A protein allergen microarray detects specific IgE to pollen surface, cytoplasmic, and commercial allergen extracts

Abstract

Background: Current diagnostics for allergies, such as skin prick and radioallergosorbent tests, do not allow for inexpensive, high-throughput screening of patients. Additionally, extracts used in these methods are made from washed pollen that lacks pollen surface materials that may contain allergens.

Methodology/principal findings: We sought to develop a high-throughput assay to rapidly measure allergen-specific IgE in sera and to explore the relative allergenicity of different pollen fractions (i.e. surface, cytoplasmic, commercial extracts). To do this, we generated a protein microarray containing surface, cytoplasmic, and commercial extracts from 22 pollen species, commercial extracts from nine non-pollen allergens, and five recombinant allergenic proteins. Pollen surface and cytoplasmic fractions were prepared by extraction into organic solvents and aqueous buffers, respectively. Arrays were incubated with <25 uL of serum from 176 individuals and bound IgE was detected by indirect immunofluorescence, providing a high-throughput measurement of IgE. We demonstrated that the allergen microarray is a reproducible method to measure allergen-specific IgE in small amounts of sera. Using this tool, we demonstrated that specific IgE clusters according to the phylogeny of the allergen source. We also showed that the pollen surface, which has been largely overlooked in the past, contained potent allergens. Although, as a class, cytoplasmic fractions obtained by our pulverization/precipitation method were comparable to commercial extracts, many individual allergens showed significant differences.

Conclusions/significance: These results support the hypothesis that protein microarray technology is a useful tool for both research and in the clinic. It could provide a more efficient and less painful alternative to traditionally used skin prick tests, making it economically feasible to compare allergen sensitivity of different populations, monitor individual responses over time, and facilitate genetic studies on pollen allergy.

Figure 1. Autofluorescence correction of fluorescence intensity values.

A) Scatter plot of green vs. red channel data of uncorrected (grey circles) and autofluorescence-corrected (black circles) fluorescence intensity values. B) Histogram of autofluorescence-corrected fluorescence intensity values on 36 mock arrays probed with secondary antibody only (no sera).

Figure 2. Comparison of surface fractions, cytoplasmic fractions, and commercial allergen extracts.

A) SDS-PAGE gels stained with Coomassie blue. B) Nitrocellulose dot blots probed with sera pooled from 500 individuals and HRP-conjugated anti-IgE.

Figure 3. IgE profiling using protein microarrays.

A) General layout of the allergen microarray, B) scanned image of an array probed only with secondary antibody to show autofluorescence, C–F) and scanned images of arrays probed with sera from four individuals showing different allergen sensitization profiles. Images were pseudocolored with a color spectrum adjusted so that blue indicates low signal and red indicates high signal.

Figure 4. Serum dilution vs. fluorescence intensity values.

Two serum samples, one with moderate levels of allergen-specific IgE (7746) and one with high levels of allergen-specific IgE (7855) were tested at five different dilutions; 50%, 25%, 12.5%, 6.25%, and 3.125%. Specific IgE in microarray units to the three fractions (surface, cytoplasm, and commercial extract) of 12 pollen allergens is shown.

Figure 5. Reproducibility of the allergy microarray.

Six individuals were serially tested 18 times for 80 allergens on the allergen microarray. A) Representative correlation plot of corrected fluorescence intensity values (circles) of one serum sample binding to 80 allergens on two replicate arrays. B) Histogram of slopes of regression curves calculated for 918 pair-wise comparisons of replicates.

Figure 6. Allergen microarray testing vs clinical diagnosis of patients.

Heatmap depiction of microarray results of six patients who were diagnosed as either not having pollen allergy, but having some other aeroallergy (P−/O+) or as having both pollen and other aeroallergy (P+/O+). Diagnoses were made at the in vitro allergy lab by ImmunoCAP RAST on 12 of the same (*) allergen species as on the array, four similar (+) allergen species as on the array, and five other weeds not present of the array. All other allergens on the allergen array were not tested by ImmunoCAP RAST.

Figure 7. Clustering of allergen-specific IgE levels.

Hierarchical clustering of allergen-specific IgE to A) recombinant, B) non-pollen, and C) pollen allergens. Within the pollens, grasses are indicated in green, trees in black, weeds in blue, and cedar (a gymnosperm) in red.

Figure 8. Comparison of specific IgE to surface, cytoplasmic and commercial extracts.

Bar graphs comparing levels of specific IgE in microarray units to surface fractions (black), cytoplasmic fractions (grey), and commercial extracts (white) of different pollen allergens. A) Bars represent the mean of 176 individuals with standard error bars. Cases where specific IgE to commercial extracts was significantly greater than to cytoplasmic extracts are denoted by a black asterisk (*) and cases where specific IgE to cytoplasmic fractions was significantly greater than to commercial extracts is denoted by a plus sign (+). B) Bars represent the mean of 18 replicates (of individual 85806) with standard error bars.

Figure 9. Comparison of specific IgE to recombinant allergens and total protein extracts.

Bar graph comparing reactivity to commercial extract vs. the corresponding recombinant major allergens in microarray units. Bars represent the percent of individuals (mold: n = 153, Tim: n = 143, Bir: n = 106, Rag: n = 124, Derp: n = 171) who show positive reactivity to commercial extract but no reactivity to recombinant major allergen (black), positive reactivity to recombinant major allergen but no reactivity to commercial extract (white), and positive reactivity to both (grey).