Mechanism of HIV-1 Acquired Resistance to Protease Inhibitors and Its Possible Limitations


Phillip Hsu

Vaccine Revolution

Human Biology 115B

Instructor: Robert Siegel

March 20, 1997


Protease inhibitors are the latest addition to the arsenal of drugs designed to combat the HIV virus. Preliminary studies show that combination therapy with reverse transcriptase inhibitors have resulted in decrease of viral load to undetectable levels.1 The FDA, encouraged by these early studies, approved protease inhibitors in late 1995 after an accelerated approval process.2 The popular media has portrayed protease inhibitors as the cure for HIV infection and AIDS. However, the medical community remains reservedly optimistic about protease inhibitors because studies have also shown that despite early potent antiviral effects, HIV eventually develops resistance to protease inhibitors.1 This review will attempt to explain the mechanism by which HIV gains resistance and the potential limits to HIV's mutability as elucidated by recent scientific literature.


HIV infection's progression into AIDS is caused by infection and subsequent depletion of CD4+ T-helper cells. As HIV replicates, more and more CD4+ T-helper cells are infected and destroyed. Eventually, opportunistic infections and cancers take advantage of the deteriorated immune system and causes AIDS. Therefore, anti-viral therapy have traditionally focused on ways to stop HIV proliferation and replication.

The protease enzyme is an obvious target for antiviral drug design because protease is essential for successful viral replication. HIV gag and gag-pol genes are translated into large polyproteins containing structural viral proteins (p17, p24, p2, p7, p1, and p6), reverse transcriptase, RNase, integrase, and protease.3 The individual viral proteins that comprise these polyproteins cannot function until after protease cleaves the polyprotein into its constituent single proteins. Studies have shown that HIV with induced mutations or deletions that result in the disappearance of protease activity do not give rise to infectious HIV particles.4 In order to successfully produce viable progeny, HIV protease must cleave each of the nine different cleavage sites to yield each of the ten different protein subunits. Since all ten proteins are essential to HIV activity, failure to cleave just one cleavage site will result in non-viable HIV.3

After the discovery of the essential nature of protease, research began to focus on the design of drugs to inhibit protease's cleaving ability. Scientists and pharmacologists designed molecules that could mimic the structure of the cleavage sites of the gag and gag-pol polyprotein. These drugs are today's protease inhibitors. Protease inhibitors look so much like protease that they competitively bind to protease active sites so that the protease cannot bind to the polyprotein cleavage sites. Through this mechanism, protease inhibitors prevent cleavage of polyproteins, and therefore prevent viral replication.5

Currently, the FDA has approved 3 protease inhibitors, saquinavir, indinavir, and ritonavir. The chemical structures of these drugs are illustrated below.


Clinical Efficacy and Resistance Development

Although all three currently licensed protease inhibitors decreased viral load and increased CD4+ T-cell counts substantially in the short run, HIV resistant strains eventually develop against each protease inhibitor in the long run. Fifty percent of patients taking 3 doses of 600 mg of saquinavir per day as monotherapy developed resistant strains by the end of the first year of treatment. Higher doses of saquinavir have resulted in only similar to slightly lower incidence of resistance development.6 Therefore, although protease inhibitors as monotherapy offer substantial improvement in AIDS progression markers, namely CD4+ T-cell count and viral load, in the short run, this level of effectiveness is not sustained in the long run due to viral resistance development.

Sequential treatment with different protease inhibitors have been moderately successful, but cross-resistance is a big problem. After 40-52 weeks of treatment with indinavir, all viruses isolated from four patients were cross-resistant to the other two currently available protease inhibitors. Ritonavir treatment has also been shown to induce cross-resistance with indinavir but no cross resistance with saquinavir. This implies that the duration of effective treatment can be increased if certain protease inhibitors are administered in sequence to avoid cross-resistance development. However, after the sequence of available protease inhibitors are exhausted by HIV resistance development, there are no other effective therapy options available, and the patient's CD4+ and viral load levels return to pre-therapy levels.6

Combination therapy with AZT has also been moderately successful in delaying onset of resistance. Saquinavir and AZT together reduced the frequency of AZT resistance at one year after initiation of treatment. Saquinavir, AZT, and zalcitabine (another reverse transcriptase inhibitor) delay the onset of resistance even further if they are taken together in triple combination therapy. HIV, however, will still eventually develop resistance, and patients will return to their normal progression to AIDS.

HIV Rapid Resistance: Reasons

HIV can rapidly adapt to new selective pressures such as protease inhibitors because of its massive viral replication rate and error-prone reverse transcriptase. Studies have shown that mutations occur at a rate of 3.4x10-5 mutations/base pair/cycle.7 Mathematical models predict that mutations can occur up to 10,000 times per day in an infected individual.8 The rapid rate of reproduction and high mutation rate result in great genotype diversity among a typical population of HIV viruses inhabiting any single patient. Studies show that differences between viral strains inside any single patient can differ by as much as 25% in their genome. HIV, in addition, seems to tolerate much variability in gene sequence. One study sequenced 167 viral isolates from 102 protease inhibitor-naïve patients and found that the DNA sequences of USA HIV-1 clade B protease had 49.5% of its 99 amino acids and 41% of its nucleotides variable. This is higher than any other clade in the world combined (40%).10 Mutations at critical sites can affect virulence, replication, competence, cytotoxicity, and response to anti-viral therapy.6 While most mutations result in defective, non-viable HIV, the high-replication rate allows HIV to overcome this to develop protease resistant strains rapidly.

Testimony to the great mutability of HIV is recent research that have found protease inhibitor resistant strains of HIV in patients who have never received protease inhibitor therapy before. In one study, an analysis of 246 protease sequences from twelve subjects who were protease inhibitor-naïve found amino acid substitutions that confer resistance to protease inhibitors.9 The samples used in the study were all collected between 1991 and 1993. Protease inhibitors were not available at that time, so it would have been very unlikely that these subjects received protease inhibitor treatment prior to the sequence analysis study. The presence of protease inhibitor resistant HIV prior to initiation of HIV treatment in patients may explain why HIV develops resistance so quickly. Instead of having to develop resistant strains through mutations after the onset of protease inhibitor selective pressures, HIV resistant viruses are already present in the viral population prior to the onset of therapy. Since the resistant genotype is already present in the population, resistant HIV can quickly become the dominant quasispecies after the onset of protease inhibitor selective pressure. For instance, nevirapine, a non-nucleoside reverse transcriptase, therapy is characterized by complete replacement of wild-type virus with resistant mutant strains after just 14 days of therapy.6

Types of Mutations

Mutations that cause resistance to protease inhibitors can be classified into major mutations and minor mutations based on phenotype effects. Major mutations are frequently associated with a severalfold decrease in sensitivity to one or more protease inhibitors. Minor mutations, on the other hand, do not cause a great decrease in sensitivity to protease inhibitors. However, if a major mutation occurs in a genetic background containing minor mutations, resistance to protease inhibitors may be augmented. In other cases, cross-resistance might be conferred. For instance, minor mutations at L63P and V82T do not confer resistance by themselves, but in combination with major mutations at I84V, the virus develops resistance to 6 different protease inhibitor drugs.10 Sometimes minor mutations are also referred to as accessory or silent mutations.

Mutations can also be classified into active site mutations, non-active site mutations, and cleavage site mutations based on location. Active site mutations are the most prevalent and are defined as mutations that cluster near or on the active site of protease. Most active site mutations are also major mutations because they have direct impact on the shape and activity of the active site. Non-active site mutations are any mutations that occur on the protease genes that do not lie in the active site. Most of these mutations are secondary silent mutations. Finally, there are cleavage site mutations. Cleavage site mutations alter the cleavage sites of polyproteins so that protease can have increased accessibility to the cleavage site substrate.11

Mutations: Sequential Mutations Confer Resistance

Selection by protease inhibitors produce several distinct major mutations that occur sequentially in the protease gene.12 Usually, a single mutation appears first, followed by more single mutations. There seems to be a direct relationship between the number of mutations that accumulate and the degree of resistance conferred.

In vitro studies of saquinavir therapy shows that the first mutation that appears is G48V in the protease gene. The resultant virus is ten-fold more resistant than wild-type HIV to saquinavir. A second mutations at L90M usually follows. This double-mutated strain is 100 fold more resistant than wild-type HIV. Surprisingly, studies show that only the L90M mutation appears in in vivo studies of saquinavir. These single L90M mutants display comparable levels of resistance as the double mutant in the in vitro studies. It is speculated that there are perhaps extra selective pressures against G48V inside the human body that are not simulated in the environment of the in vitro experiment.17

Similarly, indinavir gains resistance through four step-wise mutations at M46I, L63P, V82T, and I84V, in that order. Resistance seems to develop beginning 12-24 weeks after initiation of therapy.6 In vitro passage of HIV under selective pressure of nelfinavir yielded a 30 fold reduction in sensitivity due to mutations at M46I and I84V. Again, the mutations occur sequentially.

Another study on therapy with ABT-77003, a protease inhibitor, also shows that appearance of mutations were gradual and sequential. Accumulation of mutations occur earlier in active sites and later in nonactive site positions. The first mutation appeared in HIV after five passages of HIV through CD4+ T-cells in the presence of protease inhibitor ABT-77003. The second mutation appeared after 8 passages and conferred both better infectivity and less sensitivity. The more the mutations accumulated, the more resistant the virus. However, viruses with only the initial mutation had reduced replicative capacity compared to wild-type virus in the absence of protease inhibitors. As HIV accumulated more mutations, both drug resistance and replication capacity increased compared to the HIV with only one mutation. Surprisingly, despite its limited replicative capacity, the single mutation HIV did not revert to wild-type after passage in drug-free conditions. Even in the absence of protease inhibitor selection pressure, secondary mutation accumulations continue to occur. These secondary mutations result in resistance to protease and more replicative capacity than wild-type in the absence of protease.13

This evidence suggests that once HIV makes the initial mutation to confer resistance, it is somehow forced to continue with the accumulation of more mutations. Initial mutations appear to confer resistance to the binding of protease inhibitors, but this resistance is gained at the expense of replicative ability in the form of lower enzyme kinetics. Later mutations appear to compensate for some or all of this loss in replicative ability.

Minor Mutations Affect Major Mutation Phenotypes

Before scientists knew about the existence of minor mutations, it was believed that since major mutations conferring resistance seem to decrease the replicating ability of HIV, a possible treatment for HIV infection would be to expose HIV to as many different protease inhibitors as possible. The hope was that as HIV protease accumulates more and more mutations, the HIV will become less and less replication efficient. With less replication, the onset of AIDS can be delayed.

When laboratory constructs of expected resistance mutated HIV were tested for reproductive ability, the theory seemed correct. Tests revealed that as the number of protease inhibitor resistance mutations increased, the enzyme kinetics of protease decreased. In addition, triple mutation G48V/L90M/V82A and both quadruple mutations R8Q/M46I/A71T/V82A and G48V/L90M/A71T/V82A yielded inactive HIV and corresponding protease. However, when actual wild-type resistant viruses, viruses that are found in actual patients and not from laboratory constructs, were tested using the same methods, the wild-type resistant viruses had higher replication rates than non-resistant wild-type viruses!14

This discrepancy means that there are other silent mutations in the backbone which can compensate for the loss of replicative ability induced by major mutations. These background mutations are not rare. Indeed, one study showed that seven identified minor mutations were observed as natural polymorphisms in 165 viral isolates from 102 protease inhibitor-naïve patients. These minor mutations seem to augment resistance and to compensate for the active-site mutation induced decreases in enzyme functionality. In one experiment, the minor mutations 46I/63P/82T/84V, in the presence of protease inhibitor induced major mutations at 32I/82I, conferred cross resistance to six separate protease inhibitors.15

Cleavage Site Mutations

When HIV virus is selected with high concentration substrate analog protease inhibitors BILA 1906 BS and BILA 2185, 350- to 1500- fold resistant strains of HIV evolve. These extremely resistant strains of the virus were then sequenced and analyzed. Surprisingly, these viruses contained not only mutations in the protease gene but also mutations in the Gag precursor p1/p6 or p7/p1 cleavage sites. When these mutations are removed, viral growth decreases dramatically or disappears altogether. In addition, cleavage site mutations do not decrease the ability of the protease inhibitor to inhibit the protease active site and only appear after the first active site mutations have occurred. This suggests that cleavage site mutations are not mutations selected for directly by protease inhibitors. Instead, these mutations are selected by active site mutations.16

This means that the mutations on the Gag cleavage sites are mandatory for proper protease-substrate binding. Prior mentioned studies have shown that as protease accumulates more and more mutations, the enzyme kinetics of the protease enzyme is decreased. This is probably because most mutations that confer resistance are found on or near the active site. These mutations must alter the active site in some fashion. Perhaps, in order to gain 1500 fold resistance to BILA 2185 BS and BILA 1906 BS, the active site must be altered so much that it can no longer bind to wild-type, non-mutated cleavage site substrates. According to this line of reasoning, cleavage site mutations are a rate-limiting step in HIV resistance development. Protease can mutate only so much before it cannot mutate anymore without an accompanying mutation in the substrate cleavage site.

Limits to HIV Resistance Development

Although HIV has been able to sustain large variability in its genome, there is reason to believe that HIV's ability to sustain resistance mutations has a limit.10 First, protease is an extremely small protein, made up of only 99 amino acids. Therefore, the chances that any single mutation will result in a fatal dysfunction or at least a decrease in virulence is high.14 In addition, protease is a dimer (99 amino acids on each dimer), so any mutation that occurs must occur in duplicate on each dimer. Furthermore, major mutations seem to occur in amino acids that are traditionally well conserved in HIV-1 isolates and are probably important to viral function. The evidence that major mutations decrease protease functionality seem to corroborate this.

At extremely high levels of protease inhibitor, HIV resorts to cleavage site mutations. Evidence seems to indicate that the presence of cleavage site mutations is an indication that HIV protease is running out of options for resistance mutations that are compatible with functionality. Plus, cleavage site mutations also have some important limitations themselves. There is not much flexibility near the scissile bond, and the sequence must remain hydrophobic. The p7/p1 cleavage site has one additional constraint; its p1 and all of its p' residues are encoded by nucleotide sequences involved in ribosomal frameshifting in gag-pol expression.11 Since cleavage site mutations are also severely limited, it may be possible that there is a limit to how much HIV can mutate to survive very strong protease inhibitor therapy.

Conclusions and The Future

Protease inhibitor therapy has been shown to decrease viral load substantially, but in the long run, it has not been shown to maintain its antiviral potency. However, this does not mean that protease inhibitor therapy is a lost cause. Recent research suggests that HIV gains resistance at substantial cost to protease functionality and that increased selective pressure from more protease inhibitors may lead to less virulent strains of HIV. Therefore, more funding should be given to new and more potent protease inhibitor development.

In addition, the appearance of cleavage site mutations and the possibility that these mutations might be a rate limiting step in the evolution of resistance give hope that HIV has only a limited amount of options left for resistance mutation. Research should focus on ways to inhibit the mutated cleavage sites. If cleavage site mutations are a rate limiting step in resistance development, simultaneous inhibition of cleavage site and protease could be very effective; HIV would have to mutate at both the protease and the cleavage site simultaneously to develop resistance.

Finally, more research should be done on minor (background) mutations. The discovery that different patients with different background mutations in their HIV may have very different responses to a particular protease inhibitor therapy suggests that therapy should be tailored for every particular patient based on his/her predominant HIV backbone mutations. Due to the large variety in HIV genotypes in the United States population, it may be beneficial to analyze the HIV sequence of any patient before deciding on the specific protease inhibitor(s) to administer. Careful planning of protease inhibitor use may prolong the duration of effectiveness of these drugs significantly.

These latest studies give renewed hope that HIV can be controlled as a chronic infection. It seems that as modern science is recognizing more and more options in the fight against HIV, HIV may be running out of options to evade the ingenuity of modern medicine.


  1. Mellons John W. "Closing in on human immunodeficiency virus-1." Nature Medicine, Volume 2, Number 3, March 1996, pages 274-275

  1. Deeks, Smith, Holoduiy, Kahn. "HIV-1 Protease Inhibitors." JAMA, January 8, 1997, Volume 277, pages 145-153

  1. Petit S.C., J. Simsic, D.D. Loeb, L. Everitt, C.A. Hutchison III, and R. Swanstrom. "The specificity of the HIV-1 protease." Prescription Drug Discovery and Design, 1993, Volume 1, pages 69-83

  1. Darke P.L., and J.R. Huff. "HIV protease as an inhibitor target for treatment of AIDS." Advanced Pharmacology. 1994, Volume 25, pages 399-454

  1. Noble, Faulds. "Saquinavir." Drugs. July 1996, pages 93-112

  1. Moyle, Graeme J. "Use of Viral Resistance Patterns to Antiretroviral Drugs in Optimising Selection of Drug Combinations and Sequences." Drugs. August 1996, Volume 2, pages 168-185

  1. Mansky LM, Temin HM. "Lower in vivo mutation rate of human immunodeficiency virus type 1 than predicated from fidelity of purified reverse transcriptase." Journal of Virology 1995, Volume 69, pages 5087-5094

  1. Coffin JM, "HIV population dynamics in vivo: implications for genetic variation, pathogenesis, and therapy." Science, 1995, Volume 267, pages 483-489

  1. Lech, Wang, Yang, Chee, Dorman, McCrae, Lazzeroni, Erickson, Sinsheimer, Kaplan. "In Vivo Sequence Diversity of the Protease of Human Immunodeficiency Virus Type 1: Presence of Protease Inhibitor-Resistant Variants in Untreated Subjects." Journal of Virology, March 1996, pages 2038-2043

  1. Kozal, Shah, Shen, Yang, Fucini, Merigan, Richman, Morris, Hubbell, Chee, Gingeras. "Extensive polymorphisms observed in HIV-1 clade B protease gene using high-density oligonucleotide arrays." Nature Medicine. Volume 2, Number 7, July 1996, pages 752-759

  1. Erickson, Burt. "Structural Mechanisms of HIV Drug Resistance." Annual Review of Pharmocology Toxicology. 1996, Volume 36, pages 545-571

  1. Ho D, Toyoshima T, Mo H, Kempf D, Norbeck D, Chen C, Wideburg N, Burt S, Erickson J, Singh M. "Characterization of human immunodeficiency virus type 1 variants with increased resistance to a C2-symmetric protease inhibitor." Journal of Virology, 1994, Volume 68, pages 2016-2020

  1. Borman A, Paulos S, Clavel F. "Resistance of human immunodeficiency virus type 1 to protease inhibitors: selection of resistance mutations in the presence and absence of the drug." Journal of General Virology, 1996 Volume 77, pages 419-426

  1. Rose, Gong, Greytok, Bechtold, Terry, Robinson, Alam, Colonno, Lin. "Human immunodeficiency virus type 1 viral background plays a major role in development of resistance to protease inhibitors." Proceedings of the National Academy of Sciences, 1996, Volume 93, pages 1648-1653

  1. Condra JH. "In vivo emergence of HIV-1 variants resistant to multiple protease inhibitors." Nature. 1995, Volume 374, pages 569-571

  1. Doyon, Croteau, Thibeault, Pulin, Pilote, Lamarre. "Second Locus Involved in Human Immunodeficiency Virus Type 1 Resistance to Protease Inhibitors." Journal of Virology, June 1996, pages 3763-3769

  1. Jacobsen, Hanggi, Ott, Duncan, Owen, Andreoni, Vella, Mous. "In Vivo Resistance to a Human Immunodeficiency Virus Type 1 Proteinase Inhibitor: Mutations, Kinetics, and Frequencies." Journal of Infectious Diseases, 1996, Volume 173, pages 1379-1387