To the Editor:
Pollen from mesquite (genus Prosopis) is a potent allergen
responsible for causing immediate hypersensitivity reactions in
susceptible people in the western and southwestern United States.1
A study by Novey et al1 showed that according to skin prick testing
(SPT) results, mesquite sensitization was the most prevalent pollen
sensitization in a group of 100 atopic patients in California.
The clinical relevance of these sensitivities was confirmed with
nasal, conjunctival, and bronchial challenges.1
Over the past several years, the burning of mesquite wood for broiling and barbecuing has become popular because of the distinctive flavor that thesmoke imparts to various foods.2 Respiratory effects reported with exposure to mesquite smoke in mesquite broiler cooks have been attributed to hydrocarbons or other irritants.2 However, it is known that allergens can survive the process of combustion and remain immunologically active in smoke.3
Three patients presented to our clinic with complaints of allergy symptoms after exposure to smoke from mesquite wood. We sought to identify immunologically active allergens in mesquite smoke that could explain the symptoms in these patients.
Of the 3 patients presenting to our clinic, 2 had complained of oral allergy symptoms after eating foods cooked over mesquite wood; the other had complained of rhinitis after being exposed to mesquite smoke used for recreational barbecuing. All patients had positive SPT results with mesquite pollen extract (1:20 w/v; Hollister-Stier, Spokane, Wash).
Serum was obtained from each patient and stored at 0°C. In addition, serum was obtained from 8 mesquite pollen SPTpositive patients without known symptoms to mesquite smoke and stored at 0°C.
Commercial mesquite pollen extract was dialyzed against distilled water overnight at 4°C in 3500-d dialysis tubing. The resulting extract was lyophilized and stored at 4°C.
Commercially available mesquite wood chips used for barbecuing
were obtained and pulverized into small pieces. Protein extraction
was performed in
0.125 mol/L NH4HCO3 at 4°C overnight. The wood extract was centrifuged and the supernatant passed through 0.45-µm poresized filters. The filtrate was dialyzed against distilled water overnight at 4°C in 3500-d dialysis tubing. The resulting extract was lyophilized and stored at 4°C.
Additional mesquite woodchips were burned in an exhaust hood.
Smoke was vacuum-collected in a distilled water trap over approximately
8 hours. The
resultant extract was prepared in a manner similar to that used with the wood extract, as described above, including filtration, dialysis, and lyophilization, and stored at 0°C.
Extract protein concentrations were determined by a modification of the technique of Lowry et al.4 Pollen, wood, and smoke extracts contained 443, 325, and 925 µg of protein per milligram of extract, respectively.
SDS-PAGE of the pollen, wood, and smoke extracts was accomplished
as described previously.5 In short, the mesquite pollen, wood,
and smoke extracts were electrophoresed on BMA-PAGEr (BioWhittaker,
Rockland, Me) 10% to 20 % gradient gels along with low-molecular-weight
controls (Pharmacia, Piscataway, NJ), which provided 6 calibration
points (phosphorylase B, 94 kd; albumin, 67 kd; ovalbumin, 43
kd; carbonic anhydrase, 30
kd; trypsin inhibitor, 20.1 kd; -lactalbumin, 14.4 kd).
Gels were stained with 0.025% Coomassie blue R-250. The mo-lecular weight of each stained protein was calculated by comparison of the protein's gel position with a standard curve constructed from the molecular weight standards.
Immunoblots were accomplished as previously described.5 In
short, mesquite proteins from the 3 extracts were transferred
from unstained SDS-PAGE gels to 0.45-µm nitrocellulose membranes.
After transfer, the molecular weight standards were cut from the
nitrocellulose membrane and stained with Coomassie blue. The nitrocellulose
membranes were washed, blocked, and incubated overnight at room
temperature with one of the following: (1) a 1:5
dilution of a mixture of the "smoke-sensitive" patients' sera in 10 % FBS; (2) a 1:5 dilution of a pooled sera sample from the mesquite pollen SPTpositive patients; or (3) 1:5 pooled cord sera used as a negative control. The membranes were washed and subsequently incubated for 6 hours at room temperature in mouse antihuman IgE monoclonal antibody (Sigma, St Louis, Mo) diluted 1:5000 in 10% FBS. Finally, the membranes were incubated overnight in alkaline phosphatase-conjugated goat antimouse IgG (Chemicon International, Temecula, Calif) diluted 1:2000 in 10% FBS. Membranes were developed, and the molecular weight of each allergen was determined from the molecular weight standards.
Inhibition immunoblots were accomplished as previously described.5
The procedure duplicated that of the immunoblots, as described
preincubation of 10 mL of sera with 10 mg of mesquite pollen protein before the immunoblot procedure.
The SDS-PAGE of the mesquite pollen, wood, and smoke extracts is shown in Fig 1.
Fig. 1.SDS-PAGE of mesquite pollen, smoke,
and wood extracts. Protein bands at 66 and 59
kd are shared by the 3 extracts (arrowheads).
to view : Click on Image to view full size go to PubMed
The smoke extract contains protein bands in the 66-kd and 59-kd
ranges, which are shared by the pollen and wood extracts. The
pollen and wood
extracts have additional proteins not shared by the smoke extract.
IgE binding to the mesquite pollen and smoke extracts was demonstrated
at the 66-kd and 59-kd proteins in both the "smoke- symptomatic"
mesquite pollen SPTpositive groups, as shown in Fig 2.
Fig. 2.IgE immunoblots to mesquite pollen,
smoke, and wood extracts through use of
pooled sera from smoke-sensitive patients (A)
and pollen SPTpositive patients (B). IgE to
proteins at 59 and 66 kd is demonstrated in the
mesquite pollen and smoke extracts in both
Click on Image to view full size
In addition, both groups demonstrated IgE binding to other proteins in the pollen and wood extracts. The negative control showed no IgE binding.
The inhibition blots performed with mesquite pollen extract
partially inhibited IgE to selected bands in the smoke, wood,
and pollen immunoblots (data
not shown). The procedure was repeated to ensure that this was not due to laboratory error; the results were similar.
Nonpollen sources of airborne allergens have been implicated
as possible causes of allergy symptoms during nonpollen seasons
and at times when
airborne pollen cannot be detected.6 Although Goetz et al6 and Steir and Plunkett7 concluded that smoke collected from burning mountain cedar and
grass, respectively, contained no significant allergens, Harper et al3 determined that tobacco allergens survived combustion and remained
immunologically active in smoke.3
We have proved that mesquite smoke does contain significant
amounts of protein and that this protein elicits an IgE response
in susceptible individuals.
We describe 3 patients with mesquite smoke allergy and IgE to the 66-kd and 59-kd proteins present in mesquite smoke. Similar proteins were shared
by the pollen and wood extracts. Surprisingly, the pooled serum from the mesquite pollen SPTpositive group also contained IgE to the shared 66-kd
and 59- kd proteins in the extracts. This finding suggests that the sensitization to mesquite smoke is probably common, though the prevalence of clinical
allergy to mesquite smoke is unknown. Therefore, the potential for allergy to mesquite smoke exists in individuals who have mesquite pollen sensitivity.
IgE to the 66-kd and 59-kd proteins shared among the extracts
only partially inhibited with pollen extract. This does not necessarily
imply that these
proteins are different allergens, however. Bernstein et al8 studied cross-reactivity of tree allergens with RAST inhibition and found that "tree pollen
extracts . . . were poorer immunoabsorbents than ragweed or grass allergens" and "absorbed slightly more than 33% activity from their corresponding
Mesquite smoke might represent a new and important cause of
occupational,2 recreational, and food allergy manifesting as upper
and lower2 respiratory
tract symptoms as well as oral allergy symptoms. The transfer of allergens to foods cooked over mesquite wood might lead to symptoms in individuals
sensitized to shared proteins in mesquite pollen, wood, and smoke. Therefore, individuals with mesquite pollen allergy should consider avoidance of
exposure to mesquite smoke and foods cooked over mesquite wood.
The views expressed in this article are those of the authors
and do not represent the official policy or position of the United
States Air Force, the
Department of Defense, or the United States Government.
Daniel More, MD
Larry Hagan, MD
Bonnie Whisman, MS
Diane Jordan-Wagner, MD
Wilford Hall Medical Center
Department of Allergy and Immunology
1. Novey HS, Roth M, Wells ID. Mesquite pollen-an aeroallergen
in asthma and allergic rhinitis. J Allergy Clin Immunol 1977;59:359-63.
2. Johns RE, Lee JS, Agahian B, Gibbons HL, Reading JC. Respiratory
effects of mesquite broiling. J Occup Med 1986;28:1181-5.
3. Harper SD, Cox R, Summers D, Butler W, Hagan L. Tobacco
hypersensitivity and environmental tobacco smoke exposure in a
population. Ann Allergy Asthma Immunol 2001;86:59-61.
4. Lowry OH, Rosenbrough NJ, Farr L, Randau RL. Protein measurement with folin phenol reagent. J Biol Chem 1951;193:265-75.
5. Jordan-Wagner DL, Whisman BA, Goetz DW. Cross-allergenicity
among celery, cucumber, carrot, and watermelon. Ann Allergy
6. Goetz DW, Goetz MA, Whisman BA. Mountain cedar allergens
found in nonpollen tree parts. Ann Allergy Asthma Immunol 1995;75:256-60.
7. Stier R, Plunkett G. Is there grass allergen in smoke from burning grass field stubble? [abstract]. J Allergy Clin Immunol 1994;93:174.
8. Bernstein IL, Perera M, Gallagher J, Michael JG, Johansson
SGO. In vitro cross-reactivity of major aeroallergenic pollens
radioallergosorbent technique. J Allergy Clin Immunol 1976;57:141-52.
Back to Burning Issues