NAMING THE AIDS VIRUS

Harold E. Varmus, M.D.

Departments of Microbiology and Immunology and
     Biochemistry and Biophysics

School of Medicine,

University of California

San Francisco, California 94143

(Tel. 415-476-2824)

thr

Between April 1983 and

research groups in three locales

nc as candidates for the

4

`ee thousand miles

causative agent of th syndrome (AIDS) (1-3).

Although these viruses were each called different names by the #y%mp that

&&;Naic--ilrJ

reported them, the many isolates s now appear to

a very clos ied class known as

retroviruses (4). By the end of 1984, it was apparent that both the

scientific community and the public would be better served by a single name

for the causative agent of AIDS,  but it was much less clear what that name

should be or what criteria should be applie

-& Xi< C` cC+ *

Since March, 1985, I have chaired a committee composed of thirteen well-

known retrovirologists from four countries$ attempting to find a single
acceptable name for this important new group of human retroviruses, &Se. -

~.

This duty has fallen to me

because I chair a standing subcommittee, known as the Retrovirus Study Group,

which is empowered to rule on matters of 3hal nomenclature and classification

under the aegis of a larger group known as  the International Committee on the

Taxonomy of Viruses.   We wereeto action on this occasion w-n

.c

i i-4 tt. ,- 4

i&; I

. * i`

-calls and letters from virologists close to and distant from the

experimental activity, asking -neutral group offer a solution to

't LJ- -5,1**'

*

5 *+"I *
h

what had become an intractably confusing- and at time c5e&ious issue.

The normal workings of -i-iYi are sporadic,

II"

!'

often pedantic, and of- concern to the general public, albeit useful to

those who must work with and talk about viruses. In the case of the AIDS

virus, however, our activities have attracted extraordinary attention and

curiosity.

  Why has there been so much interest in the resolution of a problem which

would appear to have no immediate relevance to the serious issues in the AIDS

crisis---controlling the spread of this virus and treating its victims?

Some perceive, wrongly in my view,  that our final recommendation will form a

verdict upon contested issues of priority of discovery,  issues that could

influence patent rights, the awarding of major prizes, patriotic sentiments,

and financial gain.  Others have seen our deliberations as a battle between

Robert Gallo and Luc Montagnier, both members of the Committee and leaders of

the two most highly publicized groups to isolate AIDS retroviruses;  this is

certainly an oversimplified view of the proceedings.

AIDS problem closely as experimental scientists, health care workers, or

public-spirited citizens, the issue may simply be one of having a single,

convenient, and generally accepted name for the virus that causes the disease.

And for an appreciable number of scientists,
eloquently expressed by Stephen Jay Gould ( r) :

For many who follow the

the task has a higher purpose,

"...taxonomies are not neutral hatracks for the pristine facts of nature.

They are theories that create and reflect the deep structure of science and

human culture.

a hypothetical statement about the nature of things."

A taxonomy is not just a ploy for convenient arrangement,  but

   In fact, the issue before us has not been taxonomy per se,  but

nomenclature;

common or species name for a virus or collection of very similar viruses,

Since such names are in daily use by the working virologist,  much ink and some

blood has been spilled over them; the forces that influence decisions about

them are as various as the forces that influence the naming of a baby, a book,

a bridge, or a city.

we have attempted to settle upon what is usually called the

Scientific principles, political realities, aesthetics, convention, and

justice must all be served in the process of finding the species name for a

virus,

I

*.

Many questions had to be considered for each suggestion4Gkmg tho

Is the proposed name consistent with scientific facts and

f' -4 .jtCL.

ed" proc Cd

P-

with- tentative cfassificZition?

name that would prevent its being used?

IS there anything objectionable about the

P

Have those credited with discovery

been accorded their rights to contribute to the process of naming?

Does the

name distinguish the virus from those it does not resemble?

Is the name

readily remembered?

Is the name likely to create confusion in the future if

and when other viruses are isolated? Does the name conform to existing

conventions for naming viruses of this general type?

Before discussing specific names proposed for the AIDS virus,  I must

briefly introduce a debate that has been raging among namers of viruses for

the past few years.

The definition of an animal species is based upon the

ability to mate productively; membership in a species is granted to those

animals that can interbreed with other members. Since viruses do not have sex

in the conventional sense,

grouping of viruses that are so closely related as

!h Xd b4 .;"PI

to be accorded a common name was often based upon c any conveniently measured

b

biological or biochemical property---such as virus shape or immune reactivity

or presence of a certain enzyme in the virus particle.

1 1

With the application of new genetic and molecular techniques to the study

of viruses, however, it has become possible to define virus groups in a

fashion analogous to that used for animals, in a way that is likely to come
closer to establishing true evolutionary relationships among viruses (7).

this new view,  it is possible to assess which viruses are so closely related

c

In

genetically that they could exchange genes without loss of viability,  in the

manner of animals that can generate viable offspring by the intermingling of

their genes.

clearly distinguished from those that are genetically disparate, regardless of

other similarities.

Such viruses properly constitute a species and can usually be

In the context of the AIDS viruses, for example, it was apparent by the

time our deliberations began that the available isolates were sufficiently

closely related to be considered members of the same virus species.

questions, in large part, were whether and how they should be distinguished I

The

from

Most

b*$# ;$*"*-c $58.  e'? st (*1

human retroviruses that appeared to belong to other species

What names did we have to consider?

prominent among these were lymphadenopathy virus (LAV), human T cell

Several names were already in use.

lymphotropic virus-I11 (HTLV-111), AIDS-associated retrovirus (ARV), two

hybrid names (HTLV-III/LAV and LAV/HTLV-111), and a name often used by the

press (the AIDS virus). In short order, many more names, about fiftyslBrffl

all permutations

were suggested by committee members,

I

consultants, and other correspondants.

How could we decide among these many candidates?

When confronted by

confusion, it is sometimes useful to

consider what has worked well in the past.

Retroviruses form a large group

of agents united by a distinctive property:

they carry their genes  in the

form of ribonucleic acid (RNA) and convert them to deoxyribonucleic acid

(DNA) after infection, reversing the usual flow of information in nature from

DNA to RNA. spite this unifying feature, retroviruses are also

manifestly diverse: they are found in many different animal hosts,  cause

several different types of disease, and frequently appear to be only          I

                                                          i
                                                                           C'
marginally related when the substance of their genes is carefully examined (4.

This complexity has generated a need to group isolates  in a rational manner

that emphasizes biological boundaries.

Y

Traditional retroviral nomenclature has worked well in this regard. The

convention has been to name viruses according to the host species of origin

and the prominent pathology associated with the prototypic isolate of a single

(I cc 'I

type; two examples of such names are feline leukemia virus and mouse ma~ary

tumor virus." Individual isolates belonging to a single species are further

distinguished by prefixes or suffixes (numbers, letters, or [in earlier times]

c 't fr

names of discoverers), as in Rauscher murine leukemia virus or human T cell

tt

leukemia virus-2.

Many of us have favored staying close to these traditional rules,  but

there were obviously other considerations as well.

In particular, the rights

of discoverers needed to be respected,  but it was clear from outset that we

could not reach a consensus about eminent domain.

It thus seemed prudent to

set as an objective the best possible name, to include all claimants in the

proceedings, and to hope that all parties would find some solution acceptable.

If we were to adopt a traditional name for the AIDS retrovirus,  we needed

to consider a particularly troubling issueaether names that included the

term "AIDS", such as human AIDS virus or human AIDS-lymphadenopathy virus,

a I( h

were likely to be psychologically damaging in the clinicah view of the

lethality of the disease and the social stigmas attached to the prevalent

a

routes of virus transmission.

clinicians who had worked with AIDS patients for their opinions.

To explore this issue, we asked over fifty

The large

number who responded showed geniune and intelligent interest in the choice of

names, but they were almost exactly divided over the issue of whether to

exclude "AIDS" from the name. Some took the position that patients would

soon be

term that denoted the agent of AIDS

as

about a term that was explicit, a phenomenon our committee

referred to as "the evanescence of euphemism." Others felt strongly that it

was easier to explain the difference between infection by the virus  and the

disease induced in a minority of infected people if the virus had a name with

less perjorative potential.

#

There was general agreement among our membership that the latter view

should prevail, if only because strict adherence to simplicity and convention

was not important enough to outweigh the misfortune of even a rare patient

suffering needlessly from the report of an antibody test.

Moreover, it was

possible to consider use of terms that denote the principal pathological

consequences of infection- - -e. g. "immunodeficiency" or "immunosuppression"- - -

thereby avoiding the name of the disease, yet conforming to standard

nomenclature.

The many names suggested for the AIDS virus raised other taxonomic

(a(. r :'I -yk.ru*s. 1

possibilities.

Some encouraged the use of host cell

a basis for

d'>name, a view already in wide circulation through the use of "human T cell

lymphotropic virus" and endorsed by other suggestions,  such as "human T4

LC*a bLIF'L+'.*

the cellular receptor that determines whether the virus  can

(This approach is discussed more fully below.)  Another

enter a cell.

possibility was raised by experimental evidence linking the AIDS virus with a
handful of animal retroviruses that produce disease slowly and are hence known
as lentiviruses, one of the three large subgroups of retroviruses ( 7 ) .

However, enthusiasm for "human lentivirus-1" was dampened by the still weakly

defined characteristics of this class and by the inappropriateness of using as

a species name a name normally applied to a collection of diverse virus

species, Others suggested defusing the political, clinical, and genetic

issues through the use of a sequential numbering system,  in which the AIDS

virus fould become human retrovirus-3.

C

However, colorless names of this type

can be difficult to connect with the right virus, especially when the list

3

becomes lengthy; the approach does not respect the genetic basis  of virus

speciation, since human retroviruses from different species would have names

that differ only by number;

and attempts to impose similar systems upon other

virus groups, such as human herpesviruses, have not always been successful.

From a pragmatic perspective, we had to consider the already widespread

use of the term human T cell lymphotropic virus-I11 (HTLV-111) and to ask

whether it was too firmly entrenched to be

tactical , however, %% issue of how

new name. This

modern virologists view the world, and how they recreate  it for their students

through the choice of names, because the term HTLV-1x1 presents a direct

challenge to the genetic basis of virus speciation.

Proponents of HTLV-I11 wish to group it with HTLV-I and -11---viruses

orginally called human T cell leukemia viruses, but proposed to be renamed as

T cell lymphotropic viruses---based upon a long list of common features (9 ).

However, the most prominent of these similarities appear superficial and

,f.

deceptive when closely examined.  Thus, though both virus types have strong

predilections for infecting the T4 subset of thymus-derived lymphocytes,  the

mechanisms determining that preference are fundamentally different for the

leukemia viruses and the AIDS virus;  the two groups of viruses use different

host cell receptors to gain entry into the lymphocytes, and,  as a consequence,

they also differ with regard to other cell types they can infect.

Second,

though the leukemia and AIDS viruses are pathogenic for man, they cause

dramatically different types of pathology, resulting on the one hand from

inappropriate proliferation of infected cells and on the other from toxicity

and death of infected cells.  Third, despite several similarities in size>of

d-

the proteins found in leukemia and AIDS virus particles, the sequences of the

building blocks (amino acids) of these proteins are now known to be as

different as the sequences for any two ostensibly unrelated retroviruses.

Finally, though both types of viruses exhibit relatively unusual mechanisms

for improving the efficiency with which they produce the proteins encoded by

their genes, on close inspection the mechanisms appear to be largely unrelated

The explanation of these fundamental differences is immediately apparent

((9
P-

when the viral chromosomes (known as genomes) are compared. Though the

genomes of the AIDS and leukemia viruses share features that are common to all

s4- as

retroviruses ---symmetrical units govern gene expression

ICL. /"

and tka basic layout of genes for the proteins found in virus particles---

wherever they can differ, they do. Most obviously, the order of the

components of the genome differs, so that the sequence of nucleotides is no

more similar for these two groups than for any two members of the diverse

family of retroyiruses.

in which the genes are juxtapoqed assigns the viruses to different classes,

and the position and nature of genes

t are entirely unrelated.

But important subtleties are also evident:  the manner

8 %.t+
h

bq7cf'

#-

r

If an evolutionary tree is established for retroviruses by comparing the

order of amino acids in the protein most characteristic of retroviruses,  the

enzyme that converts RNA to DNA, it is apparent that the AIDS virus is most

closely related to the sheep lentivirus, called visna, whereas the human T

cell leukemia viruses are in another limb of the tree, more closely related to

other oncogenic viruses,  leukemia and sarcoma viruses of various animals,
particularly the bovine leukemia virus (7). To those who believe in the

application of genetic principles to virus nomenclature, it is apparent that a

superficial resemblance in host cell tropism should not be used to impose a

species designation upon viruses that are highly divergent according to

rigorous criteria.

Through a series of memoranda and polls that sought a&consensus on these

f&L&yPr, Y,ZiAW AT Xr&€

4 J+*+*G

& - ,t ;4 cc;g ~

issues, we narrowed the choices to a few

approved by a resounding majority but not by acclamation.  (That name and the
                                        &hC
letter officially proposing it to the scientific community

'identified a name

@

I

presented in the

Postscript.) The issue that remains is a practical one:

can we bring this

name into common use?

Or will we simply have added an academic suggestion,

albeit one enshrined in the official logs of an international agency,  to the

list of names already used?

Colleagues sometimes ask why we should care about this issue, with

convictions that go beyond the usual manifestations of good scientific

citizenship. During our deliberations, at least two components of our

commitment have emerged.

As proponents of the disciplines of virology and

genetics, we want them to operate in a way that seems to us and to students we

train scientifically sound.

Moreover, we would like the public view of our

science to be one that acknowledges the reality of disagreement,  but one that

also admires the willingness of intellectual combattants to compromise for

advancement of our shared purpose.

A prolonged battle over the issue of a

name for the AIDS virus is more than trivial, because it may come---some might

say it has come---to symbolize certain excesses of character for which

scientists are often criticized. This seems particularly unfortunate in view

of the truly extraordinary accomplishments that have been made in a very short

time by the laboratories involved in the debate.

A generally accepted

recommendation for a name for this virus could help restore to the public

whatever faith has been been lost through this issue during the confusion of

the past two years.

Pos t scriD t

group, Myron Essex and Robert Gallo, declined to sign the agreement.

To the Editor:

The undersigned are members of a subcommittee empowered by the International

Committee on the Taxonomy of Viruses to propose an appropriate name for the


retrovirus isolates recently implicated as the causative agents of the

acquired immune deficiency syndrome (AIDS). Adoption of an internationally-

acceptable name for this group of viruses has become an important issue

because of the widespread interest in AIDS and its origins and because of the

multiplicity of names currently in use.  Thus the several isolates of what are

now evidently closely related members of the same virus group have been called

lymphadenopathy associated virus (LAV), human T cell lymphotropic virus type

I11 (HTLV-111), immunodeficiency associated virus (IDAV), and AIDS-associated

retrovirus (AEW). At present, two compound names (HTLV-III/LAV and LAV/HTLV-

111) are also used in scientific publications, and the colloquial name, the

AIDS virus, is often used by the press.

1,

We are writing to propose that the AIDS retroviruses be officially designated

as the human immunodeficiency viruses,  to be known in abbreviated form as HIV.

i.,; We have considered several issues that bear upon this proposal.

(i) The name

conforms to common nomenclature for retroviruses, beginning with the host

species ("human") , ending with "virus," and containing a word that denotes a

major (though not the only) pathogenetic property of the prototypic members of

the group ("immunodeficiency") . ("Feline leukemia virus" and "mouse mammary

tumor virus" are two well-known examples of such names for retrovirus

species.)

with which the virus group is associated, it does not incorporate the term

"AIDS", which many clinicians urged us to avoid.

distinguished from all existing names for this group of viruses and has been

chosen without regard to priority of discovery.  (iv) The name is sufficiently

distinct from the names of other retroviruses to imply an independent virus

species, a group of isolates that can presumably exchange genetic  information

readily with each other but not with members of other known retrovirus

species. These other species include the human T cell leukemia viruses (e.g.,

HTLV-1 and -2), which will continue to be named according to a convention

adopted by several leading investigators in September, 1983. (Though roman

numerals are often used to indicate the type of HTLV,  arabic numbers were

originally prescribed in the agreement and are thus used here.)

Retroviruses isolated from subhuman primates and found to be genetically

related and biologically similar to HIV's should be designated as

immunodeficiency viruses of the appropriate host species (e.g., simian

immunodeficiency virus [SIV] or African green monkey immunodeficiency virus

[AGMIV]). (vi) Because HIV isolates are numerous and display considerable

genetic heterogeneity, particularly in the gene, it will be necessary for

each laboratory to assign subspecies designations to their isolates.

recommend that each laboratory adopt a code with geographically informative

letters and sequential numbers to identify their isolates (e.g.,  the 42nd

isolate at the University of Chicago could be described as HIV [CHI-421).

Initially, the existing, well-characterized isolates, such as LAV-1, HTLV-

IIIB, or ARV-2, should be identified as such in publications to ease the

(ii) Though the name clearly connects the viruses to the disease

(iii) The name is readily

(v)

We

transition to a unified nomenclature.

retroviruses with clear but limited relationship to isolates of HIV (e.g.,

more than 20% but less than 50% nucleic acid sequence identity)  should not be

called HIV unless there are compelling biological and structural similarities

(vii) Any future isolates of human

to existing members of the group.

i'

i

d[( :We hope that this proposal will be adopted rapidly by the research community

working with the viruses.

John Coffin

Tufts University, Medford, MA

Ashley Haase

University of Minnesota, Minneapolis, MN

Jay A. Levy

University of California, San Francisco, CA

Luc Montagnier

Pasteur Institute, Paris, France

Stephen Oroszlan

Frederick Cancer Research Center, Frederick, MD

Natalie Teich

Imperial Cancer Research Fund, London, England

Howard Temin

University of Wisconsin, Madison, WI

Kumao Toyoshima

University of Tokyo, Tokyo, Japan

Harold Varmus, Chairman

University of California, San Francisco, CA

Peter Vogt

University of Southern California, Los Angeles, CA

Robin Weis s

Institute of Cancer Research, Chester Beatty Laboratories, London, England