein might be possible by using triple-centrifuged
NRL (recentrifugation of double-centrifuged NRL).
 F.W. Perrella, A.A. Gaspari / Methods 27 (2002) 77–86
2.2. Creaming of NRL
Creamed NRL, which is not as common as centri-
fuged latex, is not produced by the typical centrifugation
process, but by modifying the buoyant density of the
latex and thereby accelerating the natural creaming of
NRL [3,12]. Creaming is a process of concentrating
NRL so that the rubber particles in the latex rise slowly
to the top to form a creamed rubber layer on the liquid.
This approach is not too dissimilar from that of
creaming milk. Creaming is usually carried out in large
vessels with the addition of creaming agents that in-
crease the buoyant density of latex, such as alginate and
methylcellulose type polymers. As the rubber gradually
rises to the top of the liquid and is concentrated, the
aqueous serum phase is removed from the bottom of the
vessel, leaving the creamed latex concentrate in the tank.
This process of creaming NRL is sometimes more ef-
fective but slower than centrifugation. NRL concentra-
tions up to 68% are typical of creamed latex; thus more
of the serum phase is removed than in centrifuged latex.
Following the creaming of rubber, much of the soluble
protein is removed when the aqueous phase is drained
from the bottom of the latex tank. The creaming of
NRL may be more effective in reducing smaller latex
proteins than the centrifugation process. This may be
due, in part, to the addition to NRL of colloidal
creaming agents that may release some of the proteins
that are bound to rubber in the serum phase, thus fur-
ther reducing the overall total protein levels in the
concentrate.
2.3. Prevulcanization of NRL
A process that impacts the extractable protein in
NRL is the production of prevulcanized latex [6,14–16].
The prevulcanizing of NRL is a process in which the
rubber latex is mixed with both stabilizers and vulca-
nizing chemicals and then heated at temperatures up to
70 °C. Prevulcanization is usually done on ammoniated
centrifuged NRL after the addition of compounding
ingredients. Stabilizers and crosslinking agents like
dialkyldithiocarbamate accelerators, zinc oxide, and
sulfur are necessary to precure and crosslink the rubber
latex during heating. An advantage of prevulcanizing
NRL is that when a film is formed and dried, it has good
strength and elastic properties. This type of latex re-
quires little further compounding and postvulcanization
is usually performed at lower than typical vulcanization
temperatures. A disadvantage of prevulcanized NRL is
that the wet gel strength of the dipped films may be
lower than that of nonprevulcanized latex films, thus
requiring some changes in the manufacturing process.
Typically, gloves made with prevulcanized NRL have a
higher extractable protein content than those made from
a postvulcanization process [14–17]. This may be due to
79
the higher water content and greater porosity of films
made from prevulcanized latex. This causes more of the
soluble proteins to migrate to the surface of the rubber
film during the drying process: the higher the film drying
temperature, the greater the extractable protein content
[6].
3. Protein reduction in NRL gloves
3.1. Water leaching of gloves
Leaching is the process of soaking the NRL glove in
an aqueous bath to remove hydrophilic components
[3,17–19]. This process improves film clarity, reduces the
formation of chemicals blooming to the surface during
heating and storage, and reduces the hydration and
electrical conductivity of the rubber glove film. Leaching
typically involves two types: (i) leaching the wet coag-
ulated rubber gel and (ii) leaching the surface of the dry
rubber film [13,15]. The most efficient way of reducing
water-soluble components in NRL gloves is to perform
both wet-gel and dry-film leaching. In either leaching
process, it is essential to maintain the cleanliness of the
leach water. This is usually accomplished by added and
removing some of the leach water continuously (water
turnover of at least 1–2 gal per minute). In addition to
water turnover, effective leaching is improved when the
ratio of leach water volume to total weight of rubber
gloves increases. Poor water turnover and/or low vol-
ume of unclean leach water can increase the amount of
extractable protein in gloves produced.
Although water leaching is an effective means of re-
ducing the extractable protein content of gloves, the
results can vary widely depending on how the latex was
compounded. For NRL gloves, postcure dry film
leaching is more effective than wet gel leaching in re-
ducing the extractable protein content [14]. Wet gel
leaching is important for removing salts, peptides,
soaps, and surfactants from the rubber so that a con-
tinuous rubber film with good barrier properties is
formed, but it is not effective in leaching out larger
proteins that are sequestered within the hydrated coag-
ulated film. The efficiency of leaching may be deter-
mined by several parameters. These include the leaching
time, the leach water temperature, and the rate of leach
water turnover [20]. The combination of implementing
both wet-gel and dry-film leaching with clean water is an
effective means of reducing aqueous extractable proteins
from glove products.
3.2. Post washing and chlorination of gloves
The washing and surface treatment of NRL gloves
with chlorine ions is an effective means of reducing the
extractable protein content on the glove surface
80 F.W. Perrella, A.A. Gaspari / Methods 27 (2002) 77–86
[9,13,21]. Chlorination performed in a low-pH aqueous
leach causes the breaking of bonds in the rubber poly-
mer and a hardening of the film surface. As a result of
this surface modification, the rubber film becomes more
slippery and the coefficient of friction is much reduced.
Although this process is commonly used to improve the
donnability of gloves, it can result in decreased grip on
the outside of the glove for the wearer. Moreover, the
chlorination process can be difficult to control, and re-
sult in small fissures in the rubber film that can com-
promise shelf-life and barrier integrity of the glove [21].
In comparison, extensive postwashing in water alone
can be almost as effective as chlorination [14], but
without the adverse side effects of chlorine on film
physical properties.
4. Application of enzyme-treated NRL
4.1. Proteolytic enzyme treatment of NRL
Proteases are a form of hydrolytic enzyme that cleave
peptide bonds [22,23]. The use of proteolytic enzymes to
digest NRL proteins has broad application in the rubber
industry. Proteases are available commercially in large
supply. Typically these commercial enzymes are pro-
duced from the fermentation of select nonpathogenic
strains of bacteria. Concentrated enzyme powders or
solutions should be handled carefully, however, to avoid
inhalation of aerosols, as they are potential allergens
[24–26]. Proteolytic enzymes are used in many applica-
tions such as in the manufacture of leather, in food
processing including meat tenderizing, and in the man-
ufacture of cheese, beer, and baked goods. The use of
proteases in washing detergents for home use is fairly
common.
Fig. 2. Control-untreated NRL and enzyme-treated NRL were analyzed by SDS–polyacrylamide gel electrophoresis with Coomassie blue staining of
protein. Standard protein molecular weight markers are in the far left and right lanes.
The literature contains several reports on the treat-
ment of NRL with enzymes. In 1992, Novo Nordisk A/S
presented a conference paper on enzyme applications in
NRL device production [27]. In 1993, Sumitomo Rub-
ber Industries presented on the properties of deprotei-
nized NRL prepared by enzyme treatment [28]. In 1994
and 1995, the Rubber Research Institute of Malaysia
sponsored studies on enzyme-treated NRL [13,29]. In
1996, Sumitomo Rubber Industries released the
developments of low-allergen NRL products using en-
zyme-treated latex [30]. In 1998 Tillotson Healthcare
Corporation, in conjunction with Allotex, LLC, pre-
sented data showing greater than a 95% reduction in
antigenic proteins in enzyme-treated NRL using both
ELISAs and radioallergosorbent tests (RASTs) [31].
Several publications have appeared over the years sug-
gesting the use of proteolytic enzymes to digest NRL
proteins for use in the manufacture of latex gloves
[3,9,12–14,18,27–30,32].
4.2. Procedure
Centrifuged ammoniated NRL was treated with an
alkaline protease (0.025%) for 1–2 days at 13–32 °C. The
enzyme-treated NRL (ET-NRL) was then frozen and
thawed, and centrifuged to separate the protein super-
natant (serum) from the rubber. In these studies,
ET-NRL is also defined as the serum phase of enzyme-
treated NRL. The control untreated NRL and the
ET-NRL were then analyzed by sodium dodecyl sul-
fate–polyacrylamide gel electrophoresis (SDS–PAGE)
with Coomassie blue staining to determine the protein
size distribution (Fig. 2). Standard proteins of 14,000–
66,000 Da were used as size references. Enzyme treat-
ment of NRL for at least 1 day significantly reduced the
size distribution of stainable proteins to 14 kDa or less
 F.W. Perrella, A.A. Gaspari / Methods 27 (2002) 77–86
compared with control NRL. The stained protein bands
in the gel suggested a significant reduction in protein
content over a wide molecular mass range (>100–
14 kDa).
In addition, the control centrifuged NRL (non-en-
zyme-treated) and ET-NRL were assayed for NRL al-
lergen by the RAST inhibition assay (Table 1). In the
RAST [33,34], a known amount of NRL protein antigen
is mixed with a known quantity of NRL-specific anti-
body (sera from NRL-allergic subjects) while adding in
a series of dilutions of the test NRL protein allergen
sample. The lower the percentage of NRL-specific an-
tibody that binds to the standard antigen, the higher the
level of protein allergen in the sample. The enzyme
treatment process reduced the average measurable al-
lergenic protein content of NRL by approximately 99%.
5. Evaluation of gloves made from enzyme-treated NRL
5.1. Glove physical property testing
The physical properties of gloves manufactured with
ET-NRL were tested for tensile strength at break, ulti-
mate elongation (stretch), and holes. Control NRL (non-
enzyme-treated) and ET-NRL were compounded in a
dialkyldithiocarbamate accelerator, zinc oxide, and sul-
fur cure formula. Gloves made from both the control
NRL and ET-NRL exceeded the ASTM D3578 re-
quirements of tensile strength at break and ultimate
elongation, as well as the minimum test for accelerated
heat aging. Gloves that undergo accelerated heat aging
must also pass testing for holes using the ASTM D5151
water leak test method. The current hole requirement for
medical examination gloves is an AQL of 2.5 (ASTM
D3578). Gloves made from ET-NRL were tested for
holes over a 2-month period using the water leak test.
The average incidence of holes over this period was 1.4
per 100 gloves (fractional defective ¼ 0:014  0:0057,
n ¼ 52 test days). This value ensures that the majority of
gloves are intact and below the required AQL of 2.5.
Table 1
Residual antigenic protein content of NRL and gloves
Test method RAST
(lg/mL)
Sample NRL
Untreated control (a) 3311:0  680:1
Enzyme treated
Expt 1 (b1 ) 33:0  50:5
(n ¼ 5)
Expt 2 (b2 ) — 5:8  3:6
% Reduction ð100  ðb=aÞÞ >99
81
5.2. Glove biocompatibility testing
The skin biocompatibility of NRL medical gloves [35]
made from ET-NRL was evaluated using both the
rabbit skin irritation test and the guinea pig sensitization
test (ISO Biological Evaluation of Medical Devices
10993, Part 10). Both tests were performed by NAMSA
(Northwood, OH).
The in vivo biocompatibility evaluation of gloves
made from ET-NRL in rabbit skin studies showed no
significant skin irritation reactions. In addition, guinea
pig dermal sensitization testing was used to determine
the potential of rubber materials to elicit an allergic
chemical hypersensitivity response. The guinea pig
testing of gloves made from ET-NRL showed no sig-
nificant chemical sensitization reactions.
Products that contact the skin can also be evaluated
in human clinical studies to demonstrate that they can
be applied safely to the skin without any significant risk
of irritation. The cumulative irritancy patch test is such
a test. A human clinical study using gloves made from
ET-NRL was performed to evaluate the biocompati-
bility of the product with normal human skin. After
repeated patch applications of the product every day for
2 weeks, no significant irritation by gloves made of ET-
NRL was observed. The passing of all three biocom-
patibility tests indicates that gloves made with ET-NRL
retain the quality of those made from untreated NRL.
6. Evaluation of enzyme-treated NRL glove allergenicity
6.1. Results
The residual protein content on the finished gloves
was evaluated by extracting them for 2 h in an aqueous
buffer solution of pH 7.4. The protein content of these
extracts was determined using the ASTM D5712 Lowry
assay for total protein, an immunologic ELISA for
NRL antigenic protein, and the RAST inhibition assay
for human NRL allergens (Table 1).
ELISA Lowry
(lg/g) (lg/g) (lg/g)
Gloves Gloves Gloves
364:6  129:6 211:2  43:9 89:0  30:6
13:4  9:8 0:8  0:8 101:6  13:8
(n ¼ 5) (n ¼ 5) (n ¼ 5)
1:0  0:7 —
(n ¼ 12) (n ¼ 24)
>96 >99 )14
82 F.W. Perrella, A.A. Gaspari / Methods 27 (2002) 77–86

The results of the Lowry test showed that the residual memory T lymphocytes secrete cytokines such as inter-
protein content of gloves made of ET-NRL was below leukin-4 (IL-4) that go on to stimulate B cells. The ac-
the maximum level of 200 lg=dm2 (approximates lg=g) tivated B cells then migrate to lymph nodes where they
recommended by the ASTM D3578 standard for medi- proliferate and differentiate into plasma cells that pro-
cal examination gloves. The Lowry assay can overesti- duce IgE antibodies. These antibodies then enter the
mate the amount of total protein in NRL gloves because circulation where they sensitize and bind to Fc epsilon
of the contributions of chemicals in the latex that can receptors on mast cells. The mast cells then present
interfere with the test. Moreover, the Lowry assay can- throughout the body, become primed to respond to fu-
not distinguish between intact and digested protein ture interaction with the same antigen, and release
fragments, making it difficult to evaluate the significance inflammatory mediators such as histamine. NRL-sensi-
of ET-NRL with respect to antigenicity. tized mice are expected to have T cell lymphocytes that
The ELISA is more specific and sensitive to NRL can be activated to proliferate and produce IL-4 cyto-
proteins in rubber products. This test can detect latex kine in response to an in vitro challenge with NRL
antigens in the microgram per gram of rubber range protein antigens. Moreover, these sensitized T cells can
using rabbit antisera to NRL proteins [36]. Moreover,
the ELISA can be used to determine the effectiveness of
the enzyme treatment process, since degradation of
proteins into relatively small pieces makes them less
likely to bind to the highly specific antibodies used in
this assay. The results of ELISA testing suggest that the
extractable antigenic protein content of gloves made of
ET-NRL was significantly below the values determined
for total protein by the Lowry test. The average ELISA
value of extractable antigens from gloves made of ET-
NRL was about 1 lg=g (Table 1).
In the RAST assay, human antiserum from NRL-
allergic subjects is used to detect extractable allergenic
proteins in rubber products. The RAST inhibition test
was used to evaluate whether the reduction in response
to antigenic protein (ELISA test) correlated with a re-
duction in human NRL allergens. The average RAST
measurement of human latex allergens in gloves made
from ET-NRL was 13  10 lg=g (Table 1). Although
this value of antigenic protein is slightly higher than that
of the ELISA, the test values are not expected to be
identical since different types of antiserum (human vs
rabbit) were used in the assays. When a larger sample of
gloves made from ET-NRL was evaluated over a longer
period of manufacture (Table 1), there was a closer re-
lationship between the ELISA and RAST values (1 lg=g
vs 6 lg=g, respectively).
Proteins that are enzymatically cleaved into smaller
pieces tend to lose their conformational epitopes and
may potentially lose linear epitopes [37]. The resulting
smaller peptide pieces then may have less chance to be Fig. 3. Immunobiochemical response of NRL-sensitized mouse lym-
recognized by the components of the immune mecha- phocytes to NRL protein antigens challenged in vitro. BALB/c mice
nism, thereby making them potentially less antigenic were immunized by subcutaneous injections of the serum fraction of
NRL at the base of the tail on Days 0, +14, +28, and +42. (A)
than the intact protein antigen. The immunobiology of
Lymphocyte cell proliferation: Sensitized lymph node cell suspensions
NRL allergy and the biochemical mechanisms that may from NRL-immunized mice were challenged in vitro for their ability to
be involved were studied in an animal model [39,40] to proliferate in response to NRL or the serum fraction of enzyme-treated
determine the immune response [38], if any, to ET-NRL. NRL (ET-NRL) using radioactive thymidine ([3 H]Tdr) incorporation
When antigens enter the body, they are captured by as an indicator of DNA synthesis. (B) Lymphocyte IL-4 synthesis:
Sensitized lymph node cell suspensions from NRL-immunized mice
antigen-presenting cells (APC) and then presented in an
were cultured in vitro. Cell culture supernatants from in vitro NRL or
immunogenic form to T cells (TH2). The T cells then ET-NRL antigen-challenged lymph node T lymphocytes were har-
recognize the epitopes (information) on the peptide an- vested on Day 3 of culture, and subjected to sandwich ELISA for the
tigen, and become memory cells. On restimulation these presence of IL-4 using standard techniques.
 F.W. Perrella, A.A. Gaspari / Methods 27 (2002) 77–86
promote elevated serum IgE levels which are usually
related to the allergic condition.
The animal study of NRL-sensitized mice indicated
that NRL antigens digested by enzyme treatment have a
reduced potential to activate type 2 helper T cells to
proliferate during an in vitro challenge, compared with
those stimulated in vitro by control non-enzyme-treated
NRL proteins (Fig. 3A). Furthermore, a NRL challenge
in vitro stimulated significant synthesis of the cytokine
IL-4 in the NRL sensitized T cells compared with those
challenged with ET-NRL alone (Fig. 3B). Such a con-
comitant reduction in the secretion of cytokine IL-4 by
ET-NRL could reduce the potential of B-cell isotype
switching to the IgE allergic pathway. This concept was
further supported by the lack of increase in IgE serum
levels of mice immunized with ET-NRL compared with
IgE levels of mice immunized with control NRL pro-
teins (Fig. 4).
The human skin prick test (SPT) was performed with
gloves made from ET-NRL and with control gloves
made from non-enzyme-treated NRL (Fig. 5). Neither
glove was postwashed so that the relative responses
could be compared directly. Twenty highly sensitive
people with a history of NRL protein allergy were skin
prick tested with extracts from both control and en-
zyme-treated NRL gloves. Ninety percent of the subjects
showed at least a 10-fold reduction in skin reaction to
extracts of gloves made from ET-NRL. Approximately
60% of the highly latex-allergic subjects showed a 100-
fold or greater reduction in skin response. A 1000-fold
reduction in allergic skin reaction was observed in about
15% of the clinical population. The human skin prick
testing of highly latex-allergic subjects demonstrates that
Antigen-Specific IgE
Immunogen Dose
Preimmune
50 mcg
100 mcg
100 mcg +
Alum
0.00 0.05 0.10 0.15 0.20
Absorbance Units
Fig. 4. Serum IgE levels of mice immunized with NRL or ET-NRL
protein antigens. BALB/c mice were immunized by subcutaneous in-
jections of the serum fractions of NRL or ET-NRL at the base of the
tail on Days 0, +14, +28, and +42. Sera from NRL- and ET-NRL-
immunized mice were tested for antigen-specific IgE using an ELISA
and standard techniques. The gray smooth bar represents the ET-NRL
immunogen IgE response and the darker crosshatched bar represents
NRL immunogen IgE response.
83
a much reduced response to product made from ET-
NRL is possible compared with the nontreated control
NRL product. This study was not intended to suggest
that people who are highly allergic to NRL should use
gloves made of ET-NRL, but that this material may be
useful in reducing the potential for sensitization to the
normal population. People that are known to be type I
NRL-allergic should not use NRL products of any kind.
6.2. Methodology
6.2.1. NRL reagents
The centrifuged, ammoniated NRL and enzyme-
treated NRL from Southeast Asia was provided by
Tillotson Healthcare. Ammoniated NRL was treated
with an alkaline protease (0.025%) for 1–2 days at 13–
32 °C with constant agitation to produce enzyme-treated
NRL. The serum phase of untreated NRL or ET-NRL
was prepared by freezing and thawing the NRL to
produce irreversible coagulation [7], followed by cen-
trifugation of the thawed NRL for 2 h at 15,000g. The
surface rubber material was removed from the tube, and
the supernatant was recentrifuged for 1 h at 27,000g.
The resulting supernatant of NRL (stock solution pro-
tein concentration 2.93 mg/mL) or ET-NRL (stock so-
lution protein concentration 4.48 mg/mL) was then
filter-sterilized through a 0.22-lm filter. The pH of the
NRL or ET-NRL was highly basic (pH about 10); the
pH of these solutions was adjusted to a more physio-
logic level (pH 7.4) using CO2 bubbling after dilution of
the NRL or ET-NRL in phosphate-buffered saline.
NRL or ET-NRL maintained a stable pH 7.4 after such
treatments. These sterile, pH-adjusted solutions were
used for in vivo immunizations, in vitro ELISA, IL-4
lymphocyte production assay, and T-lymphocyte pro-
liferation assay.
6.2.2. Immunization of animals
Female BALB/c mice (6–8 weeks old) were purchased
from Jackson Laboratories (Bar Harbor, ME). Animals
(N ¼ 6/group) were immunized by subcutaneous injec-
tions of NRL or ET-NRL at the base of the tail in a 50-
lL volume, with 50- to 100-lg doses of antigen or saline
alone. For the antibody responses, the mice were im-
munized on Days 0, +14, +28, and +42. The animals
were bled via the tail vein on Days )1, +21, and +56.
When alum adjuvant (Alhydrogel, Accurate Chemical,
Westbury, NY) was administered during the immuni-
zations, it was in a 1:1 ratio with the NRL solutions at
the indicated volume. For the in vitro T-lymphocyte
proliferation and IL-4 production assays, mice were
immunized with intradermal injections on Days 0 and
+7 with 50 lg of NRL or saline solution. Seven Days
later, popliteal lymph node cells were harvested and
disaggregated for the in vitro T-lymphocyte prolifera-
tion assay on Day +14.
84 F.W. Perrella, A.A. Gaspari / Methods 27 (2002) 77–86

Fig. 5. Skin prick tests (SPTs) were performed by endpoint skin prick test titration on the volar forearm of highly type I NRL-allergic human
subjects. The skin pricks were at least 2 cm apart and conducted through a test drop on the epidermis with a skin test lancet. Skin prick test sites on
the skin were wiped clean after 15 min and the wheal and flare reactions were carefully measured. A positive wheal was considered larger than 4 mm
of the negative control. The data represent the decrease in skin reactivity to aqueous extracts of gloves made from ET-NRL relative to control NRL
gloves.

6.2.3. T-lymphocyte proliferation assay the supplier’s recommendations (Pharmingen OptEIA

Unfractionated lymph node cell suspensions from
NRL-immunized mice (50 lg NRL on Days 0 and 7)
were prepared on Day +14 and assayed in vitro for their
ability to proliferate in response to the addition of me-
dium alone, NRL or ET-NRL. Cell suspensions were
prepared by disaggregation of dissected lymph nodes
between the frosted ends of sterile glass slides. A total of
1  105 lymph node cells were plated with the indicated
stimuli in 96-well round-bottom microtiter plates
(Corning Glassworks, Corning, NY) using standard T-
lymphocyte medium. The 96-well plates were incubated
in a 37 °C, 5% CO2 atmosphere for 78 h. One microcurie
of 3 [H]TdR (Amersham Life Science, Arlington Heights,
IL) was then added to each well for another 18 h of the
incubation to measure proliferation. DNA from the la-
beled cells was harvested with a PhD Model 200A cell
harvester (Cambridge Technology, Watertown, MA),
and the level of 3 [H]TdR was measured with a Beckman
LS 6500 scintillation counter (Beckman Instruments,
Fullerton, CA).
6.2.4. IL-4 lymphokine ELISA
Unfractionated lymph node cell suspensions from
NRL-immunized mice (50 lg NRL on Days 0 and 7)
were prepared on Day +14 and cultured. The in vitro
cultures were challenged with NRL or ET-NRL and
the stimulated lymph node T lymphocytes were har-
vested after 72 h of culture, and then subjected to
sandwich ELISA assay for the presence of IL-4 using
standard techniques. Briefly, 25 lL of supernatant from
each well was applied to wells of a 96-well ELISA plate
(Corning Costar, Cambridge, MA) that had been
treated with 1 lg=mL anti-mouse IL-4 capture antibody
(BVD4-1D11, BD-Pharmingen, San Diego, CA), as per
Set). Bound IL-4 was measured using 1 lg=mL bioti-
nylated anti-mouse IL-4 antibody (BVD6-24G2,
Pharmingen) followed by avidin peroxidase (Sigma
Chemical, St. Louis, MO). Plates were developed for
approximately 30 min in the presence of 2,20 -azin-
obis(3-ethybenzthiazoline-6-sulfonic acid) and H2 O2
(Sigma Chemical). Plates were read at 405 nm using a
Bio-Rad (Hercules, CA) benchmark microplate reader.
The OD readings were compared with those of re-
combinant mouse IL-4 (Pharmingen) concentration
standards, which ranged from 15 pg/mL to 4 ng/mL.
The limit of detection of this assay is 7.8 pg of IL-4 per
milliliter of culture supernatant (Becton–Dickinson/
Pharmingen Catalogue, 2000).
6.2.5. Latex-specific ELISA for mouse IgE antibodies
To test sera from NRL- or ET-NRL-immunized mice
for antigen-specific IgE, a specific ELISA was developed
using standard techniques. Briefly, ELISA plates
(Corning Glass Works) were coated with 10 lg of NRL
or ET-NRL and the plates were blocked with 0.25%
BSA in Tween/PBS. A 1:25 dilution of the sera was then
plated in triplicate to allow for antibody binding. Un-
bound antibody was washed off and biotinylated anti-
mouse IgE (BD-Pharmingen) was added at 2 lg=mL.
The level of NRL- or ET-NRL-specific immunoglobulin
was then analyzed by adding streptavidin–horseradish
peroxidase at a dilution of 1:2000. o-Phenylenediamine
dihydrochloride (Sigma) was added according to Sigma
instructions for color development. Plates were read on
a Bio-Rad benchmark microplate reader, and optical
density was read at 570 nm. For the ELISA inhibition
assay, serum from NRL-immunized mice was preincu-
bated with dilution buffer or buffer containing
 F.W. Perrella, A.A. Gaspari / Methods 27 (2002) 77–86
2–200 lg=mL of soluble NRL or ET-NRL for 60 min at
37 °C prior to use in an ELISA using NRL-coated
plates. Binding of IgE antibodies to the coated plates
was detected as described above.
6.2.6. SDS–PAGE
Using standard methods, 10–30 lg of NRL or ET-
NRL polypeptides were electrophoretically separated
using a precast 4–20% gradient Tris–HCl gel (Bio-Rad)
and stained with Coomassie brilliant blue.
6.2.7. NRL total protein measurement
The total protein content of NRL or ET-NRL for the
in vivo immunizations and the in vitro ELISA and T-
lymphocyte studies was quantified using a modified
Lowry assay (ASTM Standard Method D5712). The
same test method was used to determine the total pro-
tein level on the finished gloves.
6.2.8. Antigenic protein measurement
Ammoniated NRL polypeptides (3.3 mg total pro-
tein) and rabbit ammoniated NRL-specific antiserum
(high-titer antiserum, estimated ELISA inhibition titer
1:15,000) for ELISA testing were generously provided
by Donald Beezhold, (Guthrie Research Foundation,
Sayre, PA). The reagents were used to validate the
NRL-specific ELISA in the LEAP/ELISA test [36].
6.2.9. Allergenic protein measurement
RAST inhibition assay of NRL protein antigens was
performed by IBT Reference Laboratory (Lenexa, KS)
according to its standard protocol.
6.2.10. Human skin irritation test
Human cumulative irritation testing to evaluate
whether a product can be applied safely to the human
skin without any significant risk of irritation was done
by TKL Research, (Paramus, NJ). The study was per-
formed on 27 subjects using 2  2-cm Webril patches
from gloves made from ET-NRL and applied under
occlusive conditions. Sodium lauryl sulfate (0.2%)
served as a positive control. A total of 11 patches of
glove films were applied to the infrascapular area of the
back over a 2-week period. Twenty four hours after
application, patches were removed and the skin site
evaluated and scored for skin reaction. The patching
was continued every 24 h for 2 weeks to evaluate the
biocompatibility of the gloves with normal human
skin.
6.2.11. Human skin prick testing
The skin testing was done by Gordon L. Sussman,
(Clinical Immunology and Allergy Associates, Toronto,
Ontario, Canada) on the volar forearm of human sub-
jects with type I NRL allergy [41,42]. A dilution of 1/
100,000 of NRL glove extracts was the initial concen-
85
tration with continued testing using subsequent 10-fold
lower dilutions (higher concentrations of extracts) until
a positive skin test reaction was observed. Skin prick test
sites on the skin were wiped clean after 15 min and the
wheal-and-flare reactions were measured. A positive
wheal was considered larger than 4 mm of the negative
control wheal. The reference material was a standard
Bencard ammoniated NRL reagent, a histamine (1 mg/
mL) positive control, and a normal saline negative
control. The diluent was human serum albumin 0.3 mg/
mL) in 0.9% saline.
7. Concluding remarks
There are many ways of reducing the amount of
protein antigens in NRL. These include double cen-
trifugation or creaming of NRL, both wet-gel and
dry-film leaching, postwashing and chlorination, and
enzyme treatment of NRL. Of the above approaches,
only enzyme treatment of NRL effectively reduces the
antigenic protein content throughout the latex and its
products. The development of NRLs with a low-an-
tigenic-protein content is feasible, and can be used
successfully in the manufacturing of latex gloves and
other rubber products. Our results with the mouse
animal model of NRL sensitization and other bio-
logical studies are consistent with the hypothesis that
the treatment of NRL with proteolytic enzymes sig-
nificantly reduces the antigenicity of this material. The
enzyme treatment of NRL alters the antigenic proteins
associated with NRL by cutting them into smaller
pieces and rendering them less immunogenic. ET-NRL
has been shown to be suitable for large scale pro-
duction of NRL medical gloves with acceptable
physical, aging, and barrier properties. The broad
application of this process for consumer rubber
products would help to maintain NRL as a leading
raw material for these applications. Scientific research
on ET-NRL will continue to advance and stimulate
the development of other applications for low-anti-
genic-protein NRLs [43].
Acknowledgments
The authors thank Judy Genzale for assistance in
preparing the NRL and ET-NRL proteins and for
performing the Lowry assay and ELISA for this
study. We also thank Paul Bagley and Joseph Pieroni
for technical assistance in the ET-NRL studies, and
Gordon Sussman, M.D., for assistance in conducting
the human skin prick test clinical studies. Thanks also
go to Robert Burns, Barbara Ferbel, and Carol Tanck
for their technical assistance in the immunological
studies.
86 F.W. Perrella, A.A. Gaspari / Methods 27 (2002) 77–86
References
[1] J. Perez, Clin. Rev. Allergy 11 (1993) 355–361.
[2] E. Yip, N.K. Poon, M.K., Lang, K. Turjanmaa, in: Latex Protein
Allergy: The Latest Position (European Rubber Journal), Paris,
1995.
[3] A. Subramanian, in: J.N. Fink (Ed.), Immunology and Allergy
Clinics of North America, Saunders, Philadelphia, PA, 1995, pp.
1–20.
[4] H. De Livonniere, Clin. Rev. Allergy 11 (1993) 309–323.
[5] I. Bodycoat, in: Latex Protein Allergy: The Present Position
(European Rubber Journal), Amsterdam, 1993.
[6] A. Subramanian, E. Yip, K.P. Ng, K.L. Mok, in: Latex Protein
Workshop of the International Rubber Technology Conference
(Rubber Research Institute of Malaysia), Kuala Lumpur, 1993.
[7] M. Morris, B. Lakin, in: Latex: Educational Symposium No. 36,
Rubber Division, American Chemical Society Educational Sym-
posium on Natural Rubber, Philadelphia, PA, 1995.
[8] W. Truscott, in: J.N. Fink (Ed.), Immunology and Allergy Clinics
of North America, Saunders, Philadelphia, PA, 1995, pp. 89–121.
[9] W. Truscott, in: International Latex Conference: Sensitivity to
Latex in Medical Devices (US Food and Drug Administration,
CDRH), Baltimore, MD, 1992.
[10] R. Kekwick, in: Latex Protein Allergy: The Present Position
(European Rubber Journal), Amsterdam, 1993.
[11] T. Palosuo, in: Latex Protein Allergy: Managing the Issue
(European Rubber Journal), Amsterdam, 1996.
[12] G. Pailhories, Clin. Rev. Allergy 11 (1993) 391–402.
[13] P.F. Lai, K.P. Ng, in: Latex: Educational Symposium No. 36,
Rubber Division, American Chemical Society Educational Sym-
posium on Natural Rubber, Philadelphia, 1995.
[14] L.M. Barclay, in: Latex Protein Allergy: The Latest Position
(European Rubber Journal), Paris, 1995.
[15] T. Pendle, The Malaysian Rubber Producers’ Research Associa-
tion, Brickendonbury, UK, 1995.
[16] R. Russell-Fell, in: Latex: Educational Symposium No. 36,
Rubber Division, American Chemical Society Educational Sym-
posium on Natural Rubber, Philadelphia, 1995.
[17] M.Y. Amir Hashim, in: Latex Protein Workshop of the Interna-
tional Rubber Technology Conference (Rubber Research Institute
of Malaysia), Kuala Lumpur, 1993.
[18] A. Subramanian, in: International Latex Conference: Sensitivity
to Latex in Medical Devices (US Food and Drug Administration,
CDRH), Baltimore, MD, 1992.
[19] Z.A. Karim, in: International Latex Conference: Sensitivity to
Latex in Medical Devices (US Food and Drug Administration,
CDRH), Baltimore, MD, 1992.
[20] S.R. Kamath, M.S. Abraham, in: International Latex Conference:
Sensitivity to Latex in Medical Devices (US Food and Drug
Administration, CDRH), Baltimore, MD, 1992.
[21] N.A.A. Aziz, in: Latex Protein Workshop of the International
Rubber Technology Conference (Rubber Research Institute of
Malaysia), Kuala Lumpur, 1993.
[22] R.J. Beynon, J.S. Bond, Proteolytic Enzymes: a Practical Ap-
proach, IRL Press, Oxford, 1989.
[23] L. Stryer, in: Biochemistry, Freeman, New York, 1988, pp. 201–
232.
[24] H.R. Colten, P.L. Polakoff, S.F. Weinstein, D.J. Strieder, N. Engl.
J. Med. 292 (1975) 1050–1053.
[25] X. Baur, G. Fruhmann, Clin. Allergy 9 (1979) 75–81.
[26] X. Baur, G. Fruhmann, Clin. Allergy 9 (1979) 443–450.
[27] E. Niels, in: International Latex Conference: Sensitivity to Latex
in Medical Devices (US Food and Drug Administration, CDRH),
Baltimore, MD, 1992.
[28] N. Ichikawa, Y. Tanaka, E.A. Hwee, in: International Rubber
Conference (Rubber Research Institute of Malaysia), Kuala
Lumpur, 1993.
[29] H.M. Ghazaly, in: Latex Protein Allergy Workshop of the
International Rubber Technology Conference (Rubber Research
Institute of Malaysia), Kuala Lumpur, 1994.
[30] S. Nakade, in: Latex Protein Allergy: The Present Position
(European Rubber Journal), Paris, 1996.
[31] F.W. Perrella, in: International Latex Conference (Rubber and
Plastic News), Akron, OH, 1998.
[32] T.B. Soerensen, A. Prentoe, B. Knudsen, P.S. Skov, in: Interna-
tional Latex Conference: Sensitivity to Latex in Medical Devices
(US Food and Drug Administration, CDRH), Baltimore, MD,
1992.
[33] H. Schroder, L. Yman, Allergy 35 (1980) 234–238.
[34] J.E. Autegarden, Clin. Rev. Allergy 11 (1993) 403–412.
[35] M. Van Geffen, in: Latex Protein Allergy: Managing the Issue
(European Rubber Journal), Amsterdam, 1996.
[36] D.H. Beezhold, Guthrie J. 61 (1992) 77–81.
[37] M.J. Crumpton, in: M. Sela (Ed.), The Antigens, Academic Press,
New York, 1974, pp. 1–77.
[38] G.J. Tortora, B.R. Funke, C.L. Case, Microbiology: An Intro-
duction, Benjamin/Cummings, New York, 1995.
[39] J.E. Slater, E. Paupore, Y.T. Zhang, A.M. Colberg-Poley,
J. Allergy Clin. Immunol. 102 (1998) 469–475.
[40] M.R. Woolhiser, A.E. Munson, B.J. Meade, Toxicol. Sci. 55
(2000) 343–351.
[41] G.L. Sussman, S.M. Tarlo, J. Dolovich, JAMA 265 (1991) 2844–
2847.
[42] C.P. Hamann, S.A. Kick, in: G.A. Mellstrom, J.E. Wahlberg, H.I.
Maibach (Eds.), Protective Gloves for Occupational Use, CRC
Press, Boca Raton, FL, 1994, pp. 131–156.
[43] F.W. Perrella, J.K. Pieroni, T.N. Tillotson, Enzyme, Stabilizer
and Antioxidant Treated Natural Rubber Latex Product and
Method of Processing Same, United States Patent, US 6, 380, 283
B1, Apr. 30, 2002.