Many human diseases have been linked to malformed or malfunctioning proteins. Sickle cell disease is one example.
According to the central dogma of life, DNA becomes RNA, which becomes protein. Moving from each step to the next adds an increased layer of complexity. Proteins are expressed at different times, take on different forms, and act differently in various biological situations. Because of this, the “proteome,” previously defined as the total protein complement of a genome, can morph and change constantly.
The complexity of the proteome is also one of its strengths. Because of the diversity of available protein forms, each protein can be functionalized to fit a specific purpose – whether that is acting as a signaling molecule or catalyzing a reaction as an enzyme. Many human diseases have also been linked to malformed or malfunctioning proteins. Sickle cell disease, where red blood cells become abnormally shaped, is associated with mutated hemoglobin protein.
Given the importance of the proteome, why has proteomics (the systematic and large-scale analysis of the proteome) not been as popularized as transcriptomics/epigenomics/genomics? According to Parag Mallick, Ph.D., co-founder and chief scientist of Nautilus Biotechnology, which is developing a single-molecule platform to understand the proteome, part of this relative unpopularity is the sheer challenge of measuring something that is so diverse and constantly changing.
“Proteins are complexly structured, dynamic, and breathing,” Mallick said. “And another real challenge is the range of concentrations; most cells have one copy of your genome, but they’ve got anywhere from one copy to millions of copies of different proteins.”
These sentiments were echoed by Daniel Hornburg, Ph.D., senior director of research and tech development at Seer, a biotechnology company using nanoparticles to understand the proteome. “We probably have more than a million different proteoforms (proteins with different splice forms, modifications and a combination of those), and you basically increase the complexity of the biological information as you move toward the phenotype,” he said.
Despite the difficulties of characterizing and quantifying the proteome, researchers have developed several tools to do so. One of these tools is mass spectrometry, where either proteins or peptides (little bits of protein after digestion) are ionized and fragmented into smaller molecules to generate spectral peaks. These experimentally generated spectral peaks are compared to a library of known molecular peaks to infer the identity of the original protein.
However, scientists have noted some limitations of mass spectrometry that include incomplete detection of rare proteins and protein variants (like proteins with post-translational modifications) due to the large dynamic range of abundance.
Seer, which spun out of Omid Farokhzad’s lab at Harvard Medical School, uses nanoparticle-based technology to compress the dynamic range of protein abundances to better understand the proteome. As Hornburg explained it, Seer’s platform relies on engineered nanoparticles, which preferentially bind certain proteins based on a combination of characteristics like surface charge and chemical functionalization. The proteins bound to the nanoparticles are then fragmented into peptides and characterized through mass spectrometry via a bottom-up approach.
“Each of these particles is engineered in a specific way,” Hornburg said. “It’s a combination of, for example, having a specific surface charge, having some hydrogen donors, having specific molecule functionalization. A combination of these different factors will make it more attractive to a subset of the proteome.”
Hornburg explained that the nanoparticles are designed based on the principle that certain shared protein characteristics are broad enough to provide an adequate sampling of the proteome but narrow enough to select for rare proteoforms that may otherwise be lost. As a proof-of-concept, the company recently published a pre-print on using their platform, Proteograph, to identify several hundred protein variants in patients with non-small-cell lung cancer versus healthy controls.
The notion of preferentially selecting proteins is linked to another proteomics method known as affinity binding. Affinity binding drives the concept behind popular assays like western blotting and ELISAs, where antibodies or other molecules can bind to and select for very specific proteins. One of the limitations of this approach is that because the protein-of-interest must be known beforehand, inherent bias is introduced. Discovery of rare or previously unknown proteoforms, for instance, can be difficult to achieve with this method.
According to Mallick, Nautilus’ platform aims to utilize affinity binding on a much larger scale to profile more protein molecules in the proteome. “Instead of trying to put down your affinity reagent and grab things out of solution, we actually immobilize the sample on this fabricated nanochip in a way that each individual molecule lives in its own cell,” Mallick said. “Picture a giant chessboard, and every single one of those positions in the chessboard having exactly one protein molecule.”
Mallick explained that by probing this “chessboard” with different labeled binding molecules, proteins can be identified and quantified through imaging to look at brightness or intensity based on which binding molecule is used. The company hopes to use this approach to identify disease biomarkers or to see a drug’s effects on the proteome, among other applications.
As with the other “omics” – transcriptomics, epigenomics, and genomics – proteomics is also undergoing a push toward higher cell resolution. Isoplexis, a biotechnology company focused on developing antibody-based proteomics assays for single cells, leverages the previously mentioned affinity binding method to look into proteomes at the single-cell level. Jing Zhou, M.D., Ph.D., Isoplexis’ chief scientific officer, told BioSpace in an e-mail that “bulk methods average across all cells, losing critical cellular attributes key to understand response[s] in patients.”
Isoplexis’ platforms utilize the affinity binding approach of antibody-based selection to assay for “functional proteins” in single cells. Zhou described this as “functional phenotyping,” where, in the case of immune cells, the company’s IsoSpark platform could be used to detect “superhero” cells that secrete multiple cytokines/chemokines critical to the immune response. Recently, in a Cell paper, scientists used the IsoLight platform to determine the proteomic functionalities of various immune cells during long COVID.
And, in a true multi-omic fashion, Zhou noted the recent release of Duomic, a platform to “simultaneously measure functional protein and gene expression levels from the same cell, with the goal of identifying which genes and gene pathways have a downstream impact on protein heterogeneity in these cells.”
Despite the complexity of the proteome, researchers continue to unpack its many nooks and crannies. Much of this effort, too, is to better understand how proteins operate in human health and disease. “The vast majority of diseases that we’re struggling with today are not necessarily genetic in origin,” Mallick said. “They come from environmental factors, or aging or inflammatory conditions – [and] proteins have had an outsized impact on those conditions.”
