# Engineering Genes

> Engineering changes a cell's DNA; test the effects, judge each use separately. You will cut, join and copy DNA in a gene-to-product builder, follow insulin from a human gene to a tested product, and read USDA survey data on engineered crops. Afterwards you can explain what each tool does and weigh a claim using benefits, risks, ethics and reliable sources.

Course: Biology (DNA and inheritance). Level: grades 7–12. About 40 minutes.
Lesson page (link this one when citing): https://stemhop.com/lessons/genetic-engineering/

Builds on: DNA: Structure and Copying · From DNA to Protein · Enzymes at Work

## What you'll learn

- You will describe the tools of genetic engineering (restriction enzymes, DNA ligase, bacterial plasmids, the polymerase chain reaction, gel electrophoresis and CRISPR) and say in plain words what each one does.
- You will explain how a human gene can be put into bacteria to make a product such as insulin, and give examples of genetic engineering and synthetic biology in medicine, agriculture and research.
- You will evaluate a claim about a genetic engineering application using its benefits, its risks, ethical questions (safety, equity and the limits of human intervention) and how reliable its sources are.

After the first film, build the change yourself, test it, and ask what success shows.

**The one idea.**

Engineering changes a cell’s DNA; test the effects, judge each use separately.

**Scope:**

This rule covers the engineered cells and organisms in this lesson. Some tools, such as PCR and gels, change nothing in a cell; they test the change. A DNA change may alter what a cell makes or how it behaves, but the intended effect is not automatic.

What do you think? Decide before you read.

1. Genetic engineering is always harmful or unnatural. Agree Disagree **Scientists disagree:** Scientists and health agencies judge each product on its own evidence. The World Health Organization asks for case-by-case safety checks. Moving genes between species also happens in nature: researchers found bacterial DNA that entered sweet potatoes naturally, with no lab involved. Whether people should make a given change is a separate question of values. What scientists think **Scientists disagree:** Scientists and health agencies judge each product on its own evidence. The World Health Organization asks for case-by-case safety checks. Moving genes between species also happens in nature: researchers found bacterial DNA that entered sweet potatoes naturally, with no lab involved. Whether people should make a given change is a separate question of values.
2. Genetically modified foods contain genes, but ordinary foods do not. Agree Disagree **Scientists disagree:** Foods made of plant or animal cells, such as fruit, grains and meat, contain DNA, so they contain genes. Only highly purified ingredients, such as refined sugar, can have little or none. Scientists can extract DNA from a strawberry or a banana in a classroom. Engineering adds or changes a gene in an organism that already has thousands of them. What scientists think **Scientists disagree:** Foods made of plant or animal cells, such as fruit, grains and meat, contain DNA, so they contain genes. Only highly purified ingredients, such as refined sugar, can have little or none. Scientists can extract DNA from a strawberry or a banana in a classroom. Engineering adds or changes a gene in an organism that already has thousands of them.
3. CRISPR can change any trait instantly in a whole adult person. Agree Disagree **Scientists disagree:** The first approved CRISPR treatment edits one kind of cell, a patient’s own blood stem cells, in a lab. The cells go back into the body after chemotherapy, and doctors follow the results for months. Editing can also miss its target or change only some of the cells. What scientists think **Scientists disagree:** The first approved CRISPR treatment edits one kind of cell, a patient’s own blood stem cells, in a lab. The cells go back into the body after chemotherapy, and doctors follow the results for months. Editing can also miss its target or change only some of the cells.

## Explore: build a bacterium that makes a product

*Interactive on the lesson page:* Gene-to-product builder

**Try this.** Test Predictions 1 and 3 below. Then compare joined and empty plasmids on the gel, predicting first whether a short or a long piece travels farther. Finally, skip selection and check how many bacteria carry the gene.

## Build one change, then test it

- genetic engineering
- restriction enzyme
- DNA ligase
- plasmid
- recombinant DNA
- polymerase chain reaction (PCR)
- gel electrophoresis
- CRISPR
- synthetic biology
- genetically modified organism

**Genetic engineering** is using laboratory tools to change an organism’s DNA. DNA has the same basic structure in living things, and nearly all of them share one genetic code. So a bacterium can read intron-free DNA for a human protein when bacterial control DNA is present. Intron-free matters: most human genes contain introns, stretches that human cells cut out of the RNA message. Bacteria cannot cut them out, so they generally cannot read a natural human gene correctly. Labs instead use DNA made chemically or copied from the RNA message. Review how DNA is built in [that lesson](https://stemhop.com/lessons/dna-structure-and-copying/) and how cells read a gene in [the protein lesson](https://stemhop.com/lessons/protein-synthesis/).

A **restriction enzyme** is a protein from bacteria that cuts DNA wherever it finds one particular short sequence; many leave short single-stranded tails called sticky ends. A **plasmid** is a small, usually circular piece of bacterial DNA that copies separately from the main chromosome. **DNA ligase** is an enzyme that seals paired ends into one continuous strand, making **recombinant DNA**: DNA joined from more than one source.

**Predict 1.** In the builder, you cut the plasmid and gene with enzymes that leave ends that do not match. Will adding ligase make the intended recombinant plasmid?

> **Check your prediction.** Not in this sketch. The mismatched ends do not pair. Some real pairs of different enzymes do leave ends that can pair, but this first model keeps to matching ends.

The tools work like a word processor’s cut, paste and copy, with one difference: they are molecules acting by chance in a tube, so each step works on only some molecules or cells. As the first film showed, the plasmid usually also carries an antibiotic-resistance gene, so growing the bacteria with that antibiotic, called **selection**, leaves mostly the ones that took up a plasmid.

Now test the DNA. The **polymerase chain reaction (PCR)** uses short DNA pieces called primers, which mark where copying starts, and a DNA-copying enzyme to make many copies of one chosen DNA stretch. In the ideal model, the amount doubles each cycle.

**Predict 2.** Start with one copy. How many copies would the ideal model give after ten cycles?

> **Check your prediction.** Copies = 1 × 210 = 1,024. Real reactions copy a bit less than perfectly.

> ideal PCR: copies after n cycles = starting copies × 2 × 2 × … (n times)

**Gel electrophoresis** sorts DNA pieces by size. DNA has a negative electric charge, so current pulls it toward the positive end of a gel. Shorter pieces slip through the gel faster and travel farther. Comparing a band with a ladder of known sizes estimates the piece’s size, but not its DNA sequence.

**Predict 3.** A ladder has bands at 500 and 1,000 base pairs. Where should a piece about 700 base pairs long stop?

> **Check your prediction.** Between the 500 and 1,000 base-pair bands. That size fits the target, but sequencing would confirm that it is the intended DNA.

*Interactive on the lesson page:* Tap to enlarge negative endwellsladderlong pieceshort piecepositive end blankwellsladderblankblankblank Word bank: long 

## From a human gene to insulin

The first film’s chain, cut, join, carry, select and read, is a simplified textbook model, not a factory recipe. It skipped many details; two matter here. Many labs cut with a pair of enzymes, so the gene can enter in only one direction. And the 1979 experiment behind this model made insulin’s two chains separately (see Going deeper).

You now have tools that make a change and tools that test DNA. But which result shows which step worked?

**Film transcript: How Do We Know It Worked?**

### Film transcript: How Do We Know It Worked?

The first film built a blue plasmid carrying a green insulin gene. Now how do we know that change worked? A surviving bacterial colony is evidence, but evidence for which claim? Follow the same plasmid onto a testing bench.

One engineered bacterium can support several different claims. It may have taken up a plasmid. That plasmid may contain the intended insulin DNA. The cell may also read the gene and make insulin protein. Each claim needs its own test. A check mark belongs to a claim, not to the whole organism. Keep the blue plasmid and green product separate in your answer.

Predict before opening the first frame. A colony grows on antibiotic because the plasmid carries resistance. Does that growth alone prove the bacterium is making insulin? No. Antibiotic growth supports plasmid uptake because resistance travels on that plasmid. The insulin insert might still be missing, reversed, or unread.

Selection narrows the cells worth testing. It does not replace the tests. Next test the DNA. The polymerase chain reaction copies one chosen stretch. A gel sorts DNA pieces by size, with shorter pieces traveling farther. An expected band supports the insert's presence and size. Sequencing can confirm its identity. A DNA band still does not show insulin protein. Use a protein test, including an empty-plasmid control and a known insulin sample. A signal in the engineered group, with controls behaving correctly, supports production. Purity and function need further tests. The same logic follows other uses. A crop, an edited blood stem cell, and a research cell each begin with a specific change. Measure the intended effect, possible unintended effects, and limits.

Do not transfer one result to every use. Consider the lesson's crop survey. Its lines rise as the share of corn and soybean acres planted with engineered varieties grows. Those data show adoption. They do not show safety, environmental effect, farmer benefit, or fairness. Those claims require other evidence. Evidence can estimate benefits and risks, but it cannot choose every value. Ask who wrote the claim, what evidence supports it, and whether other scientists checked it. Then discuss access, safety, and limits for that particular use. Reasonable people may weigh the same evidence differently. A changed DNA sequence is the start of an explanation, not the end.

Match each claim to a test, keep uncertainty visible, and judge cases one by one. Engineering changes a cell's DNA; test the effects, judge each use separately.

From the film

- Antibiotic selection supports plasmid uptake, not insulin production.
- DNA tests and protein tests answer different questions.
- Adoption data do not by themselves show safety or benefit.
- Engineering changes a cell's DNA; test the effects, judge each use separately.

## From one product to many uses

The second film left out a newer way to aim the change itself. **CRISPR** comes from a bacterial defense system. In a common version, a guide RNA helps aim Cas9 at a matching DNA region, and Cas9 cuts both strands. A cell then repairs the cut. The final change must be tested, because unintended sites may also be changed.

**Tool kit:**

restriction enzymes cut, ligase joins, plasmids carry and copy, PCR copies one stretch, gels sort by size, and CRISPR cuts at one chosen sequence.

*Figure (AI-generated illustration).* The first approved CRISPR treatment edits one kind of cell, a patient’s own blood stem cells, outside the body and returns them after chemotherapy. Image: four steps from left to right. A patient's blood stem cells are taken out into a bag, edited in a dish in a lab, the patient gets chemotherapy from a drip, and the edited cells are put back. An inset shows a protein block holding a short guide strand along a DNA ladder, with a cut across both sides of the ladder.

**Predict 4.** Should Casgevy’s edit change every cell and every trait in the patient?

> **Check your prediction.** No. Only one kind of cell is edited and returned after chemotherapy, and doctors follow results for months. Sperm, eggs and skin are not edited, and nothing changes overnight.

A bacterium or crop engineered to carry a DNA change is a **genetically modified organism**. Casgevy is different: it modifies a patient’s cells outside the body, not a whole organism. Uses come from several fields:

| Field | Example | What was changed |
| --- | --- | --- |
| Medicine | Human insulin; human growth hormone | Bacteria or yeast carry a human gene |
| Medicine | Casgevy, approved in 2023 for sickle cell disease | A patient’s own blood stem cells are edited with CRISPR |
| Agriculture | Bt corn and cotton | A bacterial gene makes a protein that kills certain insect pests |
| Agriculture | Herbicide-tolerant soybeans | Plants survive certain weed-killers |
| Industry | Cheese-making enzyme chymosin | Microbes carry the gene for an enzyme once taken from calf stomachs |
| Research | Edited cells grown in a lab | Scientists switch off one gene to learn its job |

**Synthetic biology** goes a step further: it designs and builds new biological parts, devices and systems. For example, researchers built a new chemical pathway in baker’s yeast to make a malaria medicine’s starting material.

Each use raises questions that data alone cannot settle.

- **Safety:** Could an edit land in the wrong place? Could a crop affect insects it was not meant for? Can relying on one weed-killer select for resistant weeds, as in that lesson?
- **Equity:** Who can pay for a new treatment? Who can buy the seed?
- **Limits:** Should anyone edit sperm, eggs or embryos, when future people cannot agree to the change?

To the second film’s source questions (who wrote it, what evidence, checked by other scientists?), add one more: does the writer gain from your belief? In short: change, test, judge.

## Elaborate: how common are engineered crops?

Each June, USDA asks farmers what they planted. This table shows the share of planted acres with any genetically engineered variety.

| Year | Corn, engineered varieties (% of U.S. planted acres) | Soybeans, engineered varieties (% of U.S. planted acres) |
| --- | --- | --- |
| 2000 | 25 | 54 |
| 2004 | 47 | 85 |
| 2008 | 80 | 92 |
| 2010 | 86 | 93 |
| 2016 | 92 | 94 |
| 2020 | 92 | 94 |
| 2024 | 94 | 96 |
| 2025 | 94 | 96 |

Data are U.S. estimates from the USDA June Agricultural Survey, published by the USDA Economic Research Service (2000 to 2024 and 2025 tables). Virginia is not listed separately. NASS estimated that Virginia farmers planted about 630,000 acres of soybeans and 500,000 acres of corn in 2024.

**Predict 5.** For corn, was the rise larger from 2000 to 2010 or from 2010 to 2025?

> **Check your prediction.** It was much larger in the first interval: 86 minus 25 is 61 percentage points, while 94 minus 86 is 8 percentage points.

1. Which crop reached 90 percent first, and in which year shown?
2. These are national estimates. What would you need before using them to describe Virginia farms?

Claim, evidence, reasoning

Make a claim that answers the essential question, using this data: what do the data show about how people have used crop engineering, and what would you still need to decide whether its use is wise?

Claim Your answer to the question, in one sentence.

Start with: The data show that engineered crops blank, but deciding whether that is wise also needs blank.

Evidence The data that support your claim.

Start with: From 2000 to blank, the share of U.S. corn acres with engineered varieties rose from blank to blank percent, and soybeans rose from blank to blank percent.

Reasoning Why the evidence supports the claim, using the science.

Start with: Farmers chose this seed because blank, but a share of acres cannot show blank, so we also need evidence about blank.

> **Compare with a model answer.** People already use crop engineering on most U.S. corn and soybean acres. From 2000 to 2025, engineered corn rose from 25 to 94 percent of planted acres, and soybeans from 54 to 96 percent. Farmers chose this seed, likely for insect protection and easier weed control. The data cannot tell us whether those crops are safe, good for the environment or fair to small farms. They are survey estimates for the whole country, not Virginia, and they count acres, not harm or benefit. To decide, we would need safety studies on each crop, data on resistant weeds and insects, and a discussion of costs.

Saved in this browser only.

### Design problem

You put an insulin plasmid into bacteria. Plan how to show that your bacteria took up the plasmid and are making insulin, and which controls you need.

> **One way to do it.** - **Groups:** bacteria given the insulin plasmid; bacteria given the same plasmid with no insulin DNA (the control); bacteria given nothing.
> - **Selection:** grow all three on antibiotic plates. Bacteria given nothing should not grow, showing the antibiotic works. On a plate without antibiotic they should grow, showing the cells were alive.
> - **Check the DNA:** extract the plasmid from grown colonies, cut on both sides of the insert and run a gel. The insulin plasmid should give a backbone band plus an insert-sized band that the empty plasmid lacks.
> - **Check the product:** test each group for insulin with an antibody test. The empty-plasmid group should give no signal, so a signal in the insulin group points to the added DNA. A known insulin sample shows the test works. A signal shows insulin is present; purity and activity need further tests.
> - **Constants:** same strain, temperature and growth time. Give the two plasmid groups equal molar amounts of plasmid (the same number of plasmid molecules), and give the no-DNA group the same volume of plasmid buffer without DNA.
> - **Repeats:** test at least three colonies from each group that grows, on three different days.
> - **Safety:** use a harmless lab strain, wear gloves and destroy all cultures afterwards.

## Going deeper

From the 1920s on, people who needed insulin usually used insulin from cattle and pig pancreases. The FDA approved human insulin made by bacteria in 1982. In the 1979 experiment behind the simple model, scientists chemically made DNA for insulin’s two chains. Bacteria made each chain attached to a bacterial protein; the chains were cut free, purified and joined. Today bacteria or yeast usually make one precursor chain that is trimmed into insulin.

## Evaluate: check your thinking

Bacteria make human insulin because scientists joined insulin DNA into a plasmid with restriction enzymes and ligase. The bacteria copy that plasmid as they divide and read the gene to make the protein. Related tools now change crops, and CRISPR now edits blood stem cells.

**Science practices you used:** predicting and testing with a model, reading a data table and calculating a change, planning controls, judging the reliability of a source, and weighing benefits, risks and ethics.

---

From STEMhop (https://stemhop.com), free lessons by Christopher Deppmann. Lesson text © the authors. Practice questions: https://stemhop.com/lessons/genetic-engineering/
