Limits...
Signal loss due to oligomerization in ELISA analysis of amyloid-beta can be recovered by a novel sample pre-treatment method.

Janssen L, Sobott F, De Deyn PP, Van Dam D - MethodsX (2015)

Bottom Line: However, ELISA was developed for monomeric proteins and may be ill-suited for detecting aggregates.Synthetic Aβ40 monomers, Aβ42 oligomers and biological samples from mice and humans were subjected to a chemical pre-treatment protocol with: trifluoroacetic acid (TFA), formic acid (FA) or hexafluoroisopropanol (HFIP) prior to ELISA analysis.In our study we have shown that: •Aβ oligomerization leads to epitope masking and steric hindrance and results in an underestimation of the total Aβ content with ELISA.•Chemically pre-treating samples to disaggregate oligomers can (partially) recover the signal loss.•This novel sample pre-treatment method could provide a more accurate ELISA measurement of the total Aβ concentration in samples with a high oligomer content.

View Article: PubMed Central - PubMed

Affiliation: Laboratory of Neurochemistry and Behavior, Institute Born-Bunge, University of Antwerp, Antwerp, Belgium.

ABSTRACT
According to the predominant theories, soluble amyloid-beta (Aβ) aggregates are the principal neurotoxic agents in Alzheimer's disease pathology, making them a popular target for the development of therapeutics and diagnostic markers. One of the most commonly used methods for determining the concentration of Aβ is ELISA. However, ELISA was developed for monomeric proteins and may be ill-suited for detecting aggregates. Therefore, we investigated the effect of aggregation on the ELISA measurement and developed a novel chemical pre-treatment method, designed to disaggregate Aβ peptides, to improve the ELISA measurement of the total Aβ concentration. Synthetic Aβ40 monomers, Aβ42 oligomers and biological samples from mice and humans were subjected to a chemical pre-treatment protocol with: trifluoroacetic acid (TFA), formic acid (FA) or hexafluoroisopropanol (HFIP) prior to ELISA analysis. In our study we have shown that: •Aβ oligomerization leads to epitope masking and steric hindrance and results in an underestimation of the total Aβ content with ELISA.•Chemically pre-treating samples to disaggregate oligomers can (partially) recover the signal loss.•This novel sample pre-treatment method could provide a more accurate ELISA measurement of the total Aβ concentration in samples with a high oligomer content.

No MeSH data available.


Related in: MedlinePlus

Elisa results of human CSF samples. Comparison of means (±SD) between AD patient (n = 12) and control (n = 13) samples without treatment and after treatment with TFA or HFIP. All treated samples were reconstituted in 1%NH4OH. Asterisks represent significant differences between the untreated group and treatment groups (post-hoc Bonferroni test; ***p < 0.001). Abbreviations: AD, Alzheimer’s disease; HFIP, hexafluoroisopropanol; TFA, trifluoroacetic acid.
© Copyright Policy - CC BY
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4487349&req=5

fig0030: Elisa results of human CSF samples. Comparison of means (±SD) between AD patient (n = 12) and control (n = 13) samples without treatment and after treatment with TFA or HFIP. All treated samples were reconstituted in 1%NH4OH. Asterisks represent significant differences between the untreated group and treatment groups (post-hoc Bonferroni test; ***p < 0.001). Abbreviations: AD, Alzheimer’s disease; HFIP, hexafluoroisopropanol; TFA, trifluoroacetic acid.

Mentions: Finally, we explored the effect of sample pre-treatment on biological samples, i.e., protein extracts from brain tissue of APP23 mice (Fig. 5) and human CSF samples (Fig. 6). Since the FA treatment failed to produce a significant effect in the ELISA measurements of the oligomer standard, it was not included in these further experiments. Due to the more complex matrix of the brain extracts and the fact that we previously diagnosed a possible problem with reconstitution after HFIP treatment, we did include three reconstitution solutions for comparison: ultrapure water, PBS and a 1% NH4OH solution. Since WT mice do not express human Aβ, all treatment groups showed a clear difference in Aβ measurements between WT and heterozygous mice (p < 0.005). The more important question is of course whether the pre-treatments resulted in an increase of the ELISA signal in samples of heterozygous mice similar to the increase observed with the oligomer standard. When comparing the results of the heterozygote samples with and without treatment, we found no significant differences between the untreated samples and the PBS-treated (p = 0.917), the TFA–PBS-treated (p = 0.249) and the HFIP–ultrapure water-treated samples (p = 0.600). The TFA–ultrapure water treatment group, the TFA–1%NH4OH group, the HFIP–PBS group and the HFIP–1%NH4OH group all exhibited a significant increase in signal (p < 0.05). Ideally, the WT samples are unaffected by the treatment. In reality, however, the signal of WT samples was lowered significantly in the TFA–PBS, the HFIP–ultrapure water and the HFIP–PBS group (p < 0.05). On the other hand, in the PBS-treated group a slight increase was observed (p = 0.046). The other treatment groups did not display a significant difference between untreated and treated WT samples. As 1%NH4OH appeared to be the most efficient reconstitution buffer for both TFA and HFIP-treated brain extracts, we selected this reconstitution buffer for the experiments with human CSF. Contrary to the results from the brain extracts, treatment of the CSF samples resulted in a decrease in the ELISA signal for both TFA-treated and HFIP-treated samples when compared to the untreated samples (p < 0.001, Fig. 6). No significant difference could be observed between the control samples and the AD samples (p = 0.217) and the effect of the various treatments was also similar for both groups (p = 0.230). Treatment of the brain extract samples resulted in an increased scatter of the data. Whether this simply reflects the biological variability in Aβ aggregation or is an issue of consistency/reproducibility of the treatment procedure, remains to be investigated. The human CSF samples, however, did not show an increased scatter after treatment. Before applying the protocol, it should be extensively validated for reproducibility and proper controls should be included in any experiment.


Signal loss due to oligomerization in ELISA analysis of amyloid-beta can be recovered by a novel sample pre-treatment method.

Janssen L, Sobott F, De Deyn PP, Van Dam D - MethodsX (2015)

Elisa results of human CSF samples. Comparison of means (±SD) between AD patient (n = 12) and control (n = 13) samples without treatment and after treatment with TFA or HFIP. All treated samples were reconstituted in 1%NH4OH. Asterisks represent significant differences between the untreated group and treatment groups (post-hoc Bonferroni test; ***p < 0.001). Abbreviations: AD, Alzheimer’s disease; HFIP, hexafluoroisopropanol; TFA, trifluoroacetic acid.
© Copyright Policy - CC BY
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4487349&req=5

fig0030: Elisa results of human CSF samples. Comparison of means (±SD) between AD patient (n = 12) and control (n = 13) samples without treatment and after treatment with TFA or HFIP. All treated samples were reconstituted in 1%NH4OH. Asterisks represent significant differences between the untreated group and treatment groups (post-hoc Bonferroni test; ***p < 0.001). Abbreviations: AD, Alzheimer’s disease; HFIP, hexafluoroisopropanol; TFA, trifluoroacetic acid.
Mentions: Finally, we explored the effect of sample pre-treatment on biological samples, i.e., protein extracts from brain tissue of APP23 mice (Fig. 5) and human CSF samples (Fig. 6). Since the FA treatment failed to produce a significant effect in the ELISA measurements of the oligomer standard, it was not included in these further experiments. Due to the more complex matrix of the brain extracts and the fact that we previously diagnosed a possible problem with reconstitution after HFIP treatment, we did include three reconstitution solutions for comparison: ultrapure water, PBS and a 1% NH4OH solution. Since WT mice do not express human Aβ, all treatment groups showed a clear difference in Aβ measurements between WT and heterozygous mice (p < 0.005). The more important question is of course whether the pre-treatments resulted in an increase of the ELISA signal in samples of heterozygous mice similar to the increase observed with the oligomer standard. When comparing the results of the heterozygote samples with and without treatment, we found no significant differences between the untreated samples and the PBS-treated (p = 0.917), the TFA–PBS-treated (p = 0.249) and the HFIP–ultrapure water-treated samples (p = 0.600). The TFA–ultrapure water treatment group, the TFA–1%NH4OH group, the HFIP–PBS group and the HFIP–1%NH4OH group all exhibited a significant increase in signal (p < 0.05). Ideally, the WT samples are unaffected by the treatment. In reality, however, the signal of WT samples was lowered significantly in the TFA–PBS, the HFIP–ultrapure water and the HFIP–PBS group (p < 0.05). On the other hand, in the PBS-treated group a slight increase was observed (p = 0.046). The other treatment groups did not display a significant difference between untreated and treated WT samples. As 1%NH4OH appeared to be the most efficient reconstitution buffer for both TFA and HFIP-treated brain extracts, we selected this reconstitution buffer for the experiments with human CSF. Contrary to the results from the brain extracts, treatment of the CSF samples resulted in a decrease in the ELISA signal for both TFA-treated and HFIP-treated samples when compared to the untreated samples (p < 0.001, Fig. 6). No significant difference could be observed between the control samples and the AD samples (p = 0.217) and the effect of the various treatments was also similar for both groups (p = 0.230). Treatment of the brain extract samples resulted in an increased scatter of the data. Whether this simply reflects the biological variability in Aβ aggregation or is an issue of consistency/reproducibility of the treatment procedure, remains to be investigated. The human CSF samples, however, did not show an increased scatter after treatment. Before applying the protocol, it should be extensively validated for reproducibility and proper controls should be included in any experiment.

Bottom Line: However, ELISA was developed for monomeric proteins and may be ill-suited for detecting aggregates.Synthetic Aβ40 monomers, Aβ42 oligomers and biological samples from mice and humans were subjected to a chemical pre-treatment protocol with: trifluoroacetic acid (TFA), formic acid (FA) or hexafluoroisopropanol (HFIP) prior to ELISA analysis.In our study we have shown that: •Aβ oligomerization leads to epitope masking and steric hindrance and results in an underestimation of the total Aβ content with ELISA.•Chemically pre-treating samples to disaggregate oligomers can (partially) recover the signal loss.•This novel sample pre-treatment method could provide a more accurate ELISA measurement of the total Aβ concentration in samples with a high oligomer content.

View Article: PubMed Central - PubMed

Affiliation: Laboratory of Neurochemistry and Behavior, Institute Born-Bunge, University of Antwerp, Antwerp, Belgium.

ABSTRACT
According to the predominant theories, soluble amyloid-beta (Aβ) aggregates are the principal neurotoxic agents in Alzheimer's disease pathology, making them a popular target for the development of therapeutics and diagnostic markers. One of the most commonly used methods for determining the concentration of Aβ is ELISA. However, ELISA was developed for monomeric proteins and may be ill-suited for detecting aggregates. Therefore, we investigated the effect of aggregation on the ELISA measurement and developed a novel chemical pre-treatment method, designed to disaggregate Aβ peptides, to improve the ELISA measurement of the total Aβ concentration. Synthetic Aβ40 monomers, Aβ42 oligomers and biological samples from mice and humans were subjected to a chemical pre-treatment protocol with: trifluoroacetic acid (TFA), formic acid (FA) or hexafluoroisopropanol (HFIP) prior to ELISA analysis. In our study we have shown that: •Aβ oligomerization leads to epitope masking and steric hindrance and results in an underestimation of the total Aβ content with ELISA.•Chemically pre-treating samples to disaggregate oligomers can (partially) recover the signal loss.•This novel sample pre-treatment method could provide a more accurate ELISA measurement of the total Aβ concentration in samples with a high oligomer content.

No MeSH data available.


Related in: MedlinePlus