What is proteomics?

What is a protein and why does it matter?

A protein is a chain of building blocks called amino acids, folded into a particular shape, and the shape determines its function. DNA codes what proteins could be produced. Proteins do the actual work in the body. Hemoglobin carries oxygen; insulin regulates blood sugar; enzymes digest lunch; antibodies fight infection.

One individual's genome is essentially the same in every cell and throughout its life. A caterpillar that turns into a butterfly has the same genome throughout its life; what differs is which proteins are being made.

The same is true for a neuron and a skin cell within an individual. RNA helps to pass on the instructions to produce certain proteins, but it is an imperfect proxy for proteins. RNA levels correlate loosely with the protein it encodes.

DNA tells you what could happen. Proteins tell you what is happening.

The full set of proteins present in a cell, tissue, or sample – its proteome – is also far larger than the genome that encodes it. Humans carry roughly 20,000 protein-coding genes, but the number of distinct protein forms, or proteoforms, runs past a million.

What is proteomics?

Proteomics is the large-scale measurement of proteins: which ones are present, in what quantities, and in what forms.

Analyzing DNA has become relatively routine. A human genome – three billion letters – can be sequenced in about a day for a few hundred dollars in sequencing costs. For more details, see the primer on genetic sequencing.

Systematically quantifying proteins, on the other hand, is much more difficult.

Today, proteins are typically measured one at a time – a pregnancy test detects a single protein hormone, a PSA screens for one protein from the prostate, a troponin test looks for one protein that leaks from injured heart muscle.

Proteomics is doing this at a much larger scale: reading the full set of proteins in a sample. This primer focuses on blood samples.

Why is measuring proteins harder than reading DNA?

Proteins are difficult to measure at scale because:

• Proteins cannot be copied and amplified easily.
• Proteins are built from 20 different amino acids and can take many different forms, instead of just 4 nucleotide bases.
• Abundance of certain proteins in a sample can "drown out" the proteins of interest.

Genetic sequencing has a trick that proteins do not allow: DNA can be copied. PCR (polymerase chain reaction) turns a trace of DNA into billions of identical copies, so even tiny amounts can be read. Proteins cannot be copied the same way. So protein detection is limited to the amount present in a sample.

Reading DNA is essentially categorizing the order of 4 nucleotide bases. But the human body contains roughly 20,000 proteins encoded directly by our genes, and the total number of unique protein forms can reach into millions.

Proteins of interest tend to be over-crowded by the abundant proteins in a sample. In a typical blood sample, albumin alone makes up more than half of the protein in plasma. The top 20 most abundant proteins make up ~99% of the protein in the sample, and the remaining thousands are ~1%.

Exhibit 1Illustrative view of protein distribution in a sample

How are proteins measured today?

This primer groups proteomics into three main types:

• Immunoassay (using antibodies)
• Mass spectrometry
• DNA-tagged

Exhibit 2Trade-offs of different types of proteomics

How does immunoassay work?

It uses an antibody as a lock that fits one protein key, and attaches something readable to the moment of binding.

For decades the standard protein test has been the immunoassay, and its most common form, the ELISA (enzyme-linked immunosorbent assay), works by attaching a color-producing enzyme to the antibodies.

An antibody is a type of protein that binds to a specific targeted spot within a certain protein molecule. Over the past several decades, scientists have learned to produce an antibody against almost any protein they choose – and those antibodies are used to detect proteins.

In ELISA, enzymes are attached to antibodies, and the antibodies bind to specific proteins. Then a chemical is added, and the enzyme turns that chemical into a specific color. Whether or not the test shows color (and the intensity of the color) is the readout: it tells you the protein is present, and roughly how much. This format is cheap and standardized, and it sits behind most routine protein and hormone tests.

These traditional immunoassay tests are proven, but limited because each test only measures one protein.

Next-gen immunoassay uses the same antibody-and-dye principles, but does more per run. Instead of one color, each test run can consist of a handful of antibodies with distinct signals, and modern imaging technology quantifies each signal. The most sensitive next-gen tests also count protein molecules one at a time rather than reading a pooled color, and can detect proteins at concentrations roughly a thousand times lower than a standard ELISA.

Exhibit 3How immunoassay works

The benefit of the immunoassay methodology is that the tests are highly specific. The disadvantage is that even next-gen immunoassay tests are limited in how many proteins each test run can detect, which can make it costly and time-intensive to do proteomics at scale.

How does mass spectrometry work?

Mass spectrometry splits proteins into tiny fragments and measures the mass of each of them. An enzyme chops proteins into short fragments called peptides. A column then spreads the peptides out over time by their chemistry – a step called liquid chromatography, which is why the method is usually written LC-MS – so they reach the instrument a few at a time rather than all at once. Each is given an electric charge and flown through electric fields; the instrument computes its mass to within a few parts per million, and software matches those masses to identify each fragment.

Exhibit 4How mass spectrometry identifies a protein

Mass spectrometry is considered the gold standard in protein analysis because it is highly specific – it confirms what a protein is from its mass, rather than trusting an antibody – and because it does not require a pre-defined target list. It is particularly useful for research, in case there are proteins in the sample that nobody thought to look for.

The disadvantage of mass spectrometry is that it tends to be expensive and requires specialist operators. Furthermore, abundant proteins dominate a plasma sample: run as-is, mass spectrometry mostly sees the few hundred most common proteins. To go deeper, labs either strip out the abundant proteins or enrich the rare ones first. This preparation step is labor and time intensive, which the immunoassay technology does not require.

How does DNA-tagged proteomics work?

There is an emerging category of proteomics that uses DNA tags. An antibody – or another engineered binder – is bound to a short DNA tag (sometimes called a barcode). When the binder finds its protein, the tag is amplified by PCR and counted on a sequencer. It is another way of reading out a binder, without relying on color detection.

Exhibit 5How a DNA tag stands in for a protein

Because the tags are read by a sequencer rather than by color, one run can carry thousands of different tags at once – which is how these methods measure thousands of proteins from a few drops of blood.

DNA-tagged proteomics shares immunoassay's main limitation: it only finds the proteins it was told to look for. Mass spectrometry is the only one of the three that needs no target list. That makes DNA-tagged methods less suited to open-ended discovery research.

Why does proteomics matter?

Medicine already runs on protein tests. Troponin, PSA, and the pregnancy test are all protein tests, and only about 150 proteins are measured this way in routine clinical practice.

And now research proteomic tools are able to detect 10,000+ proteins in blood. As proteomics technology has advanced, three key things have changed:

• A single run can now measure thousands of proteins from a few drops of blood.
• The most sensitive methods can detect proteins that leak from the brain into the bloodstream, at concentrations far too low for older tests.
• Large population studies have measured thousands of proteins across tens of thousands of people and linked the levels to genes, diagnoses, and outcomes – the raw material for working out which proteins mean something.

Exhibit 6What proteomics enables

For research, proteomics is becoming what sequencing became for DNA – a routine, population-scale measurement rather than a specialist's experiment. The vast majority of approved drugs act on a protein. Measuring thousands of proteins across large populations is how new targets and biomarkers are found, and how a trial can see whether a drug is working.

For medicine, a panel of proteins could be read at once, disease could be detected earlier and more easily, and a disease or a treatment could be monitored over time.

Takeaways

• Proteins do the work; DNA is the blueprint. The genome is fixed for life, but the proteins show what is actually happening.
• Proteomics is measuring proteins at scale. Medicine has measured proteins one at a time for decades; proteomics reads thousands from a single sample.
• Three categories of proteomics, each with trade-offs. Immunoassay is specific but narrow; mass spectrometry needs no target list but is costly and laborious; DNA-tagged methods are broad and sensitive but only find what they are told to look for.
• Proteomics technology is far ahead of current clinical practices. Routine clinical tests are limited to about 150 proteins, while research platforms are now able to detect 10,000+ proteins. Advances in proteomics could lead to major changes in how patients are diagnosed and treated.

General, educational, and informational research only, not tailored to your situation. Nothing here constitutes investment, legal, medical, or other professional advice; an offer to sell or a solicitation of an offer to buy any security; promotional or marketing material; or a recommendation. The author may hold positions in the securities or sectors discussed. Do your own research and consult a licensed professional. Full disclosures at www.eastrivernotes.com/disclosures.