Research

Systems biology of intracellular signaling networks at the single-cell resolution

Our lab uses Mass Cytometry to profile the states and dynamics of up to 40 intracellular phosphorylation sites on key signaling proteins in single cells. This comprehensive measurement enables us to precisely determine the signaling and physiological states of each individual cell, including growth, proliferation, and responses to stress and inflammation, and to understand how these signaling activities integrate to determine cell fate.

Single-cell signaling network profiling with mass cytometry
Figure 1. Single-cell signaling network profiling with mass cytometry. a, Using mass cytometry-based single-cell analysis, we quantified how the expression variances of signaling proteins modulated the states and dynamics of EGFR signaling networks. b, Expanding this analysis to a kinome and phosphatome-wide scale, we systematically investigated how expression variances of each signaling protein influence network dynamics and cell fate. This work led to the identification of key tyrosine kinases as novel therapeutic targets in drug-resistant melanoma cells.

Based on this technology, we established the first single-cell quantitative analysis of EGFR networks using mass cytometry that allowed us to interrogate the protein abundance-dependent signaling effects1. We coupled this approach with transient overexpression and a novel measure, called BP-R2, to reveal the intricate modulation of signal amplitudes and peak times as functions of continuous protein abundance1 (Fig. 1a). These analyses characterized how the overexpression of key signaling proteins within the MAPK/ERK pathway alters intracellular signaling networks to favor cancer cell proliferation and enabled the quantitative identification of protein abundance thresholds determining cell differentiation, transition, and transformation at phenotypical switching points. Additionally, our study indicated that treatments for tumors with high heterogeneity in signaling protein abundance need to be designed to account for abundance-dependent modulation of network state.

Applying this single-cell signaling network analysis in a human kinome- and phosphatome-wide study, we systematically assessed how 649 signaling proteins, spanning four orders of magnitude in abundance, modulated cancer-related signaling in HEK293T cells2 (Fig. 1b). These comprehensive analyses broadened the functional classification of human kinases and phosphatases, and unveiled their non-catalytic roles in driving cancer-related signaling. More importantly, our work revealed a drug-resistant mechanism through which overexpression of tyrosine kinases, including SRC, FES, YES1, and BLK, induced MEK-independent ERK activation in melanoma A375 cells. These proteins could predict drug sensitivity to BRAF-MEK concurrent inhibition in cells carrying BRAF mutations. This result underscores the applicability of our innovative mass cytometry-based approach in the identification of novel therapeutic targets2 (Fig. 1b). 

Fig. 2
Figure 2. CellCycleTRACER analyzes and accounts for heterogeneities caused by cell size and cell cycle effects.

Like many other single-cell technologies, mass cytometry is greatly confounded by the variations of cell size and cell cycle in measured single cells. To address these impacts, we developed a supervised comprehensive analysis tool called CellCycleTRACER3. This novel approach includes an NHS chemistry-based total protein staining method to correct for cell size confounding effects. A machine learning algorithm is then applied to reconstruct a cell cycle pseudotime that can be used to normalize the variance caused by cell cycle effects (Fig. 2). CellCycleTRACER profiles heterogeneities caused by differential cell cycle phases and effectively addresses the influences of cell cycle and cell size effects within the context of single-cell signaling network analysis3. This method is particularly important when highly heterogeneous cell populations with deregulated cellular processes, as typically found in tumors, are analyzed. 

Through comprehensively interrogating the heterogeneous intracellular phosphorylation networks involved in cancer progression, our goal is to elucidate the previously unprecedented intracellular signaling responses underlying the diverse susceptibility to cancer treatments and specific kinase inhibitors employed in clinical settings. By employing advanced screening methods and novel computational models, our strategies will result in the identification of new druggable targets within cancer signaling networks and the development of novel therapeutic approaches to address drug resistance caused by signaling network abnormalities. In the long term, we envision that the knowledge generated from these studies will contribute to the establishment of a cancer signaling network atlas that maps various intracellular signaling network profiles to cell susceptibility to available therapeutic options. Ultimately, our objective is to enhance the outcomes for cancer patients by driving the discovery of new druggable targets to improve current cancer treatment options.

 

ACE fig
Figure 3. Signal amplification by ACE enables high-sensitivity mass cytometry analysis. a, ACE implements thermal-cycling-based DNA in situ concatenation in combination with 3-cyanovinylcarbazole phosphoramidite (CNvK)-based ultra-fast DNA crosslinking to achieve signal amplification in mass cytometry analysis. b, ACE amplifies the signal in suspension mass cytometry, allowing detection and quantification of low-abundance markers which do not reach detection limits of a mass cytometer using conventional approaches. c, Coupled with imaging mass cytometry, ACE enables high-sensitivity multi-parametric spatial profiling of healthy and diseased tissue samples. 

Advancing single-cell and 'omics technologies through the engineering of programmable DNA probes

We engineer dynamic and reactive DNA probes and integrate them with advanced single-cell and 'omics methods to develop technologies for in-depth exploration of intracellular signaling networks, protein-protein interactions, and spatial proteomics, as well as for comprehensive investigation of disease mechanisms through multi-omic single cell profiling. 

We addressed the sensitivity limitation in mass cytometry-based single-cell and spatial tissue profiling through the development of a signal amplification approach termed Amplification by Cyclic Primer Extension (ACE). ACE combines thermal-cycling DNA in situ concatenation with 3-cyanovinylcarbazole phosphoramidite (CNvK)-based ultra-fast DNA crosslinking to achieve >500-fold signal amplification simultaneously on >30 protein epitopes, allowing unprecedented questions in the realm of the low-abundance proteome to be addressed (Fig. 3a). We demonstrated the utility of ACE in low-abundance protein quantification with suspension mass cytometry to characterize key phenotypical switches and the dynamics of signaling responses in mammalian cells (Fig. 3b). We further showed the application of ACE in imaging mass cytometry (IMC)-based multi-parametric tissue imaging to identify tissue compartments and profile pathological states in diseased human kidney tissues. This implementation of ACE also allows for improvement in the spatial resolution of IMC by compensating for the signal loss resulting from reducing the laser ablation crater size (Fig. 3c).

apr mechanism
Figure 4. Profiling intracellular protein complexes with smAPR.  The smAPR DNA nanodevices are conjugated with antibodies designed to label specific intracellular proteins of interest. When these labeled proteins bind together, they bring two smAPR probes into close proximity, initiating the APR reaction and resulting in the release of a full-record (double-stranded DNA) carrying both probe identities (barcodes).  Upon the full-record release, parental probes are disassembled and reset to their initial states, ready for the next round of reaction between the same pair of probes or with a different partner. Through next generation sequencing, the identities of interactive proteins can be revealed to allow the interrogation of multi-subunit protein assemblies.   

As a recent focus in our lab, we have developed DNA technology-based methodologies to advance the understanding of protein assembly and disassembly dynamics. We engineered a DNA nanodevice called single-molecule auto-cycling proximity recording (smAPR), which allows systematic profiling of intracellular protein-protein interactions. Conjugated to antibodies and applied to cell or tissue samples, smAPR uniquely generates repeated proximity records from barcoded intracellular neighboring epitopes, effectively "reversely translating" single-molecule protein-protein interaction (PPI) information into DNA sequences that can be read by next-generation sequencing (NGS). The recorded sequences are clustered by overlapping barcodes (e.g., proximity records A-B, B-C, and A-C can be made from targets A, B, and C in a protein complex) to reveal the protein-protein interactions and protein assembly arrangements in examined samples (Fig. 4).

smAPR is the first method of its kind to record all interactions within a pre-formed protein complex, allowing for the detection and quantification of multivalent interactions and protein assemblies. With its high-throughput analysis capability and compatibility with other sequencing-based 'omics approaches, smAPR is a versatile tool for systematically interrogating various biological mechanisms. In the near future, our laboratory will apply smAPR to decode the mechanisms behind drug resistance in complex cancer cell populations.

By combining programmable DNA reactions with advanced single-cell and 'omics approaches, we aim to develop innovative technologies for the systematic investigation of heterogeneous phosphorylation and interaction networks associated with patient outcomes across diverse cancer types.

References

  1. X.-K. Lun, V. R. T. Zanotelli, J. D. Wade, D. Schapiro, M. Tognetti, N. Dobberstein, and B. Bodenmiller, “Influence of Node Abundance on Signaling Network State and Dynamics Analyzed by Mass Cytometry,” Nat. Biotechnol., Jan. 2017. 

  2. X.-K. Lun, D. Szklarczyk, A. Gábor, N. Dobberstein, V. R. T. Zanotelli, J. Saez-Rodriguez, C. von Mering, and B. Bodenmiller, “Analysis of the Human Kinome and Phosphatome by Mass Cytometry Reveals Overexpression-Induced Effects on Cancer-Related Signaling,” Mol. Cell, May 2019. 

  3. M. A. Rapsomaniki*, X.-K. Lun*, S. Woerner, M. Laumanns, B. Bodenmiller, and M. R. Martínez, “CellCycleTRACER Accounts for Cell Cycle and Volume in Mass Cytometry Data,” Nat. Commun., Feb. 2018.

  4. X.-K. Lun, K. Sheng, X. Yu, C. Y. Lam, G. Gowri, M. Serrata, Y. Zhai, H. Su, J. Luan, Y. Kim, D. E. Ingber, H. W. Jackson, M. B. Yaffe and P. Yin, “Signal Amplification by Cyclic Primer Extension Enables High-Sensitivity Single-Cell Mass Cytometry.” Nat. Biotechnol., Jul. 2024.