Upload
andrew-j
View
215
Download
0
Embed Size (px)
Citation preview
The interplay between the humanimmunodeÞciency virus (HIV)and the immune system turns
out to be signiÞcantly more dynamicthan most scientists would have sus-pected. Recent research indicates thatHIV replicates prodigiously and destroysmany cells of the immune system eachday. But this growth is met, usually formany years, by a vigorous defensive re-sponse that blocks the virus from mul-tiplying out of control. Commonly, how-ever, the balance of power eventuallyshifts so that HIV gains the upper handand causes the severe immune impair-ment that deÞnes full-blown AIDS.
We have put forward an evolutionaryhypothesis that can explain the ultimateescape of the virus from immune con-trol, the typically long delay between in-fection and the onset of AIDS, and thefact that the extent of this delay can varyconsiderably from patient to patient.Most infected individuals advance toAIDS over the course of 10 years or so,but some patients are diagnosed withintwo years of infection, and others avoidAIDS for 15 years or more.
We argue that the powerful immuneresponse enabling many patients to re-main healthy for years is Þnally under-mined by continuous mutation of thevirus. As will be seen, within any givenindividual, new viral variants mayemerge that are able to evade the pro-tective forces somewhat. In our view,the accumulation of many such vari-ants can muddle the immune systemto the point that it can no longer Þghtthe virus eÝectively.
To understand how we came to thishypothesis, which is gaining clinicalsupport, it helps to know a bit abouthow the immune system eradicates vi-ruses in general and how it respondsto HIV in particular. When any virus en-ters the body and colonizes cells, de-fensive forces launch a multipronged
but highly targeted attack. Macrophag-es and related cells engulf some of thefree particles and break them up. Thenthe cells Þt certain protein fragments,or peptides, into grooves on proteinsknown as human leukocyte antigens(HLAs). The cells subsequently displaythe resulting complexes on their sur-face for perusal by the white blood cellscalled helper T lymphocytes.
The Body Fights Back
Each helper cell bears receptors ableto recognize a single displayed pep-
tide, or epitope. If it encounters the rightepitope on a macrophage or similar cell,it binds to the peptide, divides and se-cretes small proteins. The proteins helpto activate and promote replication ofstill other components of the immunesystemÑnotably cytotoxic, or killer, Tlymphocytes and B lymphocytes.
Under the right circumstances, thekiller T cells directly attack infectedcells. Like macrophages, infected cellsbreak up some viral particles, combinecertain of the fragments with HLA mol-ecules and exhibit the complexes onthe cell surface. If a cytotoxic T lympho-cyte, through its receptors, recognizesone of the epitopes on a diseased cell,it will bind to the epitope and destroythe cell before more viral particles canbe generated. Activated B lymphocytessecrete antibodies that recognize spe-ciÞc peptides on the viral surface. Theantibodies mark free viral particles,those not yet sequestered in cells, fordestruction.
All these responses are believed toparticipate in the defense against HIV.In the initial stage of HIV infection, thevirus colonizes helper T cells and mac-rophages. It also replicates uncheckedfor a while. As the amount of virussoars, the number of helper cells falls;macrophages die as well, but the eÝects
on them have been less studied. Theinfected T cells perish as thousands ofnew viral particles erupt from the cellmembrane. Soon, though, cytotoxic T
and B lymphocytes mount a strong de-fense and kill many virus-infected cellsand viral particles. These eÝects limitviral growth and give the body an op-portunity to restore temporarily itssupply of helper cells to almost normalconcentrations. Nevertheless, the viruspersists. In the early phase, which maylast for a few weeks, about 30 percentof infected patients display some symp-toms, often a fever that may be accom-panied by a rash and swollen lymphglands. Even those individuals, though,usually go on to enter a prolongedsymptom-free stage.
Throughout this second phase theimmune system continues to functionwell, and the net concentration of mea-surable virus remains relatively low.Nevertheless, the viral level rises grad-ually, in parallel with a decline in thehelper population. Accumulating evi-dence indicates that helper cells are lostbecause the virus and cytotoxic T cellsdestroy them, not because the bodyÕsability to produce new helper cells be-comes impaired. It is a sad irony thatthe killer cells required to control HIVinfection also damage the helper T cellsthey need to function eÛciently.
Patients are generally said to crossthe line to AIDS when the helper cellcount, which in healthy individuals mea-sures 1,000 cells per microliter of blood,falls below 200. During this stage, theviral level climbs sharply, and measuresof immune activity drop toward zero. Itis the loss of immune competence thatenables normally benign microorgan-isms (particularly protozoa and fungi)to cause life-threatening diseases inAIDS patients. Once AIDS develops, peo-ple rarely survive for more than twoyears.
58 SCIENTIFIC AMERICAN August 1995
How HIV Defeats the Immune System
A plausible hypothesis suggests the immune devastation that underlies AIDS stems from continuous—and dangerous—evolution of the human immunodeficiency virus in the body
by Martin A. Nowak and Andrew J. McMichael
LEN
NA
RT
NIL
SS
ON
Boe
hrin
ger
Inge
lhei
m In
tern
atio
nal
Copyright 1995 Scientific American, Inc.
Persistence of a good immune re-sponse in the face of constant attack byHIV raises the issue of why the immunesystem is unable to eradicate HIV com-pletely in most, if not all, cases. Severalyears ago various features of HIV ledone of us (Nowak) and his colleagues inthe zoology department of the Univer-sity of Oxford to suspect the answerslay with an ability of the virus to evolvein the human body.
Evolutionary Theory Predicts Trouble
Evolutionary theory holds that chancemutation in the genetic material of
an individual organism sometimesyields a trait that gives the organism a
survival advantage. That is, the aÝectedindividual is better able than its peersto overcome obstacles to survival andis also better able to reproduce prolifi-cally. As time goes by, oÝspring thatshare the same trait become most abun-dant in the population, outcompetingother membersÑat least until anoth-er individual acquires a more adaptivetrait or until environmental conditionschange in a way that favors diÝerentcharacteristics. The pressures exertedby the environment, then, determinewhich traits are selected for spread in apopulation.
When Nowak and his co-workers con-sidered HIVÕs life cycle, it seemed evi-dent that the microbe was particularly
well suited to evolve away from anypressures it confronted (namely, thoseexerted by the hostÕs immune system).For example, its genetic makeup chang-es constantly; a high mutation rate in-creases the probability that some genet-ic change will give rise to an advanta-geous trait. This great genetic variabilitystems from a property of the viral en-zyme reverse transcriptase. In a cell,HIV uses reverse transcriptase to copyits RNA genome into double-strandDNA. This DNA is inserted into a chro-mosome of the host, where it directsthe production of more viral RNA andviral proteins. These elements, in turn,assemble themselves into viral particlesthat can escape from the cell. The virus
SCIENTIFIC AMERICAN August 1995 59
PARTICLES OF HIV (blue spheres ), the virus that causes AIDS,bud from an infected white blood cell before moving on to
infect other cells. The immune system controls such spreadat Þrst but is eventually outmaneuvered by the virus.
Copyright 1995 Scientific American, Inc.
mutates readily during this process be-cause reverse transcriptase is rather er-ror prone. It has been estimated thateach time the enzyme copies RNA intoDNA, the new DNA on average diÝersfrom that of the previous generation inone site. This pattern makes HIV themost variable virus known.
HIVÕs high replication rate further in-creases the odds that a mutation use-ful to the virus will arise. To appreciatethe extent of HIV multiplication, consid-er Þndings released early this year fromteams headed by George M. Shaw ofthe University of Alabama at Birming-ham and by David D. Ho of the AaronDiamond AIDS Research Center in NewYork City. The groups reported that atleast a billion new viral particles areproduced in an infected patient eachday. They found that in the absence ofimmune activity, the viral populationwould on average double every twodays. Such numbers imply that viral par-ticles present in the body 10 years af-ter infection are several thousand gener-ations removed from the original virus.In 10 years, then, the virus can under-
go as much genetic change as humansmight experience in the course of mil-lions of years.
A Scenario of Disease Progression
With knowledge of HIVÕs great evo-lutionary potential in mind, Nowak
and his colleagues conceived a scenariothey thought could explain how thevirus resists complete eradication andthus causes AIDS, usually after a longtime span. Their proposal assumed thatconstant mutation in viral genes wouldlead to continuous production of viralvariants able to evade to some extentthe immune defenses operating at anygiven time. Those variants wouldemerge when genetic mutations led tochanges in the structure of viral pep-tidesÑthat is, epitopesÑrecognized bythe immune system. Frequently suchchanges exert no eÝect on immune ac-tivities, but sometimes they can causea peptide to become invisible to thebodyÕs defenses. The aÝected viral par-ticles, bearing fewer recognizable epi-topes, would then become more diÛ-
cult for the immune system to detect.The hypothesis proposed that a mu-
tation able to reduce recognition of anepitope would give a viral variant a sur-vival advantage, at least until the im-mune system discovered and reactedto the altered peptide. This responsewould reduce the viral load for a time,but meanwhile other Òescape mutantsÓwould begin to break out, and the cyclewould continue, preventing full elimi-nation of the infection.
Such a scheme is extremely hard toverify with clinical tests alone, largelybecause the nonlinear interactions be-tween the virus and the immune sys-tem are impossible to monitor in detail.Consequently, Nowak turned to a com-puter simulation in which an initiallyhomogeneous viral population evolvedin response to immunologic pressure.He reasoned that if the mathematicalmodel produced the known patterns ofHIV progression, he could conclude theevolutionary scenario had some merit.
The equations that formed the heartof the model reßected features that No-wak and his colleagues thought were
60 SCIENTIFIC AMERICAN August 1995 Copyright 1995 Scientific American, Inc.
important in the progression of HIV in-fection: the virus impairs immune func-tion mainly by causing the death ofhelper T cells, and higher levels of virusresult in more T cell death. Also, the vi-rus continuously produces escape mu-tants that avoid to some degree thecurrent immunologic attack, and thesemutants spread in the viral population.After a while, the immune system Þndsthe mutants eÛciently, causing theirpopulations to shrink. The model addi-tionally distinguished between twokinds of immune responses: those rec-
ognizing epitopes that undergo muta-tion readily and those recognizing con-served epitopes (ones that appear in anunchanging form on every viral particlein the body, because the virus cannottolerate their loss or alteration).
The simulation managed to reproduce
the typically long delay between infec-tion by HIV and the eventual sharp risein viral levels in the body. It also pro-vided an explanation for why the cycleof escape and repression does not goon indeÞnitely but culminates in un-controlled viral replication, the almost
SCIENTIFIC AMERICAN August 1995 61
COURSE OF HIV INFECTION typicallyruns many years, during most of whichthe patient has no symptoms. Striking-ly, the bodyÕs defensesÑas indicatedby levels of antibodies, killer T cellsand helper T cells in the bloodÑremainstrong throughout much of the asymp-tomatic period, eradicating almost asmuch virus as is produced. At somepoint, however, the immune defenseslose control of the virus, which repli-cates wildly and leads to collapse of theimmune system.
DIM
ITR
Y S
CH
IDLO
VS
KY
DIM
ITR
Y S
CH
IDLO
VS
KY
Copyright 1995 Scientific American, Inc.
complete loss of the helper T cell pop-ulation and the onset of AIDS.
In particular, the model indicated thatthe immune system can often mount astrong defense against several viral vari-ants simultaneously. Yet there comes apoint, usually after many years, whenthere are too many HIV variants. Whenthat threshold is crossed, the immunesystem becomes incapable of control-ling the virus. This Òdiversity thresh-old,Ó as we call the breaking point, can
diÝer from person to person. For in-stance, if the immune system is rela-tively weak from the start, a few vari-ants may be suÛcient to overcome thebodyÕs defenses.
There is an intuitive explanation forwhy the presence of multiple HIV vari-ants in an individual can impair theeÛciency of the immune system. Thisexplanation considers the battle be-tween HIV and the bodyÕs defensiveforces to be a clash between two armies.Each member of the HIV army is a gen-eralist, able to attack any enemy cell itencounters. But each member of theimmune army is a specialist, able torecognize an HIV soldier only if the sol-dier is waving a ßag of a precise color.
Suppose the armies would be equallypowerful if every specialist in the im-mune army recognized the same ßagand every HIV soldier carried that ßag.Now suppose that the HIV army con-sisted of three groups, each carrying adiÝerent ßag and that, in response, theimmune specialists also divided intothree groups, each recognizing a sepa-rate ßag. Under these conditions, theimmune army would be at a signiÞcantdisadvantage. Any given immune spe-cialist would recognize and attack onlyone out of every three enemy soldiersit encounteredÑthe one carrying theright ßag. The HIV soldiers, meanwhile,would continue to pick oÝ every spe-cialist they met and would ultimatelywin the war.
Predicting the Course of Disease
Beyond giving us the concept of a di-versity threshold, the model oÝered
a possible explanation for why somepatients progress to AIDS more quicklythan do others. If the initial immune re-sponse to conserved epitopes is strong,the eÛciency of the defensive attack onHIV will not be undermined very muchby mutation in other epitopes. (Manyactive members of the immune system
will continue to recognize every infect-ed cell or viral particle they encounter.)Hence, the body should control thevirus indefinitely, in spite of quite highlevels of viral diversity. In such individ-uals, progression to AIDS is likely to beslow (or may not happen at all ).
If the immune response to conservedepitopes is not strong enough to con-trol the viral population on its own, butthe combined eÝort of the responsesagainst conserved and variable epitopescan initially manage the virus, the de-fensive forces could do well for quite awhile. But the reaction against variableepitopes should eventually be under-mined by the emergence of escape mu-tants and increasing viral diversity. Inthis case, HIV levels should rise as theresponse to variable epitopes becomesless eÛcient. This is the pattern thatapparently occurs in most patients.
If the combined immune responsesto conserved and variant epitopes aretoo weak to control HIV replication fromthe start, AIDS should develop rapidly.In that situation, the original viral parti-cles would proliferate without encoun-tering much resistance, and so the viruswould be under little pressure to gen-erate mutants able to escape immunereconnaissance. Such patients mightprogress to AIDS even in the absenceof signiÞcant viral diversity.
The simulation also provided insightinto probable properties of the viralpopulation during each stage of HIVdisease. In the earliest days, before theimmune system is greatly activated, theviral variants that replicate fastest willbecome most abundant. Hence, even ifa patient were infected by several vari-ants at once, after a short time most ofthe virus in the body would probablyderive from the fastest-growing version.And so we expect little genetic diversi-ty during the acute phase of disease.
After the immune system becomesmore active, survival becomes morecomplicated for HIV. It is no longer
62 SCIENTIFIC AMERICAN August 1995
HIV EVADES IMMUNE CONTROL byevolving. In particular, it gives rise toÒescape mutantsÓÑvariants able to eludeimmune recognition to some extent. Ina simpliÞed example, a viral populationbearing just one recognizable epitope(green in 1 and 2 ) undergoes repeatedmutations in that epitope (3Ð5 ). The im-mune systemÑrepresented here by an-tibody-producing B lymphocytesÑcankeep pace with such maneuvers for awhile, but emergence of too many newviral variants apparently underminesthe bodyÕs ability to cope with the virus.
Copyright 1995 Scientific American, Inc.
enough to replicate freely; the virusalso has to be able to ward oÝ immuneattacks. Now is when we predict thatselection pressure will produce increas-ing diversity in epitopes recognized byimmune forces. Once the defensive sys-tem has collapsed and is no longer anobstacle to viral survival, the pressureto diversify evaporates. In patients withAIDS, then, we would again anticipateselection for the fastest-growing vari-ants and a decrease in viral diversity.
Long-term studies involving a smallnumber of patients have conÞrmedsome of the modeling predictions.These investigations, done by severalresearchersÑincluding Andrew J. LeighBrown of the University of Edinburgh,Jaap Goudsmit of the University of Am-sterdam, James I. Mullins of the Univer-sity of Washington and Steven M. Wol-insky of Northwestern University Medi-cal SchoolÑtracked the evolution of theso-called V3 segment of a protein in theouter envelope of HIV for several years.V3 is a major target for antibodies andis highly variable. As the computer sim-ulation predicted, viral samples ob-tained within a few weeks after patientsbecame infected were alike in the V3region. But during subsequent years,the region diversiÞed.
Focus on Killer Cells
The original mathematical modelstreated the immune system as a
unit and did not distinguish among theactivities of the various cell types. Be-cause killer T lymphocytes seem to ex-ert tremendous immunologic pressureagainst HIV, the two of us and our co-workers have recently designed models
that speciÞcally examine the behavior ofthose cells. These newer models taughtus even more about the way HIVÕs abilityto diversify can erode the defensivecompetence of the immune system.
We began working on these simula-tions early in 1994, after one of us (Mc-Michael) became perplexed by the re-sults of studies in which he and sever-al collaborators tracked responses ofcytotoxic T cells to HIV in initiallyasymptomatic patients. Those studiesfollowed the patients for about Þveyears and were undertaken in part toassess the inßuence of diÝerent HLAmolecules on the ability of patients tocombat the virus.
HLA molecules play a critical part inthe defensive response because theydetermine which viral peptides will bedisplayed on cells and how eÝectivelythey are showcased. Any two patientsare likely to diÝer in the precise mix ofHLA molecules they possess. In conse-quence, they will also diÝer in the pep-tide epitopes their cells exhibit and inthe ability of the HLA-peptide units toattract the attention of the immune sys-tem. Most patients infected with HIVseem to recognize just a few of themany potential epitopes generated fromthe virusÕs proteins, usually betweenone and 10.
The clinical investigations examinedthe response of cytotoxic T cells to var-ious epitopes in an internal HIV proteincalled gag. Three of the patients usedthe HLA variant B27 for such display,and two patients used HLA-B8. In theB27 patients, cytotoxic T cell responseswere directed at a single fragment of thegag protein, which underwent insigniÞ-cant variation during the course of the
study. In the B8 patients, killer T cellactivity was directed against a set ofthree other segments in gag. All threeepitopes spawned mutants during thestudy, and many of the mutant peptidesescaped recognition by the hostÕs cyto-toxic cells. It also turned out that therelative strength of the responses di-rected against the three epitopes ßuc-tuated markedly.
These studies were the Þrst to docu-ment the existence of mutant virusesable to evade killer T cells in the humanbody. Yet they also raised some puz-zling questions, especially this one: Whydid the strength of the T cell responsesto the several epitopes ßuctuate somuch? In most other viral infectionsthe responses, which are usually direct-ed against one or a few epitopes, aremuch more stable.
Why Killer Cells Go Astray
It was partly to answer this questionthat our groups collaborated on mak-
ing computer models of cytotoxic T cellresponses to HIV. The programs as-sumed that breakdown of viral particlesin infected cells would result in the dis-play of many epitopes recognized bycytotoxic T cells. The models also pre-sumed that most of the epitopes wouldbe capable of mutating and hence ofgiving rise to viral variants bearingchanges in some of their epitopes.
The models introduced random mu-tations in epitopes and then traced thegrowth of every new viral variant as wellas the abundance of cytotoxic T cells di-rected against each epitope. The abun-dance of T cells recognizing a given epi-topeÑand, hence, the killing power of
SCIENTIFIC AMERICAN August 1995 63
SPEED AT WHICH HIV LEVELS RISE (linear plots) over theyears may depend greatly on the composition of the initialimmune response (insets). Modeling suggests that if the im-mune attack directed against conserved epitopes (ones foundon every viral particle) can limit viral growth on its own (left ),the body might keep viral levels low indeÞnitelyÑeven afterthe response to readily changeable epitopes inevitably de-
cays. This pattern is uncommon. If the combined responsesare weak (center ), viral levels will rise quickly. If the com-bined responses are strong but the ÒconservedÓ responsecannot by itself control the virus (right ), the typical, fairlyslow course of viral multiplication should result. In that situ-ation, levels will begin to soar when the ability to respondeÛciently to changeable epitopes is lost.
VERY SLOW OR NO PROGRESSION TO AIDS
FAST PROGRESSION TYPICAL PROGRESSION
HIV
LE
VE
L
TIME (YEARS)
Response tochangeable
epitopes
Responseto conserved
epitopes
Rateof HIV growth
INITIAL IMMUNE RESPONSE TO HIV
DIM
ITR
Y S
CH
IDLO
VS
KY
DIM
ITR
Y S
CH
IDLO
VS
KY
Copyright 1995 Scientific American, Inc.
these populationsÑwas made to dependon the number of viral particles bearingthat epitope and on the excitatory powerof the peptide. (Some epitopes evokemore T cell replication than do others.)
The results of the multiple-epitopemodels were complex, to say the least.In essence, though, the overall eÛcacyof the immune system declined overtime, and the drop resulted from muchthe same kind of ßuctuation in immunereactivity seen in the two patients whoproduced HLA molecules of the B8 type.The ßuctuation seemed to derive froma kind of competition among killer Tcell populations.
Our calculations suggest that in thebody, one clone of killer T cells (a pop-ulation recognizing one epitope) essen-tially vies with all others for dominance.As the initial killer cell response, which
involves many clones, takes eÝect, theviral population gets smaller, therebyreducing the number of stimulatory sig-nals received by the T cells. Ultimately,only the T cell clones recognizing themost stimulatory epitopes remain ac-tive, and the T cell response may evenbe dominated by a single clone.
Such a process could be beneÞcialand could potentially eliminate a virusif the microbe did not change. On theother hand, if the epitope fueling thedominant response mutates, the corre-sponding T cell clone may not recognizethe mutant. Viral particles bearing thispeptide may then multiply virtually un-noticed. Sometimes the immune systemwill catch up with the renegade groupand mount a defense targeted againstthe new version of the epitope, but oth-er times the defensive system may
switch its attention to a diÝerent, andoriginally less stimulating, epitope. Thisswitching can be repeated many times,producing a very intricate pattern inwhich the relative abundances of T cellclones ßuctuate continuously. Emer-gence of an unrecognized form of anepitope can thus cause trouble in atleast two ways. In addition to reducingdirectly the strength of the attack onthe altered viral variant, it can inducethe immune system to shift its eÝortstoward less stimulating epitopes.
The global picture taking shape fromour recent simulations is one in whichdiversity of epitopes gives rise to ßuc-tuations of immune responses and di-version to weaker and weaker epitopes.Such diversion results in high levels ofHIV, leading to faster killing of helpercells and macrophages and to reducedcontrol of the overall viral population.Put another way, viral diversity seemsto drive disease progression. Thesemultiple-epitope simulations can be ap-plied to antibody responses as well.
Thoughts on Therapy
Someone unfamiliar with such Þnd-ings might reasonably suspect that
patients who respond to many diÝer-ent epitopes will enjoy better controlof a viral population, because a micro-bial particle not noticed by one cloneof immune cells would probably be no-ticed by another clone. Yet our modelspredict that in the case of HIV, a re-sponse to many diÝerent epitopes canbe a bad signÑan indication that im-portant epitopes may have undergoneunrecognized mutations. The simula-tions imply that patients whose immunedefenses stably recognize one or a fewepitopes probably control the virus bet-ter than those who respond to a largenumber of epitopes. This view is sup-ported by an interesting Þnding fromthe HLA study described earlier. Thetwo patients who displayed ßuctuatingT cell responses progressed towardAIDS more quickly than did patientswho had consistent responses to a sin-gle epitope. This study involved too fewpatients to allow for deÞnitive conclu-sions, however.
If the models reßect the course ofHIV disease accurately, the Þndingshave implications for the developmentof vaccines (for prevention or treat-ment) and chemical-based therapies. Inthe case of vaccines, it would probablybe counterproductive to stimulate im-mune activity against a variety of HIVepitopes in an individual. After all, suchstimulation would probably elicit an un-desirable competition among immuneforces. Rather it may be better to boost
64 SCIENTIFIC AMERICAN August 1995
COMPUTER SIMULATION tracked levels of killer T cells in a hypothetical patient.Initially (top) the T cells responded to a homogeneous population of HIV particles,each of which carried seven recognizable epitopes; epitope 5 elicited the strongestresponse (yellow). After a viral mutant carrying an altered, unrecognized versionof this epitope emerged (middle panel ), the dominant response became focused ona less stimulatory epitopeÑnumber 2 (red ). And after epitope 2 mutated (bottom),dominance shifted again, to number 4 (green), an even weaker epitope. Such shiftscould contribute to reduced immunologic control in HIV-infected patients.
KILLER T CELL RESPONSES TO UNIFORMPOPULATION OF VIRUS
RESPONSES AFTER MUTATION IN EPITOPE 5
RESPONSES AFTER MUTATION IN EPITOPE 2
RE
LAT
IVE
AB
UN
DA
NC
E O
F K
ILLE
R T
CE
LLS
(A
RB
ITR
AR
Y U
NIT
S)
EPITOPE
DOMINANTRESPONSE
0
1
1 2 3 4 5 6 7
2
3
4
8
0
1
2
3
4
8
0
1
2
3
4
8
DOMINANTRESPONSE
DOMINANTRESPONSE
DIM
ITR
Y S
CH
IDLO
VS
KY
Copyright 1995 Scientific American, Inc.
the response against a sin-gle conserved epitope, evenif that epitope is not nor-mally recognized most read-ily. This response could ide-ally evoke a persistent, con-trolling response to HIV. Thetrick, of course, would be toidentify conserved epitopesand Þnd the best way to de-liver them.
Another striking implica-tion relates to the fact thatthe virus replicates quicklyand continuously in all stag-es of infection. This realiza-tion has made many physi-cians conclude that chemi-cal agents able to halt viralreplication are probably mosteÝective when delivered ear-ly, before the virus has achance to expand too much.Combination therapies mayalso be more effective thansingle drugs, because even ifthe virus generated a mu-tant population resistant toone of the substances, theother drugs could still con-tinue to be eÝective. By re-tarding the rate of replica-tion, such strategies shouldslow the speed at which mu-tants are produced and solimit viral diversity. Ourmodels further suggest that reducingviral levels and curtailing diversity inthis way would help the natural im-mune system to contain the virus.
A Broad View of HIV Dynamics
The collected clinical and mathemat-ical Þndings show that in addition
to replicating massively in infected pa-tients, HIV mutates repeatedly and thus
spawns an enormous diversity of viralpopulations. These features enable thevirus to evolve in response to thethreats it encounters during the courseof an individual infection. Mutants ableto evade immune attack to some degreeappear and predominate until the im-mune system gathers the strength toquell themÑbut meanwhile new escapemutants begin to multiply. Power thusmoves repeatedly from the virus to the
immune system and backfor a time.
The reversals do not go onendlessly, though, apparent-ly because the evolution ofviral diversity gradually tiltsthe balance toward the vi-rus. Diversity favors the mi-crobe in part because thevariability befuddles the pa-tientÕs immune system,which becomes less eÛcientand therefore enables the vi-ral population to grow andto kill increasing numbers ofhelper cells.
Of course, killing of helpercells impairs the functioningof killer T cells and B cells,which react strongly onlywhen they are stimulated byproteins released from help-er cells. As these two celltypes become even less ef-fective, a potentially lethalspiral ensues in which virallevels rise further, more help-er T cells are killed and theoverall responsiveness of theimmune system declines.
Generation of mutantsthus stimulates a continuousreduction in the eÛciency ofthe immune system. At somepoint, the diversity becomestoo extensive for the immune
system to handle, and HIV escapescontrol completely. As the viral load in-creases, the killing of helper cells accel-erates, and the threshold to AIDS iscrossed. Finally, the immune systemcollapses. In short, it seems that anevolutionary scenario can go a longway toward explaining why HIV infec-tion usually progresses slowly but al-ways, or almost always, destroys theimmune system in the end.
SCIENTIFIC AMERICAN August 1995 65
Further Reading
ANTIGENIC DIVERSITY THRESHOLDS AND THE DEVELOPMENT OF AIDS. M. A.Nowak, R. M. Anderson, A. R. McLean, T.F.W. Wolfs, J. Goudsmit and R. M.May in Science, Vol. 254, pages 963Ð969; November 15, 1991.
HUMAN IMMUNODEFICIENCY VIRUS: GENETIC VARIATION THAT CAN ESCAPECYTOTOXIC T CELL RECOGNITION. R. E. Phillips et al. in Nature, Vol. 354, No.6353, pages 453Ð459; December 12, 1991.
HOW DOES HIV CAUSE AIDS? Robin A. Weiss in Science, Vol. 260, pages1273Ð1279; May 28, 1993.
VIRAL QUASISPECIES. Manfred Eigen in ScientiÞc American, Vol. 269, No. 1,pages 42Ð49; July 1993.
MULTIFACTORIAL NATURE OF HUMAN IMMUNODEFICIENCY VIRUS DISEASE: IM-PLICATIONS FOR THERAPY. Anthony S. Fauci in Science, Vol. 262, pages1011Ð1018; November 12, 1993.
ANTIGENIC OSCILLATIONS AND SHIFTING IMMUNODOMINANCE IN HIV-1 INFEC-TIONS. M. A. Nowak, R. M. May, R. E. Phillips, S. Rowland-Jones, D. Lalloo, S.McAdam, P. Klenerman, B. K�ppe, K. Sigmund, C.R.M. Bangham and A. J.McMichael in Nature, Vol. 375, pages 606Ð611; June 15, 1995.
MOVING IN FOR THE KILL, cytotoxic T lymphocytes attack acancer cell in much the way they ambush virus-infected cells.Many lymphocytes attach to a target cell and secrete sub-stances that drill holes into the cell membrane.
LEN
NA
RT
NIL
SS
ON
Boe
hrin
ger
Inge
lhei
m In
tern
atio
nal
The Authors
MARTIN A. NOWAK and ANDREW J. MCMICHAEL arecollaborators at the University of Oxford. Nowak is a Well-come Trust Senior Research Fellow in the department ofzoology and at Keble College. He earned his Ph.D. fromthe University of Vienna, where he studied biochemistryand mathematics. Although Nowak concentrates on the in-teractions between HIV and the immune system, he hasdeveloped a wide variety of mathematical models relatingto evolutionary biology. McMichael, who became excitedby science after reading a series of ScientiÞc American ar-ticles on DNA in the 1960s, is a Medical Research CouncilClinical Research Professor of Immunology at Oxford andhead of the Molecular Immunology Group at OxfordÕs In-stitute of Molecular Medicine. He is also a consultant toCelltech and a Fellow of the Royal Society. McMichael hasclimbed the highest mountain in Austria, Nowak the high-est mountain in England.
Copyright 1995 Scientific American, Inc.