Defending the Castle

How HIV Attacks, and How Medicines Fight Back

Most of us have heard that HIV creates complex chemical reactions to fool our white blood cells into producing new baby HIVs (that is, virions). Our white blood cells make up our immune system, which is the invisible armor that protects us from colds and other diseases. Following are some illustrations that describe how HIV takes over. (I'll put the technical description in italics, like this, after each section.)

Since our immune system's white blood cells shield us from the full effects of most pathogens, imagine that our bloodstream is filled with white fortresses, or castles, which protect us from enemies. Castles are designed to be sturdy, and our castles are just that: they can stand up against almost any disease we might come up against. But our deadliest enemy, HIV, has figured out how to beat our castles. How? Through four basic steps.

Step 1

If any enemy wanted to attack a castle, what would be his very first objective? Well, he'd have to make it across the moat (the deep water that surrounds a castle, keeping enemies from running up to the door). HIV knows how to get on a raft to get across our moats. (In other words, it fuses and attaches to our cells, using the T-cell co-receptors, such as CKR-5 and CXR-4.) Now, if an enemy makes it across the moat, has it won the war? Of course not. You've seen movies that show how armies secure their castle doors. Picture faithful soldiers pouring boiling oil onto an enemy who is trying to break down the door to a castle. Your body can also kill off HIV that is still crossing the moat. But it's a race to see whether HIV will sneak in before the troops can catch it.

Step 2

What if an enemy that makes it across the moat also has a key to the castle's door? HIV does, and it can turn the key to get in the door. (It turns its RNA into DNA in the cell, through a process called reverse transcriptase. RNA is just a piece of information; DNA is the operating code for cells.) If an enemy makes it through your castle's door, has he won? No, you still have a sturdy castle, but your castle is more vulnerable now, as the enemy can start causing havoc inside.

Step 3

Once inside the castle, a smart enemy would know that he's outnumbered, so he'd want to mess up the castle's defense system. So an enemy might sneak down to the castle's map room, and pin a phony map up on top of the one that is supposed to correctly direct the troops in the event of attack. Then he would sound the alarm, and watch all the soldiers running the wrong way. In effect, the enemy would turn the soldiers against the castle they're supposed to defend! (At this point, HIV integrates its new DNA into the cell's nucleus, using an enzyme called integrase. This step overtakes the cell's primary function, and directs the cell to start producing strips of new material to make future virions.) In other words, the enemy fools the castle into actually sending new soldiers out to battle other castles, rather than the castle's enemies.

Step 4

But the enemy's work still isn't done. Not only does the enemy want to take over this castle; it wants to take over all of the castles in the "kingdom" of your body. So it sneaks down to the weapons room, to arm itself for battle in this and other castles. It does this by cutting up pieces of "metal" to make new weapons. (Or, more exactly, HIV cuts up virion strands, using its protease enzyme. Until these strips are separated, like pieces of a model car, they can't be made into new HIV particles.) The enemy would then put on the special war gloves needed to carry these new, sharp weapons out to battle. Then he breaks out of the castle to go attack other castles fully armed. (That is, HIV packages new virions using zinc fingers, and then buds from the cell. The new weapons are in reality new HIV pieces, which break out to infect new cells.)

Medicine first worked at trying to stop HIV at the castle door, Step 2, in its overall attack plan. Medicines like AZT basically try to fake out HIV, by putting a phony keyhole on the door, so that it won't turn its key in the lock. (They set a decoy so that HIV can't turn its RNA into DNA through reverse transcriptase; this class of drugs is called nucleoside analogues.)

These drugs work fairly well, especially if two are combined. A newer strategy to holding HIV outside the door tries to gum up the lock, in order to stall HIV even if it does luck out and place its key in the right hole. (These drugs are called NNRTIs, or non-nucleoside reverse transcriptase inhibitors.) The newest defense of all at this step tries to bend HIV's key itself, making it harder for HIV to turn the RNA into DNA. (These drugs are called nucleotide analogues -- not nucleoside. Though quite potent, they are proving difficult to make in a way that is safe.)

These days, most people are talking about "the cocktail," which is a daily drug combination made up of three or more different medicines taken during the course of a day (but, unlike a real cocktail, you don't actually mix them in a glass). There are actually many cocktails, because there are many combinations that you could take. Cocktails are also called HAART (highly active anti-retrovial therapy). The earliest cocktails typically featured two nucleoside analogue RTIs and one protease inhibitor (PI). Today, some include two PIs, while others use potent NNRTIs instead of PIs. Another uses the newest nucleoside analogue, Ziagen, for a total of three RTIs.

Powerful protease inhibitors have helped many patients lower their virus below detection (though we know virus is still there, hiding out). Unfortunately, these meds can also bring the worst side effects some patients have ever experienced: kidney stones, anemia, neuropathy, nausea and vomiting, and even lipid redistribution ("protease paunch" and "buffalo hump"). In addition, PIs have a weakness of cross-resistance. The tricks that HIV pulls to outsmart one of the drugs seem to give it an advantage against other similar drugs later (HIV develops cross-resistance). Non-peptide PIs (like the coming tipranavir) may help thwart cross-resistance.

For a while, some scientists thought that cocktails might push HIV out of a person's body. We now see that this probably won't work. These days, doctors are talking about "remission," rather than "eradication." We've learned that it may not be necessary to hold the virus completely below the limits of detection. Our bodies have some ability to recover their immune systems, if the virus is brought down to fairly low (but still detectable) levels.

Scientists also continue to work on a new drug from another new class, integrase inhibitors. Unfortunately, so far, tests suggest that these drugs are difficult to make, and don't work as well as hoped.

Early in the years of our war on HIV, scientists tried to stop HIV at Step 1, before it crossed the moat. Some of the medicines looked good in test tube studies, but then failed when we tried them in live patients (the anti-oxidants in our blood absorbed the medicines). Science moved to working on protecting the door, because it was easier to develop medicines at this step.

Then three years ago, researchers were startled to discover a few individuals who were virtually immune to traditional modes of HIV transmission, despite numerous exposures to the virus. These people don't have a boat in their moat! They were born with pieces missing in their immune system. Normally, this would be a bad thing, but scientists learned that the missing part was just what HIV uses to get across the moat to the castle's door. Less than 1% of persons tested lacked a boat (or a CC-CKR-5 co-receptor).

The discovery that people could live without the boat turned scientists on to new ideas, and new ways to try to slow HIV down at its first step. Scientists want to learn how to create this "defect" in other people, so that HIV will be stopped at the moat. It will take many years to figure out how to program our bodies to lose its boats. We're working on gene therapy, which would pull the boats out of every moat surrounding every new castle your body produces (forcing the body to produce CD4 white blood cells that are "born" without the CKR-5 co-receptor). In the meantime, medicines such as T-20 (pentafuside) and T-1249 try to block HIV's fusion and attachment in slightly different ways.

Taking whatever combination of medicines works well for you can buy time. By taking your medicines correctly now, you make it more likely that your castle will still be intact when medicine finally develops new reinforcements to bring to the battle.

Stephen J. Fallon is president of Skills4 Inc. in Fort Lauderdale, Florida (


Viral load tests typically look for HIV in the bloodstream, which you might think of as being the "highways" of the body. When we can't find any HIV traveling those highways, that's a good sign. But there could still be lots of HIV hidden away in the "buildings" of the body (such as the lymph nodes, the spleen, the liver, the brain, the testes, etc.). Typically, about 98% of the HIV in a person's body resides in the tissues and organs, not in the bloodstream. The good news is that medicines that work in the bloodstream usually have some effect in the tissues as well. The bad news is that they don't always reach everywhere they need to go (for example, they may not effectively cross the blood/brain barrier). As such, patients with undetectable viral load may still shed drug resistant HIV through their sexual fluids. HIV positive persons can speak frankly with their partners about the possible risk of "undetectable, yet transmittable" HIV.

-- Stephen J. Fallon

Skipping Pills

Many patients are trying very hard to take their meds as scheduled, so that the drugs will keep working. Unfortunately, drug failure rates now reach 50% in urban populations. HIV can eventually develop resistance on its own, even if you don't skip any pills. But you don't want to help HIV win any sooner. The lesson is simple: fighting HIV requires commitment and a wise strategy.

Recent studies show that patients miss doses of their medicines frequently, either due to forgetfulness or intolerable side effects. Missing doses, though, gives HIV time to "regroup." One study found that 56% of patients miss doses, and of those who take all of their pills, only 43% take them on time. This is dangerous because HIV can begin mutating around some protease inhibitors in just an hour and a half if patients miss a dose! A new study shows that skipping 5% of your pills leads to a one-in-six chance that your meds will fail within a year. If you miss 10% of your pills, the risk doubles to a one in three chance. And if you skip 20% of your pills, your chances are three-out-of-four that HIV will break through your medicines.

Now, you might think that with so many medicines available, failing on one combination is no big deal. But remember that doctors can't just pick any two, three, or four anti-HIV medications off of the shelf. Certain drugs do not work well together, while others work only in concert with certain others. That means that there are only so many possible ways to combine the available medicines.

In a disturbing recent study, researchers used a computer model to predict what would happen if a patient did not adhere well to medicines, and was thus forced to switch to new combinations frequently. This patient "exhausted her choices of AIDS drugs fairly early in her disease." On the other hand, if you take your pills exactly as scheduled, and with the right food considerations, we've seen patients hold their virus to undetectable levels for many, many years.

-- Stephen J. Fallon