Protein Structure

Lecture 11/12 Video 3

🧬 Big Picture: What is Proteomics?

Proteomics = studying all proteins (the proteome) in a system at a given time

📌 Key idea:

Genome = what could happen
Proteome = what is actually happening right now

Why not just study genes?

Because biology adds layers of complexity:

Level	Approx. complexity
Genome	~20,000 genes
Transcriptome	~100,000 RNAs
Proteome	>1,000,000 proteins 😳

👉 Why the explosion?

Alternative splicing
mRNA editing
Post-translational modifications (PTMs)

➡️ Conclusion: Proteins are far more dynamic than genes They reflect the real-time state of the cell.

📄 Source:

🔬 Two Main Proteomics Strategies

1. Top-Down Proteomics

Analyze intact proteins
Then fragment them

✅ Pros:

Keeps full protein info (including PTMs)

❌ Cons:

Hard to handle (big, complex, messy spectra)

2. Bottom-Up Proteomics (focus of your file)

💡 Flip the strategy:

Cut proteins → analyze peptides

Workflow:

Digest protein → peptides
Separate peptides (LC)
Analyze with MS1
Fragment selected peptides → MS2

📄 Source:

💥 Shotgun Proteomics (Real-World Mode)

Now scale it up:

Instead of 1 protein → many proteins

Example:

10,000 proteins
~20 peptides each 👉 ~200,000 peptides 🤯

Problem:

Too complex to analyze directly

Solution:

👉 LC (liquid chromatography) separates peptides over time

⚙️ Full Workflow Explained

Step 1: Digest proteins

Break into peptides

Step 2: LC separation

Spread peptides over time (critical!)

Step 3: MS1 (survey scan)

Detect all peptide ions
Measure:
- m/z (mass-to-charge)
- charge states

Step 4: Precursor selection

Pick one peptide ion

Step 5: MS2 (fragmentation)

Break peptide → fragments

Step 6: Database matching

Compare experimental spectra to theoretical ones

📌 Important insight: You don’t identify proteins directly 👉 You identify peptides → then infer proteins

📄 Source:

🤖 Why Databases Are Essential

Manual interpretation = impossible

Instead:

Generate theoretical peptides from genome
Simulate fragmentation
Match to real data

📌 This is why: 👉 Genome → needed to build proteome database

✂️ Protein Digestion (Very Important)

Why digest proteins?

Proteins are:

Too big
Too highly charged (+40, +50, etc.)

👉 Leads to messy spectra

Ideal peptide properties

Feature	Requirement
Length	6–30 amino acids
Why?	Balance between uniqueness & detectability

⚠️ Key reasoning (often tested!)

Too short → not unique (many proteins share them)
Too long → poor ionization

🔬 Combinatorics Insight

With 20 amino acids:

3 aa → 20³ = 8,000 combinations ❌ (too few)
7+ aa → huge diversity ✅

👉 That’s why peptides must be ≥6–7 aa

🧪 Sample Preparation (Critical Step)

1. Reduction

Break disulfide bonds

2. Alkylation

Prevent re-formation

👉 Keeps proteins linear and analyzable

✂️ Cleavage Methods

Chemical cleavage

Example:

Cyanogen bromide → cuts at methionine

❌ Problem:

Too few cuts → peptides too large

Enzymatic cleavage (BEST)

⭐ Trypsin (the GOAT)

Cuts after:

Lysine (K)
Arginine (R)

Why trypsin is perfect:

✅ Predictable → easier database search ✅ Produces ideal peptide lengths (~13 aa avg) ✅ Adds positive charges → better MS detection

📌 Key insight:

Peptides end with K or R → better ionization

📄 Source:

⚡ Peptide Fragmentation

Peptides break at backbone bonds → generate ions

Ion types:

Type	Description
b-ions	from N-terminus
y-ions	from C-terminus
a/x/c/z	less common

👉 Most important: b and y ions

🧠 How Sequencing Works (Core Concept)

Key rule:

b-ions → read from N-terminus
y-ions → read from C-terminus

🧮 Mass Logic

You identify amino acids by:

👉 mass differences between peaks

Example:

Difference = 113 → Leucine (L) or Isoleucine (I)

⚠️ Important limitation:

L and I cannot be distinguished (same mass)

Fragment Mass Calculations

b-ion:

Sum of residues + H

y-ion:

Sum of residues + H₂O + H

📄 Source:

🔍 Example Logic (What You Actually Do)

Identify y1 peak → C-terminal amino acid
Subtract adjacent peaks → find residues
Build sequence step-by-step
Reverse to N→C direction

🤯 Real Data is Messy

Not clean like textbook examples:

You’ll see:

Mixed b and y ions
Noise
Overlapping peaks
PTMs (e.g. phosphorylation)

Example complication:

Phosphorylated serine → can lose phosphate → multiple peaks

👉 This is why: software is absolutely required

⚡ Fragmentation Methods

CID / HCD (collision-based)

Most common
Produces b and y ions

ETD (electron-based)

Produces c and z ions
Useful for PTMs

🔑 Key Takeaways (Exam-Level)

Core Concepts:

Proteome ≠ genome (much more complex)
Bottom-up = digest → analyze peptides
Trypsin = standard enzyme
LC-MS/MS = essential workflow
Identification = database matching

Critical Understanding:

Peptide length matters (6–30 aa)
Fragmentation gives sequence info
b/y ions are primary tools
Mass differences = amino acid identity

Practical Reality:

Data is huge → requires software
Spectra are messy
Database quality is crucial

🧠 Mental Model (Simplified)

Think of it like this:

🧩 Protein = sentence ✂️ Digest = cut into words 💥 Fragment = break words into letters 🧠 Database = dictionary 👉 Match letters → words → sentence

Quiz

Score: 0/31 (0%)

Q0. In shotgun bottom-up proteomics, why are proteins usually digested into peptides before MS analysis?

Because peptides are generally easier to ionize, separate, fragment, and analyze than intact proteins

Because intact proteins cannot contain post-translational modifications

Because peptides always have only one possible charge state

Because digestion removes all database-searching requirements

Q1. What is the main idea of proteomics compared with genomics?

Proteomics studies only DNA mutations, while genomics studies proteins

Proteomics investigates the proteins present in a system at a given time, while genomics describes the genetic potential

Proteomics always requires no prior genome information

Proteomics only measures mRNA expression

Q2. Why can a relatively small number of genes produce a much larger number of protein forms?

Because every gene is translated into exactly one protein with no variation

Because alternative promoters, alternative splicing, mRNA editing, and post-translational modifications increase molecular diversity

Because proteins are copied directly from lipids

Because mass spectrometry artificially creates new proteins during ionization

Q3. In bottom-up proteomics, protein identification is usually built from which type of evidence?

Only intact protein mass

Only the color of the chromatogram

Multiple peptide identifications that map back to the same protein

Only the number of genes in the genome

Q4. Why is liquid chromatography especially important in shotgun bottom-up proteomics?

It converts proteins into DNA

It separates complex peptide mixtures over time before MS analysis

It prevents all peptides from becoming charged

It replaces the need for MS/MS fragmentation

Q5. What happens during the MS1 stage in an LC-MS/MS shotgun proteomics workflow?

Peptides are enzymatically digested

Precursor peptide ions are detected and measured by m/z

The database is physically loaded into the mass spectrometer

Only y-ions are produced

Q6. What is a precursor ion in tandem mass spectrometry?

The selected ion that is isolated and fragmented to generate product ions

A protein before digestion

A neutral amino acid that cannot be detected

A database entry with no experimental signal

Q7. Why is database searching essential in shotgun proteomics?

Because the amount of spectral data is too large and complex for manual interpretation

Because databases replace the need for experimental MS data

Because databases physically digest the sample

Because MS/MS spectra contain no sequence information

Q8. What does an in silico digest do during database searching?

It chemically digests the experimental sample

It predicts which peptides would be generated from database proteins using a specified cleavage rule

It removes all PTMs from the sample

It converts peptide spectra into RNA sequences

Q9. Why is trypsin especially useful in bottom-up proteomics?

It cuts randomly after every amino acid

It cleaves predictably after lysine and arginine, often generating peptides of useful length and charge

It only cleaves proteins at methionine residues

It prevents peptides from being ionized

Q10. Why are peptides of about 6 to 30 amino acids generally useful in bottom-up proteomics?

They balance sequence uniqueness with manageable size for ionization and fragmentation

They are too short to be identified

They cannot carry post-translational modifications

They are always singly charged

Q11. Why are very short peptides problematic for protein identification in complex samples?

They are always too heavy for MS detection

They have limited sequence combinations and may occur in many proteins

They cannot be produced by enzymes

They always fragment into only one ion

Q12. What is the purpose of reducing proteins before digestion?

To break disulfide bonds and make the protein chains more linear

To add phosphate groups to serine residues

To convert peptides into RNA

To remove all lysine and arginine residues

Q13. Why is alkylation commonly performed after reduction?

To block free thiols and prevent disulfide bonds from reforming

To restore all original disulfide bonds

To cleave specifically after methionine

To remove the need for LC separation

Q14. Which fragment ion series are most commonly produced by collision-based peptide fragmentation methods such as CID and HCD?

Only x and z ions

Mainly b and y ions

Only intact precursor ions

Only neutral amino acids

Q15. Proteomics is useful because the proteome reflects the current biological state more directly than the genome.

True

False

Q16. In bottom-up proteomics, intact proteins are usually analyzed first and only then digested into peptides.

True

False

Q17. Shotgun bottom-up proteomics usually handles complex mixtures containing many proteins rather than only one purified protein.

True

False

Q18. LC separation is unnecessary in shotgun proteomics because MS/MS can easily analyze hundreds of thousands of peptides at once.

True

False

Q19. A high-quality protein database is important because experimental spectra are matched against theoretical peptide spectra.

True

False

Q20. Trypsin cleaves mainly after arginine and lysine residues.

True

False

Q21. Trypsin is useful partly because it often leaves a basic residue at the C-terminus of the peptide.

True

False

Q22. Chemical cleavage with cyanogen bromide cuts after methionine, but this may generate peptides that are too large for ideal bottom-up proteomics.

True

False

Q23. The main reason peptides should not be too short is that short sequences may not be unique enough to identify a specific protein.

True

False

Q24. b-ions are read from the C-terminus, while y-ions are read from the N-terminus.

True

False

Q25. Mass differences between neighboring fragment-ion peaks can reveal the identity of amino acid residues.

True

False

Q26. Leucine and isoleucine are usually difficult to distinguish by standard MS/MS because they have the same mass.

True

False

Q27. CID and HCD mainly generate c- and z-ion series.

True

False

Q28. ETD can be useful in proteomics because it produces c- and z-ion series and can preserve some labile modifications better than collision-based methods.

True

False

Q29. A real MS/MS spectrum is always clean and contains only one complete y-ion series.

True

False

Q30. Post-translational modifications such as phosphorylation can complicate MS/MS interpretation.

True

False