Sensitive and Accurate Multi-Color RPPA Using Grace Bio-Labs ArrayCAM Protein Microarray System

Introduction

Profiling sets of multiple protein markers and signaling pathways is becoming the standard method for biomarker discovery, accurate disease diagnosis, and monitoring of therapeutic response. Low-cost, high throughput multiplexed assay platforms that can provide specific, sensitive, and reproducible measurements, featuring reliable and robust assay controls along with easy to use methodologies will yield increased discovery throughput and decreased disease mortality.The application of Reverse Phase Protein Array (RPPA) has developed into a valuable multiplexed tool for the precise study of cell signaling. Detection of low abundance proteins from small or labile samples is a primary obstacle for RPPA users while enzyme catalyzed colorimetric detection is the traditional method for obtaining measurements of sufficient sensitivity for quantitation of these proteins of interest. However, accuracy of robust multi-color inter- and intra- experimental data normalization for consistent measurements is prohibitively challenging with single channel colorimetric detection. Further, dynamic range of colorimetric detection is significantly curtailed. Grace Bio-Labs has developed a complete proteomic microarray system incorporating high protein-binding porous nitrocellulose (PNC), multi-color multiplexing using near IR- nanocrystal detection, optimized reagents and protocols, along with rapid image acquisition and automated software analysis to address the needs of RPPA users for both sensitivity as well as accuracy (Figure 1).

Method

Platform Description – The Grace Bio-Labs ArrayCAM™ system employs a suite of components specifically optimized to function together for easy, rapid, sensitive and accurate RPPA.  ONCYTE PNC film is employed to immobilize cell lysates, providing 500X greater binding capacity than other microarray substrates (Figure 2).  This translates into maximum signal to noise (S/N) and 50X lower LOD.  Porous nitrocellulose is optimal in combination with near-IR detection modalities due to emission and excitation light resonant scattering.  Optimized pore size and structure amplify emitted fluorescent signal achieving a 150-fold signal enhancement (Figure 3.)  When considering fluorescence detection, the inherent native PNC auto-fluorescence at visible wavelengths must be considered.  Migrating detection to far-red and near infra-red wavelength mitigates PNC auto-fluorescence and provides 4x greater S/N than visible wavelengths (Figure 4).  Optimal fluorescence in the near IR range is provided by quantum nanocrystals QNC which are bright, photo-stable, protein sized fluorescent inorganic particles.  The QNC is excited at a single wavelength, while emitting in narrow emission bands, facilitating robust multi-color multiplexing at red, far red and near-IR wavelengths (Figure 5).  Furthermore, these particles do not quench or photo-bleach under repeated scanning or degrading environmental conditions such as at high ozone levels, providing an archivable record of experimental data.  Finally, to compliment the unique single excitation/multiple emission wavelength properties of (QNC) the low cost multi-wavelength reading ArrayCAM instrument provides array image data in under 1 minute.  Following acquisition the automated software rapidly provides spot finding and data analysis results, reliably eliminating lengthly manual and subjective data processing.  This novel system exploits the properties of PNC and QNC, using the three channels simultaneously to increase throughput and integrate total protein normalizers, improving data accuracy (Figure 6). 

Three Color RPPA – To illustrate the benefits of high throughput RPPA three-color multiplexing arrays, two microarrays containing Raf 1 kinase inhibitor time-course/dose-response treated C32 cell lysate arrays were probed with antibodies for three markers and detected with QNC.  One array was probed with anti-Erk, total and phosphorylated (T202/Y204), and the second with anti-S6, total and phosphorylated (S235/236) then both detected with Q800 and Q655, respectively.  Both arrays were also simultaneously probed for GAPDH for total protein normalization and detected with Q585.  Excellent time-course dosage patterns are observed (Figure 7) with both changing (phosphoprotein) and stable (non-phosphorylated) markers clearly exhibited.  As expected, signal for P-Erk drops rapidly in response to the Raf 1 kinase inhibitor and remains low over time.  The temporal effect on the downstream P-S6 is clearly observed at the later timepoints. Cellular levels of the non-phosphoryated proteins remain constant across dose and time, indicating that the observed changes in phosphorylation states are not due to general decreases in Erk or S6. 

Additionally, to demonstrate the benefit of using intra-spot multi-color signal normalization , data from cell lysate replicate spots (n=8, each cell lysate) for P-pERK were normalized four different ways; 1) Un-normalized, 2) normalized to GAPDH measured on another slide from the same print run, 3) normalized to total protein signals using FastGreen protein stain on another slide from the same print run and 4) normalized to intra-spot GAPDH measured on the same experimental slide using the third (Q585) channel.  For cell lysates on this array, intra-spot normalization provided the best replicate consistency (average CV = 10.1%) when compared to no normalization (average CV = 17.1%), external total protein normalization (average CV = 13.1%) and external GAPDH protein normalization (average CV = 17.1%) (Figure 8).  This illustrates how implementing intra-spot normalization mitigates typical variability inherent with microarray printing and customary assay processing providing the most accurate signal quantitation within and between experiments.

Near IR Detection Systems Comparison – Multi-color near-IR detection is not a novel technique with other suppliers offering near-IR organic fluors and confocal microarray scanners capable of performing similar multi-color methods.  Here, we compare a complex existing system to the faster and less expensive ArrayCAM system.  Multi-color detection using quantum nanocrystals (QDots) and LI-COR probes were compared on the same C32 cell lysate time-point arrays as previously described.  Slides were probed with P-ERK, cPARP or P-S6 antibodies (normalized with GAPDH) then detected with multiple wavelength probes for both QDots and LI-COR respectively.  For QDot detection, the slides were imaged in under 1 minute per color using the ArrayCAM.  For LI-COR probes the slides were imaged in 5 minutes per color using an Innopsys 710-IR scanner.  The QDot/ArrayCAM method proved as or more sensitive than the more widely used and expensive LI-COR/Innopsys method, detecting the same signaling changes and in some cases detecting them earlier (Figure 9).  This demonstrates the rapid and lower cost ArrayCAM/QDot method is equivalent and in some cases superior to the LI-COR/Innopsys system in terms of sensitivity and ease of use.

Quantum Nanocrystal versus Amplified Colorimetric Detection – The absolute protein mass detection sensitivity was demonstrated between quantum nanocrystals (QDot) and colorimetric detection using DAB with catalyzed signal amplification.  Both systems were employed to detect GAPDH in arrayed serially titrated NIH 3T3 and Jurkat cell lysates of known total protein concentrations.  The colorimetric system is traditionally performed using an auto-stainer instrument.  Therefore, in the present case, the QDot assay was observed using both manual benchtop and automated staining methods.  Sensitivity was determined for each method using respective signal minus background (S-B) of replicate blank plus two standard deviations to derive an LOD using linear regression against the S-B cell lysate titration series of known concentrations.  Results obtained using unamplified QDots were consistently more sensitive than the standard amplified colorimetric detection system (Figure 10).

Conclusions

We have shown that the Grace Bio-Labs ArrayCAM Microarray System provides a valuable sensitive, accurate, easy to use and economical tool for investigators or clinicians in their study of multiple protein signaling pathways or detection and monitoring of important clinical biomarkers.  The system delivers sensitivity equivalent or superior to current benchmark RPPA methods for detection of low-abundance proteins.  Utilizing multi-wavelength detection assay methods provides improved data normalization accuracy and higher throughput not possible using standard methodologies.  This translates into improved consistency and more robust data sets within and between assays.  Rapid, easy to use instrumentation and software present a low barrier to adoption when compared to more complex and costly systems.  This facilitates rapid implementation and greatly accelerates access to data and publication.  

Technical Resources

Institute Curie Reverse Phase Protein Array Platform –

http://www.institut-curie.org/recherche/reverse-phase-protein-array-platform

Reverse-Phase Protein Array (RPPA) Assay System User Guide –

https://www.gracebio.com/media/RPPA_Assay_System_Guide_1.0.pdf

Assay Kinetics for Quantum Dots on Protein Arrays –

https://www.gracebio.com/research/assay-kinetics-for-quantum-dots.html

Reverse phase protein array based tumor profiling identifies a biomarker signature for risk classification of hormone receptor-positive breast cancer – Sonntag, et al. doi:10.1016/j.trprot.2014.02.001

http://www.sciencedirect.com/science/article/pii/S2212963414000023

Improved protein arrays for quantitative systems analysis of the dynamics of signaling pathway interactions – Wang et al. Proteome Science 2011, 9:53 – doi:10.1186/1477-5956-9-53

http://www.proteomesci.com/content/9/1/53

Multiplex Protein Arrays with Quantum Dots

https://www.gracebio.com/research/multiplex-protein-arrays.html

References

1. Espina, Virginia, Lance A. Liotta, and Emanuel F. Petricoin III. “Reverse-phase protein microarrays for theranostics and tatient tailored therapy.” Tumor Biomarker Discovery. Humana Press, 2009. 89-105.
2. Becker, Karl-Friedrich.  “Using tissue samples for proteomic studies – Critical considerations.” PROTEOMICS-Clinical Applications (2015).
3. Wang, Xiaoyu, et al. “Improved protein arrays for quantitative systems analysis of dynamic signaling pathway interactions.” Proteome Science 9.1 (2011): 1-13
4. Calvert, Valeri S., et al.  “Development of multiplexed protein profiling and detection using near infrared detection of reverse-phase protein microarrays.” Clinical Proteomics 1.1 (2004): 81-89
5. De Luca, Antonella, et al.  “The RAS/RAF/MEK/ERK and the PI3K/AKT signaling pathways: role in cancer pathogenesis and implications for therapeutic approaches.”  Expert opinions on therapeutic targets 16.S2 (2012): S17-S27