BIOLOGY OF REPRODUCTION 63, 667676 (2000)

EDITORS PREFACE
The first paper in this issue of Biology of Reproduction was presented as an invited lecture at the 32nd Annual
Meeting of the Society for the Study of Reproduction (from 31 July to 3 August 1999) at Washington State University,
Pullman, WA. The lecture was part of the Presidents Symposium, entitled Sexual Differentiation, organized by Dr.
Michael D. Griswold, President of the Society for the Study of Reproduction.
This paper, Human Y Chromosome, Sex Determination, and SpermatogenesisA Feminist View, by Jennifer A.
Marshall Graves (Department of Genetics, La Trobe University, Melbourne, Victoria 3083, Australia), did not undergo
peer review. Except for minimal editorial changes, the paper is published as submitted.
Virendra B. Mahesh, Ph.D., D. Phil.
Editor-in-Chief, Biology of Reproduction

Human Y Chromosome, Sex Determination, and Spermatogenesis

A Feminist View1
Jennifer A. Marshall Graves2
Department of Genetics, La Trobe University, Melbourne, Victoria 3083, Australia
ABSTRACT

and when this happens, the phenotype of the baby is female. Female development has therefore been said to be
the default pathways, and females have even been called
mutant males.
The production of a male is likely to require many more
genes than just the TDF gene. Some of these are becoming
known through studies of patients with a variety of sex
reversal syndromes. There are also likely to be many
geneshundreds or thousandsrequired for germ-cell differentiation and male fertility. Of course, there are likely to
be just as many genes required for ovarian differentiation
and egg development, and so far we know rather little about
these genes or how they are switched on in the absence of
testis development.

In this review I want to argue that, far from being a macho

entity with an all-powerful role in male development, the human
Y chromosome is a wimp. It is merely a relic of the X chromosome, and most or all of the genes it bearsincluding the
genes that determine sex and control spermatogenesisare relics of genes on the X chromosome that have other functions
altogether.

Sertoli cells, spermatogenesis

HUMAN SEX, GENES, AND CHROMOSOMES

In humans, as in other mammals, sex determination depends on the testis, and testis differentiation depends on the
Y chromosome.

Sex Chromosomes

Sex Determination

In mammals, the male-specific Y chromosome plays a

pivotal role in sex determination, and also bears genes that
are required for spermatogenesis. However, not all the
genes that are needed to make a testis or to make germ
cells need to be on the Y chromosome, and many are
known to be located on the X chromosome or on the autosomes (chromosomes other than the X and Y).
Like other mammals, human females have two X chromosomes (XX) and males have a single X and a single Y
chromosome (XY). The X is large (5% of the total length
of a single set of chromosomes) and bears a proportional
number of genes (3000 or 4000), which have a variety of
functions much like those of genes located on other chromosomes. To ensure fair play between the sexes, only one
X chromosome is genetically active in female cells. The set
of genes on the X chromosome is almost completely conserved between different species of eutherian (placental)
mammals, probably because breaking it up would disrupt
this chromosome-wide X-inactivation mechanism.
The Y chromosome is much smaller than the X chromosome and contains only a few genes. It is largely composed of repetitive sequences that preferentially bind fluo-

The event that marks human male determination is the

differentiation of the testis at about 5 wk of gestation. The
embryonic testis then churns out testosterone and Mullerian-inhibiting substance, powerful hormones that divert developmental pathways along male lines. The pivotal genetic
event is the switching of the indifferent gonad (the genital
ridge) to testis differentiation by the action of a gene called
the testis-determining factor (TDF). After this event, male
development unfolds automatically, unless something goes
wrong. A number of genetic accidentsanything from absence or mutation of TDF to a block of the action of testosteronecan interrupt the male-determining pathway,
Supported by Australian Research Council grant A00000786 and National Health and Medical Research Council (Australia) grant 980990.
2
Correspondence: Jennifer M. Graves, Department of Genetics, Building
NW7, La Trobe University, Kingsbury Drive, Bundoora, Melbourne, Victoria 3093, Australia. FAX: 613 9479 2480; e-mail: j.graves@gen.
latrobe.edu.au
1

Q 2000 by the Society for the Study of Reproduction, Inc.

ISSN: 0006-3363. http://www.biolreprod.org

667

668

GRAVES

model is the one for which I shall argue. It holds that the
Y chromosome is a wimp, a pale shadow of its former
self, having degraded to almost nothing. The genes that it
contains are just relics of genes that were originally on an
autosome and have been retained intact on the X chromosome.
The Dominant Y Chromosome

FIG. 1.

Models of the Y chromosome.

rescent dyes, so that it literally glows in the dark. These

sequences have no known coding function and are good
candidates for hard-core junk DNA. The X and Y chromosomes are homologous over a tiny region at the end of
the short arms, and they regularly pair and undergo crossing
over within this 2600-kilobase-pair (kbp) pseudoautosomal
region) (PAR). The human X and Y chromosomes also
have a 500-kbp PAR at the ends of their long arms.
A grand total of 33 genes have been characterized on
the Y chromosome, nine of these within PAR1 and four
within PAR2. Many are inactive pseudogenes. This may be
a reasonably exhaustive list, because the human Y chromosome was thoroughly searched by screening a testis
cDNA library with large fragments of the human Y chromosome cloned into yeast artificial chromosomes (Y-specific YACs) [1]. The 19 genes on the nonrecombining, Yspecific (differentiated) region of the Y chromosome are a
peculiar lot. Several are expressed only in the testis, suggesting a male-specific function in spermatogenesis. The Y
chromosome is unique in its functional coherence.
Lahn and Page [1] divided up the 19 genes on the nonrecombining region of the Y chromosome into two classes.
Class I genes are single-copy genes on the Y chromosome
that are ubiquitously expressed and have homologues on
the X chromosome. Class II genes (the interesting ones) are
multicopy and testis specific, and have no homologues on
the X chromosome. However, as I will point out, this classification breaks down when we consider the origin of two
of the best-characterized male-specific genes, one involved
in spermatogenesis and the other, TDF itself.
The Y chromosome is therefore peculiar in that it contains few active genes and a lot of junk, and it is unique
in that it is male specific and bears several genes important
for male functions. Let us consider alternative models of
the organization, function, and evolution of the human Y
chromosome.
MODELS OF THE Y CHROMOSOME

There are many models put forth to account for the peculiarities of the Y chromosome, and I have distilled them
into three categories (Fig. 1). The first represents a concept
that we have all grown up withthe concept of a dominant
entity, acting to determine a male, regardless of which other
chromosomes are present. An alternative model is that the
Y is a selfish entity, which somehow accumulates genes
that are handy in a male and/or bad in a female. The third

The reason for the traditional view of the human Y chromosome as an all-powerful entity is the dominant role it
plays in sex determination. This is evident from the sex of
patients with abnormal numbers of sex chromosomes.
Males with Klinefelter syndrome (XXY) have two X chromosomes, like a female, plus a Y chromosome, like a male
[2], whereas females with Turner syndrome (XO) have a
single X chromosome and no Y chromosome [3]. Even
patients with multiple X chromosomes (up to five) are
male, if they possess a Y chromosome, and female, if they
lack a Y chromosome. The phenotype (outward appearance) of patients is remarkably normal because all but one
X chromosome is largely inactive anyway; the major effect
of abnormal numbers of X chromosomes is infertility.
Variations in the number of X chromosomes have also
been observed in other eutherian mammals. Monkeys, cats,
horses, sheep, cattle, pigs, rats, and mice with an XO karyotype are female, and cats and mice with an XXY karyotype are male. As in humans, the abnormal number of X
chromosomes has little effect, except on fertility. Thus, the
rule for eutherian mammals is that the presence of a Y
chromosome makes the mammal a male, no matter how
many copies of the X chromosome are present. This has
been thought to imply that there is a gene on the Y chromosome (TDF) that has a positive action in determining
the testis, and that there is no dosage effect of the X chromosome. We therefore think of the Y chromosome as dominant, because it bears TDF, which determines the testis
even in the presence of several X chromosomes.
How dominant is the Y chromosome really? Given that
only one of the X chromosomes is active anyway, the phenotypes of individuals with XO and XXY karyotypes say
little about a potential balance. There is at least one report
of a human fetus with an XXXY karyotype that was female; perhaps there are non-inactivated genes on the X
chromosome that balance the effect of the testis-determining factor. We know of at least one gene, DAX1 on the
short arm of the X chromosome, that causes XY female sex
reversal when duplicated [4]. At other Y loci that control
spermatogenesis, the Y chromosome is certainly not dominant over extra X chromosomes, for any organismman
or mousewith two X chromosomes is sterile [5]. In this
case, the presence of two X chromosomes is a killer of
fertility, if not of virility
In marsupial mammals, a dosage effect of the X chromosome is yet more obvious, because mammals with XXY
and XO karyotypes are intersexes, neither completely male,
nor completely female. Marsupial mammals with XXY karyotypes have intra-abdominal testes, but their scrotum is
replaced by a pouch with mammary glands. Mammals with
an XO karyotype have no testes, but their pouch and mammary glands are replaced with an empty scrotum [6]. Thus,
the testis does not control all of the downstream events in
marsupials, and at least some sexual differentiation (scrotal
or mammary development) seems to be a function of the
numbers (or parental origin) of the X chromosomes [7].
Even so, the Y chromosome controls whether animals have

THE HUMAN Y CHROMOSOME AND SEXA FEMINIST VIEW

a testis or not, implying that marsupials also have a dominant TDF on the Y chromosome. Thus, eutherian and marsupial TDF has what looks like a male-dominant action,
although the evidence for this action is less convincing than
has been appreciated. This dominant action has traditionally
been interpreted to mean that TDF codes for some kind of
activator that turns on transcription of other genes in the
male-determining pathway. However, I shall argue that this
male-dominant action results from completely the opposite
actionthe testis-determining gene on the Y chromosome
acts as a spoiler that turns off genes that turn off testis
determination.
The Selfish Y Chromosome

The Y chromosome is unique in that it is found only in

males. This chromosome would be the perfect place for
evolution to relocate genes that have a function only in
males, as well as genes that would have an adverse effect
in females. Has the Y chromosome become a repository for
such genes?
Hurst [8] argued that the Y chromosome acts as an attractor for selfish growth factors. Such selfish, Y-borne
genes favor implantation and embryonic growth, even at
the expense of the future reproduction of the female (perhaps with another male), as has been suggested for the action of imprinted genes expressed only from the paternal
genome. An embryo carrying a Y chromosome containing
such a gene will be selected, and this selfish Y chromosome
will spread through the population.
As predicted by this hypothesis, there are some genes
on the Y chromosome that affect growth. However, three
of these, the osteogenesis gene (SHOX) and three growth
factor receptor genes (IL3RA, CSF2RA in PAR1, and Il9R
in PAR2), lie within the PAR, and are therefore shared with
and recombine with genes on the X chromosome. Another
factor has been identified only as a region that, when deleted, causes overgrowth of the undifferentiated gonad
(gonadoblastoma), so this factor is also more likely to represent a growth inhibitor (or gonad differentiation promoter) than a growth factor.
The selfish Y hypothesis also requires that selfish genes
on the Y chromosome are truly Y specific and would therefore correspond to Class II genes as defined by Lahn and
Page [1]. These Class II genes have no X homologue and
hail from other regions of the genome. It is therefore particularly interesting to examine the origin of these genes. I
will show here that at least some of them have homologues
on the X chromosome and were therefore homegrown on
the sex chromosomes. Some beautiful experiments on fruitflies [9] suggested that a gene already on the Y chromosome could evolve a selfish male-advantage/female-disadvantage function. In these experiments, an allele of an autosomal gene was artificially selected in females only. After
only 30 generations, the region containing this allele was
shown to have a detrimental effect on male viability and
fertility.
The Wimp Y Chromosome

The third model proposes that the Y chromosome is

merely a relic of a former autosome, and all that the genes
that it bears are relics of genes that happened to be on the
chromosome that accidentally became a Y chromosome.
This model derives from the suggestion that the mammalian
X and Y chromosomes differentiated from a homologous

669

pair of ancient autosomes (proto-sex chromosomes) by the

progressive degradation of the Y chromosome [10].
The Y chromosome is thought to have been originally
defined by its acquisition of a male-determining gene [11].
Other genes with male-specific functions then accumulated
nearby, and selection kept a male-specific package together
by suppressing crossing over. Within this genetically isolated region, all kinds of genetic accidents occurred and
could not be repaired because there was no crossing over
between the X and Y chromosomes, so the region was rapidly degraded and deleted. The only genes that could survive on the Y chromosome were those that became indispensable in male reproduction.
Information on the possible origins of genes on the Y
chromosome can help to distinguish between the models.
Class I genes with homologues on the X chromosome are
likely to be relics of ancient homology, as predicted by the
wimp Y hypothesis. However, Class II genes that are testis
specific and have no homologues on the X chromosome
may have been recruited by a selfish Y.
THE ORIGIN AND EVOLUTION OF THE Y
CHROMOSOME

These models and their predictions about the origin of

genes on the Y chromosome can be examined by looking
for evolutionary intermediates of Y degradation in the
mammals most distantly related to humans. Comparisons
of eutherians with marsupials (pouched mammals that diverged 130 million years ago) and monotremes (egg-laying
mammals that diverged 170 million years ago) have provided a rich source of variation to study the origin and
function of mammalian sex chromosomes and sex-determining genes.
Sex Chromosome Differentiation

The wimp Y model has its origins in a hypothesis put

forward three decades ago to explain the peculiarities of
snake sex chromosomes. The pattern in snakes is completely opposite from the pattern in mammals: snakes with a
ZW karyotype are female and snakes with a ZZ karyotype
are male. Like the mammalian X chromosome, the Z chromosome is large, containing about 6% of the genome in all
snake families. The W chromosome is much more variable.
Some snake families have a tiny, heterochromatic W chromosome, whereas others have a W chromosome that is
much the same size as the Z chromosome. Ohno [10] suggested that these patterns represent stages in the gradual
breakdown of the W chromosome, starting from a pair of
equivalent proto-sex chromosomes.
Birds also have a ZZ male and ZW female system, and
cross-species chromosome painting (in situ hybridization
with a mixture of DNA sequences from an isolated Z chromosome) shows that even birds as distantly related as the
chicken and the emu have a genetically identical Z chromosome [12]. As in snakes, the W chromosome in birds
has different sizes in different families; it is large in ratite
(flightless) birds and small in carinate birds. Chromosome
painting with the chicken Z chromosome sequences shows
that the W chromosome is largely homologous to the Z
chromosome in the emu, but has become small and heterochromatic in the chicken. A process of W chromosome degradation has therefore taken place to different extents, independently in different bird and snake lineages.
The same sort of event is proposed to have occurred in
mammals. Although the X and Y chromosomes are very

670

GRAVES

different in size and gene content, there is excellent evidence that they were once homologous, and the Y chromosome has been largely degraded during 200 million
years of mammalian evolution. Significantly, the human X
and Y chromosomes share considerable homology. First,
the PAR is homologous, and this region contains at least
nine genes. Second, many genes on the differentiated region of the Y chromosome also have obvious homologues
on the X chromosomegenes that share DNA sequence
and code for similar protein products. Since comparative
mapping shows no homology between the mammalian XY
pair and the bird ZW pair, these two sex chromosome systems must have evolved independently.
The Ancient X and Y Chromosomes

The gene content of the ancient mammalian sex chromosomes can be deduced by comparing eutherian sex chromosomes with those of the distantly related marsupial and
monotreme mammals and by defining the shared regions.
Sex chromosomes have been compared in two groups of
marsupialsmacropodids (kangaroos and wallabies) and
dasyurids (small rodent-like insectivores)as well as in
two of the three species of monotreme, the fabled platypus
and its spiny cousin, the echidna.
Marsupials have a genome about the same size as that
of eutherian mammals, but their DNA is packaged into a
few very large chromosomes. However, the X and Y chromosomes are unusually small, the dasyurid X chromosome
constituting only about 3% of the haploid genome, and the
Y chromosome barely visible under the microscope as a
tiny dot. Gene mapping reveals that these small marsupial
X and Y chromosomes are minimal mammalian sex chromosomes, little changed from the ancient mammalian X
and Y chromosomes in a common ancestor of all mammals
170 million years ago.
Comparative mapping of the marsupial X chromosome
was accomplished by making hybrids between rodent and
marsupial cells. The hybrids retained only one or two marsupial chromosomes in a background of rodent chromosomes, so it was possible to identify marsupial proteins in
hybrids by isozyme typing and marsupial genes by Southern blotting. The presence or absence of a particular gene
or gene product could then be correlated with the presence
or absence of a particular marsupial chromosome. More
recently, a number of marsupial homologues of human Xlinked genes have been cloned and mapped by fluorescence
in situ hybridization.
This gene mapping showed that much of the human X
chromosome is homologous to the smaller marsupial X
chromosome [13]. All of the genes on the long arm of the
human X chromosome and on the region around the centromere were also found on the X chromosome in marsupials, implying that this region represented an ancient mammalian X chromosome at least 130 million years old. The
same set of genes was also mapped to the X chromosome
in monotremes, pushing back the age of this chromosomal
region to 170 million years. This region (called the X conserved region, XCR) is equivalent to the ancient mammalian X chromosome.
The Y chromosome is also partially homologous between humans and marsupials, but represents only a tiny
region. Four genes from the human Y chromosome have
now been found to map to the tiny marsupial Y chromosome and must have been on the Y chromosome for at least
130 million years (personal communication with Waters).

These genes, SRY, RPS4Y, SMCY, and RBMY, map to two

small regions at either end of the human Y chromosome,
which constitutes the Y conserved region (YCR). They
were probably originally contiguous at the border of the
large heterochromatic region, until an inversion separated
them a few million years ago. The small conserved region
of the human Y chromosome (constituting as little as 10
Mb) is all that is left of the original mammalian Y chromosome.
Regions Added to the Sex Chromosomes

Surprisingly, the genes on the rest of the short arm of

the human X chromosome were all found on autosomes in
marsupials and in monotremes. Most of them cluster on the
short arm of chromosome 5 in the tammar wallaby and on
chromosome 1 in monotremes. This pattern could mean
either that the marsupial and monotreme X chromosomes
both lost a large chunk after their divergence from eutherians, or that the human X chromosome gained a chunk.
The latter possibility is favored, since the same piece of X
chromosome is unlikely to have been lost independently
from the marsupial and monotreme X chromosomes. This
conclusion suggested that the human X chromosome contains an X added region (XAR), which is still located on
autosomes in other mammals [13].
Recently, the conserved and added regions of the human
X chromosome have been directly demonstrated by chromosome painting between the marsupial and human X
chromosomes [14]. A DNA probe derived from a flowsorted wallaby X chromosome was amplified and tagged
with fluorescent dye, then hybridized in situ to human chromosomes. The probe bound to the long arm and pericentromeric region of the human X chromosome, identifying
the XCR that represents the ancient X chromosome. It did
not bind to the rest of the short arm that represents the
added region, XAR (Fig. 2).
The human Y chromosome contains an equivalent added
region. The presence of this region was first obvious from
a map of the X chromosome, since many genes on the
human Y chromosome have partners on Xp within the
XAR. They, too, must therefore have been added. More
recently, the addition has been directly demonstrated by
cloning and mapping the wallaby homologues of genes on
the human Y chromosome (personal communication with
Waters). Eight of these genes mapped to chromosome 5p
in the wallaby, to the same position as the XAR genes.
These genes define a Y added region (YAR), which shares
homology with the XAR. The YAR forms the great majority of the human Y chromosome and includes the PAR.
How could a piece of an autosome be added to both the
X and Y chromosomes? We know that a piece of autosome
added to the differential region of the X chromosome could
not be transferred to the Y chromosome, because the X and
Y chromosomes do not pair over this region. Instead, a
compound X chromosome is formed, which pairs with a
Y1 (the original Y) region and a Y2 (the original autosome)
region at meiosis. The addition is therefore likely to have
occurred at the stage when the X and Y chromosomes were
only partially differentiated. If a piece of autosome was
added to the X chromosome above the PAR, it would cross
over onto the Y chromosome at the next meiosis. Once the
piece was on the X and Y chromosomes, it would enlarge
the PAR [13]. There may have been two or more separate
additions, since the XAR is present as at least two autosomal clusters in monotremes and marsupials.

THE HUMAN Y CHROMOSOME AND SEXA FEMINIST VIEW

671

FIG. 2. Cross-species chromosome painting of the wallaby X chromosome onto the human X chromosome [14]. The conserved region is labeled; the
added region is not.

Thus, most of the human Y chromosome derives from a

relatively recent addition to the sex chromosomes. The
original and added regions of the Y chromosome have both
been greatly degraded in size and gene content, and the
older YCR has all but disappeared.
The Y ChromosomeGoing, Going . . .

Thus, gene mapping shows that the human Y chromosome evolved from the proto-X/Y chromosome plus the
added region, now represented by the human X chromosome, with which it was once homologous.
But what happened to all the genes on the original Y
chromosome? More than 1000 genes have been cloned
from the human X chromosome, compared with only about
30 on the Y chromosome, despite exhaustive screening. A
clue to the genes fate is given by the finding that several
X-linked genes detect inactive homologues on the Y chromosome. For instance, the sex-linked steroid sulfatase
(STS) gene on the X chromosome detects homologous sequences on the Y chromosome, but these belong to a partially deleted pseudogene [15]. The Y copy of UBE1X has
disappeared altogether from the human Y chromosome, although bits of pseudogenes can be detected on the Y chromosome in monkeys; an active copy of the gene still resides
on the mouse Y chromosome, and the gene is pseudoautosomal in the platypus [16]. Thus, the process of Y degradation continues.
Indeed, additions may have saved the human Y chromosome from disappearing altogether. In addition to the
added autosomal segment, the Y chromosome has also been

enlarged by the addition or amplification of repetitive-sequence DNA. Almost half of the human Y chromosome is
composed largely of two simple repeats. In the kangaroo
and wallaby, the ribosomal genes and associated heterochromatin have been added to both the X and Y chromosomes and have all but degenerated on the Y chromosome
[17]. In other species, large aggregates of heterochromatin
have been added solely to the Y chromosome. Maybe additions of heterochromatin act as a kind of ballast.
Is the Y chromosome therefore headed for extinction?
Without further additions to the Y chromosome in dasyurid
marsupials, the unfortunate Y chromosome has been whittled down to a minimal 10 Mb. This little package evidently
contains a TDF needed to determine maleness, as well as
sundry spermatogenesis genes, but little else. Indeed, there
are species of marsupials that dispense with the Y chromosome during somatic growth [18] by eliminating it from
the embryo. This process must mean that the Y chromosome no longer bears any genes required for general functions. There are two rodent species (mole voles) that have
completely lost the Y chromosome [19]. Presumably, in
these species, the Y chromosome first lost everything but
the TDF, whose control function was then taken over by a
new gene elsewhere on the genome.
If the Y chromosome is a relic of the original undifferentiated proto-X/Y chromosome that derived from a pair of
autosomes with no sex-determining function, it follows that
all of the genes on the Y chromosome evolved from genes
now represented on the X chromosome. It is therefore of
interest to examine the origins of genes on the Y chromo-

672

GRAVES

some, particularly those suspected of male-specific functions, like spermatogenesis and sex determination.
ORIGIN OF SPERMATOGENESIS GENES ON THE Y
CHROMOSOME

There is evidence that one or more genes on the Y chromosome are essential for spermatogenesis. Deletion of parts
of the human Y chromosome near the heterochromatic region have been found in men producing few or no sperm.
Deletions of part of the tiny short arm of the mouse Y
chromosome and of several parts of the long arm are associated with an absence of sperm or abnormal sperm.
There are several genes on the Y chromosome that map
within these regions, and these become candidate spermatogenesis genes. Their origin is of particular interest for
evaluating models of the Y chromosome.
Spermatogenesis Genes from the Ancient Proto-X/Y
Chromosome

The RBM gene (named for the RNA-binding motif that

its protein product contains) is present in about 30 copies
on the human Y chromosome, but only two of these copies
are transcribed [20]. The RBM gene appeared to be a prototypical Class II gene: it is multicopy and testis specific,
and apparently has no homologue on the X chromosome.
Where did it come from, then? Was it an ancient, autosomal
male-specific gene that found an appropriate spot on a selfish Y chromosome? A close homologue (called HNRPG)
has been described on chromosome 6 and codes for one of
a large family of ubiquitously expressed RNA-processing
and transport proteins. The speculation was that a copy of
this autosomal gene, introns and all, somehow made its way
onto the Y chromosome and became modified and testis
specific. It acquired a role in spermatogenesis and was then
amplified in tandem on the human Y chromosome.
However, a spin-off from recent studies of the gene in
marsupials shows that RBM is really a Class I gene after
all. Both X-borne and Y-borne copies of this gene were
detected, first in marsupials and then in eutherian mammals.
When human RBM was used to detect homologous sequences in Southern blots of marsupial DNA, dosed
fragments (with twice the intensity in females as in males)
were observed, as well as a male-specific band. An RBM
homologue with a sequence identical to that of HNRPG was
then isolated from the human X chromosome; in fact, the
supposedly autosomal HNRPG turned out to be an X-borne
gene after all, and the chromosome 6 homologue was exposed as an intronless pseudogene that could not be translated into a sensible protein [21]. Both X and Y copies of
Rbm were also found in the mouse [22]. Thus, the Y-borne
RBMY (as it is now called) has a homologue on the X
chromosome, RBMX, just as do all the ubiquitously expressed Class I genes on the human Y chromosome. Like
these genes, RBMY must have evolved from genes on the
original proto-X/Y chromosome. The RBMY gene is therefore a relic of the conserved ubiquitous RBMX.
How could a gene coding for a generic RNA-processing
protein be molded into a spermatogenesis gene? The sequence similarity between RBMX and RBMY implies that
their functions in RNA processing are likely to be similar.
The most obvious sequence differences between RBMX and
RBMY are the internal amplification of an intron containing
a SRGY box containing serine-arginine-glycine-tyrosine, but since this amplification is not present in the mouse
or marsupial gene, it is likely that the major difference lies

in its testis-specific expression. Perhaps its product wields

post-transcriptional control over other genes specifically in
the testis.
Spermatogenesis Genes from the Added Region

Two genes within the YAR on the Y chromosome may

also have a function in spermatogenesis. Both ZFY and
DFFRY have copies on the X chromosome, from which the
Y-borne genes were obviously derived. Since their marsupial homologues are autosomal, they must derive from
genes on the autosomal region that were added recently to
the eutherian X and Y chromosomes. ZFY and ZFX are both
ubiquitously expressed in humans and code for a zinc finger
protein that resembles a transcription factor. In mice, Zfx is
expressed ubiquitously, but Zfy is testis specific and lies in
a critical chromosomal region for spermatogenesis [23].
Perhaps human Zfy retains the original housekeeping function that it shares with Zfx, but Zfy has recently acquired a
specific function in spermatogenesis in rodents.
The DFFRY gene has a similar origin in the YAR, but
may have already had an ancient function in germ-cell differentiation, before it was relocated to the sex chromosomes. Like ZFY, this gene is ubiquitously expressed in
man, but testis specific in the mouse [24]. Significantly,
mutations in DFFRY have been detected in sterile men [25].
Even more telling is the observation that the gene has a
homologue in Drosophila that affects ovarian function, as
well as eye development (indeed, its name derives from
Drosophila fat facets). Evidently, then, this gene has an
ancient function in the development of germ cells, as well
as in other tissues, and its relocation on the mammalian Y
chromosome allowed it to specialize in sperm development
in males. Both DFFRX and DFFRY belong to a family of
genes that affect protein stability, so DFFRY may be another testis-specific, posttranscriptional regulator that
evolved from a gene with an ancient function in germ-cell
development in both sexes.
The three candidate spermatogenesis genes RBMY, ZFY,
and DFFRY, would all be classified as Class I genes with
their origin in homologous genes on the X chromosome,
either within the conserved or the added region. Are there,
after all, any Class II genes that were relocated from other
chromosomes to a selfish Y chromosome? Two candidate
spermatogenesis genes, which apparently have no X chromosome homologues, may point to alternative origins for
Y-borne spermatogenesis genes.
Genes Relocated to a Selfish Y Chromosome?

The Deleted in Azoospermia (DAZ) gene was also

cloned from a deletion interval near the heterochromatin on
the human Y chromosome and is deleted in many sterile
men [26]. This gene, too, codes for an RNA-binding protein, though one in a different family from RBMY. The DAZ
gene also exists in multiple copies and is expressed only in
the testis. However, there appears to be no X chromosome
homologue, and the only obvious source for the gene is an
autosomal homologue, DAZLA, on human chromosome 3.
In mice, there is an autosomal Dazla homologue, but no Yborne sequence, and in marsupials, only autosomal homologues can be detected [27]. The DAZLA gene is also gonad
specific. These observations suggest that a copy of the sequence somehow moved from its autosomal location early
in primate evolution. Since the gene possesses introns, it
could not have moved by retrotransposition (i.e., by integration of a DNA copy of the RNA transcript), unless the

THE HUMAN Y CHROMOSOME AND SEXA FEMINIST VIEW

673

relocation occurred via an unprocessed transcript. Significantly, DAZ is homologous to the gonad-specific boule
gene in Drosophila, mutations of which cause meiotic
block and sterility. Thus, DAZ, too, has an ancient function
in germ-cell differentiation, and a copy of this handy gene
was somehow moved from an autosome onto a selfish Y
chromosome.
The recently discovered CDY gene on the human Y
chromosome has an even more recent origin. This gene is
testis specific and multicopy, and has no X chromosome
homologue (i.e., it is a Class II gene). There is no Y chromosome homologue in any nonprimate or even in prosimians, suggesting that it moved onto the primate Y chromosome only recently. There is a homologue, CDYL, on
chromosome 6 that is transcribed ubiquitously, but in the
mouse, the autosomal homologue Cdyl is alternately transcribed into ubiquitous and testis-specific products. The autosomal homologue contains introns, but the Y-specific
CDY does not, so it evidently moved onto the Y chromosome by retroposition [28].
Both DAZ and CDY therefore do appear to be dinkum
(real) Class II genes with an autosomal origin. At least DAZ
appears to have capitalized on an ancient function in germcell differentiation, even before it somehow moved onto a
selfish Y chromosome.
Multiple Origins of Spermatogenesis Genes on the Y
Chromosome

Thus, the candidate spermatogenesis genes represent at

least four different origins for genes on the Y chromosome.
Some, like RBMY, were present on the original proto-X/Y
chromosome, and others, like DFFRY, were acquired along
with a large autosomal addition to the proto-X/Y chromosome. They all became male specific after the X and Y
chromosomes differentiated. Other genes were relocated to
the Y chromosome from autosomal copies, one by retroposition, and another by movement of a genomic copy or
an unprocessed transcript. Some candidate spermatogenesis
genes, like DFFRY and DAZ, evidently had an ancient function in germ-cell differentiation, whereas others, like RBM
and ZFY, were evidently innocent housekeeping genes that
were remodeled to fill a male-specific role.
How does this information help us to evaluate the different models of Y chromosome origin and function? It is
evident that many or most of the genes on the Y chromosome, including three with suspected functions in spermatogenesis, have copies on the X chromosome from
which they were derived, as proposed by the wimp Y model. However, at least two bona fide Class II genes must have
arisen by relocation of genes from autosomes, and at least
one of these had an ancient function in germ-cell differentiation, as proposed by the selfish Y hypothesis.
THE ORIGIN AND FUNCTION OF TDF

The origin of the male-dominant TDF itself is of special

interest, since it is the properties of this gene that have
given the Y chromosome its undeservedly macho image.
Cloning the TDF Gene

The TDF gene on the Y chromosome was pinpointed

using DNA from patients having only parts of a Y chromosome (deletion mapping). Females with an XY karyotype have been described, in whom most of the Y chromosome was present, but obviously the TDF gene was

FIG. 3. Human and marsupial X and Y chromosomes, showing the original, conserved region (blue) and the added region (red). The heterochromatic region of the Y is gray.

missing. Conversely, males with an XX karyotype have

been described, in whom a tiny piece of Y chromosome
was added to one X chromosome [29]. Running down the
small piece of the Y chromosome that harbored TDF and
cloning the SRY gene from this region [30], was a major
triumph of this positional cloning strategy.
The SRY gene was confirmed to be the long-sought TDF
by mutation analysis in humans. Several female patients
have been described with an XY karyotype and single base
changes within the SRY gene [31]. The equivalent gene,
SRY in the mouse, was shown to be sex determining; insertion of this gene into XX embryos produced transgenic
males [32]. Other species of eutherian mammals had an
equivalent gene on the Y chromosome, and marsupials also

674

GRAVES

had a Y-borne SRY homologue [33] (a criterion that eliminated an earlier candidate gene, ZFY).
The SRY gene proved to be one of a large family of
genes that code for architectural proteins containing an
HMG box, a short stretch of amino acids that binds to DNA
and bends it through a specific angle. Presumably, this
bending brings sequences on either side of the binding sequence, or the proteins bound to them, into juxtaposition.
Other members of this SOX gene family have been shown
to code for factors that turn other genes on in development.
It was therefore expected that SRY, too, would act as a
transcriptional activator, accounting for its male-dominant
role.
The Origin of SRY

The SRY gene, the all-important male-determining

switch, might be expected to be a classic Class II gene: a
gene unique to the Y chromosome and expressed only in
the genital ridge. It was therefore a shock to find that SRY,
too, has a homologue on the X chromosome. The presence
of this homologue was first noted in marsupials, when the
human SRY detected both male-specific bands (denoting a
Y-borne gene) and bands that showed clear dosage differences between males (one dose) and females (two doses)
on Southern blots, which revealed an X-borne homologue
[34]. This SOX3 gene was eventually also localized to the
X chromosome in humans and mice, and is expressed in
the central nervous system (CNS), as well as in the genital
ridge [35]. Thus, SRY is really a Class I gene that, like most
of the other genes on the human Y chromosome, evolved
from a counterpart on the X chromosome.
How did the CNS-determining SOX3 gene evolve into
the testis-determining SRY? Sequence comparisons between
species suggest that SRY is a degraded relic of SOX3 with
quite a different (and perhaps opposite) action. It may even
have different modes of action in different species. It is
poorly conservednot what you would expect of a gene
critical for reproduction. Even within the critical HMG box,
the amino acid sequence is only about 80% identical between the human, mouse, and marsupial SRY genes, and
outside the box, the sequences cannot even be lined up.
Strangely, it appears that the mouse Sry gene contains an
activating domain, essential for sex determination, that is
absent in SRY genes of other species [36]. Another surprise
is that the transcription pattern of this gene is inconsistent.
As we would expect, it is transcribed in the genital ridge
in the mouse, just before testis determination; but it is expressed more widely in the human embryo and is virtually
ubiquitous in marsupials. The angle of bending, too, appears to be different in mouse and human. Is it possible
that this critical gene has different actions in different species?
The Action of SRY

Is SRY a transcriptional activator, as we would expect

from its male-dominant action? Strangely, its properties
look more like those of an inhibitor, and it has been suggested that it may operate not by turning on testis-differentiating genes, but by turning off a testis inhibitor (perhaps
SOX3) [37]. Another related gene, SOX9, was found to be
involved in testis determination in mammals and other vertebrates [3840] and may interact with related genes to effect testis determination. This idea has been supported by
the recent finding that mice with an insertional mutation
upstream of SOX9 are male, even in the absence of SRY,

suggesting failure to bind an inhibitor at an upstream control site (Bishop et al., unpublished communication).
The SRY gene need not necessarily be the original mammalian sex-determining gene. Indeed, there is no direct evidence that SRY is sex determining in marsupials, and no
SRY has yet been discovered in monotremes. A possible
forerunner to SRY is a mutation of SOX3 that no longer
inhibits SOX9. The SOX3 gene may have originally worked
in a dosage-sensitive manner, requiring two doses of the
normal SOX3 to inhibit SOX9 and produce a female [37].
There may have been an even earlier mammalian sexdetermining gene, perhaps represented by a new candidate
sex-determining gene, ATRY, that was recently cloned from
the marsupial Y chromosome. The ATRY gene is expressed
appropriately in the marsupial embryo and is testis specific,
whereas SRY is ubiquitous. The ATRY gene, too, has an X
homologue from which it obviously evolved. In humans
and mice, there is an X homologue, ATRX, and mutations
in this gene causes XY female sex reversal. However, the
ATRY gene has evidently been lost from the Y chromosome
in eutherian mammals. Perhaps SOX3 first took over from
ATRY in eutherians, then SRY later took over from SOX3.
In rodents, Sry has evidently acquired a new function, and
in mole voles, it has been superseded yet again with the
loss of the entire Y chromosome.
Thus, even the mighty TDF is a wimpa degraded relic
of a normal CNS-determining gene that just gets in the way
of another gene (SOX3?) in repressing SOX9. The SRY gene
probably was not the first sex-determining gene in mammals and will not be the last, if mole voles are a portent of
the future.
CONCLUSIONS
Of the three models of the human Y chromosome presentedthe dominant Y chromosome, the selfish Y chromosome, and the wimp Y chromosomethe last model has
the most explanatory power. Homology of the human Y
chromosome with the X chromosome within the PAR and
outside of the PAR suggests an origin as an ancient autosome pair. The genetic paucity and high content of repeated
sequences and pseudogenes of the Y chromosome tells a
tale of genetic degradation. Homology of a small part of
the human Y chromosome with the tiny marsupial Y chromosome reveals the last vestiges of the ancient mammalian
Y chromosome, a region of only a few megabases, containing only four known genes. Homology of the bulk of
the human Y chromosome with autosomes in marsupials
reveals a relic of a recent autosomal addition.
The ancient and added regions of the human Y chromosome may be represented on a map (Fig. 3). There is
practically nothing left of the ancient mammalian Y chromosome (blue), and most of the chromosome is derived
from the autosomal addition (red), which itself is degrading
fast. This degraded Y chromosome is a wimp by any measure. It is deficient at every locus within the differentiated
region. From its beginning as a gene-rich autosome, equivalent in gene content to the X chromosome, it has lost most
of its 3000 or 4000 genes, as they were progressively mutated and deleted. A few pathetic relics remain as inactive
pseudogenes. A mere handful of genes have clung to survival because of mutations that allowed them to adopt a
male-specific function essential for sex determination or
spermatogenesis.
However, there are at least a few genes on the human Y
chromosome that do not seem to originate from the ancient
proto-X/Y chromosome or from autosomal addition. These

THE HUMAN Y CHROMOSOME AND SEXA FEMINIST VIEW

genes must have been incorporated quite recently into the

primate Y chromosome, by integration either of the DNA
copy of an RNA transcript, or of a genomic copy of an
autosomal gene. These genes would have to be colored
green on our map (Fig. 3). So the Y chromosome is not
only a wimp, it is a selfish wimp.
Perhaps, then, it is time to reassess the adage that maleness is dominant, and femaleness is a default condition. The
view of females as some sort of deficient, mutant males
arises from observations that an embryo with an XY karyotype will develop into a female if anything goes wrong
with the male sex-determining pathway, from the absence
of a functional SRY gene to the failure to bind and use
testosterone. In genetic terms, however, it is the Y chromosomethe genetic element that defines malenessthat
is the mutant element.
This wimp Y chromosome is disappearing fast, and, indeed, its future looks grim. In marsupials, there is only a
tiny remnant left, and this remnant can be dispensed with
in somatic tissues of some species. Even in eutherians, animals in which the Y chromosome has been salvaged by
the addition of a large bit of autosome and some heterochromatic ballast, the Y chromosome is small and degenerate, and has few essential functions left. If we come back
in 10 or 100 million years, we may well find that humans,
like mole voles, have dispensed with the Y chromosome
entirely. Perhaps they will have evolved an entirely new
pair of sex chromosomes, beginning with an autosomal
gene that took over a sex-determining function from SRY,
the last in a long line of apparently disposable sex-determining genes.
REFERENCES
1. Lahn B, Page DC. Functional coherence of the human Y chromosome.
Science 1997; 278:675680.
2. Jacobs PA, Strong JA. A case of human intersexuality having a possible XXY sex-determining mechanism. Nature 1959; 183:302.
3. Ford CE, Jones KW, Polani PE, de Almeida JC, Briggs JH. A sex
chromosome anomaly in the case of gonadal dysgenesis. Lancet 1959;
1:711713.
4. Zanaria E, Muscatelli F, Bardoni B, Strom TM, Guioli S, Guo W, Lalli
E, Moser C, Walker AP, McCabe ERB, Meitinger T, Monaco AP,
Sassone-Corsi P, Camerino G. An unusual member of the nuclear
hormone receptor superfamily responsible for X-linked adrenal hypoplasia congenita. Nature 1994; 372:635641.
5. Burgoyne P. Y chromosome function in mammalian development.
Adv Dev Biol 1992; 1:129.
6. Sharman G, Hughes R, Cooper D. The chromosomal basis of sex
differentiation in marsupials. Aust J Zool 1990; 37:451466.
7. Cooper DW, Johnston PG, Graves JAM, Watson JM. X-chromosome
inactivation in marsupials and monotremes. Semin Dev Biol 1993; 4:
117128.
8. Hurst LD. Embryonic growth and the evolution of the mammalian Y
chromosome. I. The Y as an attractor for selfish growth factors. Heredity 1994; 73:223232.
9. Rice WR. Genetic hitchhiking and the evolution of reduced genetic
activity of the Y sex chromosome. Genetics 1987; 116:161167.
10. Ohno S. Sex Chromosomes and Sex Linked Genes. Berlin: Springer
Verlag; 1967: 524.
11. Charlesworth B. The evolution of sex chromosomes. Science 1991;
251:10301033.
12. Shetty S, Griffin DK, Graves JAM. Comparative painting reveals
strong chromosome homology over 80 million years of bird evolution.
Chromosome Res 1999; 7:289295.
13. Graves JAM. The origin and function of the mammalian Y chromosome and Y-borne genesAn evolving understanding. Bioessays
1995; 17:311320.
14. Glas R, Graves JAM, Toder R, Ferguson-Smith MA, OBrien PC.
Cross-species chromosome painting between human and marsupial
directly demonstrates the ancient region of the mammalian X. Chromosome Res 1999; 10:11151116.

675

15. Yen PH, Marsh B, Allen E, Tsai SP, Ellison J, Connolly L, Neiswanger
K, Shapiro LJ. The human X-linked steroid sulfatase gene and a Yencoded pseudogene: evidence for an inversion of the Y chromosome
during primate evolution. Cell 1988; 55:11231135.
16. Mitchell MJ, Wilcox SA, Watson JM, Lerner JL, Woods DR, Scheffler
J, Hearn JP, Bishop CE, Graves JAM. The origin and loss of the
ubiquitin activating enzyme gene on the mammalian Y chromosome.
Hum Mol Genet 1998; 7:429434.
17. Toder R, Wienberg J, Voullaire L, OBrien PCM, Maccarone P, Graves
JAM. Shared DNA sequences between the X and Y chromosomes in
the tammar wallabyEvidence for independent additions to eutherian
and marsupial sex chromosomes. Chromosoma 1997; 106:9498.
18. Watson CM, Margan SH, Johnston PG. Sex chromosome elimination
in the bandicoot Isoodon macrourus using Y-linked markers. Cytogenet Cell Genet 1998; 81:5459.
19. Just W, Rau W, Vogel W, Akhverdian M, Fredga K, Graves JAM,
Lyapunova E. Absence of Sry in species of the vole Ellobius. Nat
Genet 1995; 11:117118.
20. Ma K, Inglis JD, Sharkey A, Bickmore WA, Hill RE, Prosser EJ,
Speed RM, Thomson EJ, Jobling M, Taylor K, Wolfe J, Cook, HJ,
Hargreave TB, Chandley AC. A Y chromosome gene family with
RNA-binding protein homology: candidates for the azoospermia factor AZF controlling human spermatogenesis. Cell 1993; 75:1287
1295.
21. Delbridge ML, Disteche CM, Graves JAM. The candidate spermatogenesis gene RBMY has a homologue on the human X chromosome.
Nat Genet 1999; 22:223224.
22. Mazeyrat S, Saut N, Mattei M-G, Mitchell MJ. RBMY evolved on the
Y chromosome from a ubiquitously transcribed X-Y identical gene.
Nat Genet 1999; 22:224226.
23. Koopman P, Ashworth A, Lovell-Badge R. The ZFY gene family in
humans and mice. Trends Genet 1991; 7:132136.
24. Brown GM, Furlong RA, Sargent CA, Erickson, RP, Longepied G,
Mitchell M, Jones MH, Hargreave TB, Cooke HJ, Affara NA. Characterisation of the coding sequence and fine mapping of the human
DFFRY gene and comparative expression analysis and mapping to the
srx (b) interval of the mouse Y chromosome of the Dffry gene. Hum
Mol Genet 1998; 7:97107.
25. Sun C, Skaletsky H, Birren B, Devon K, Tang Z, Silber S, Oates R,
Page DC. An azoospermic man with a de novo point mutation in the
Y-chromosomal gene USP9Y. Nat Genet 1999; 23:429432.
26. Reijo R, Lee T-Y, Salo P, Alagappan R, Brown LG, Rosenberg M,
Rozen S, Jaffe T, Straus D, Hovatta O, de la Chapelle A, Silber S,
Page DC. Diverse spermatogenic defects in humans caused by Y chromosome deletions encompassing a novel RNA-binding protein gene.
Nat Genet 1995; 10:383393.
27. Delbridge ML, Harry JL, Toder R, Graves JAM. A human candidate
spermatogenesis gene, RBM1, is conserved and amplified on the marsupial Y chromosome. Nat Genet 1997; 15:131136.
28. Lahn BT, Page DC. Retroposition of autosomal mRNA yielded testisspecific gene family on human Y chromosome. Nat Genet 1999; 21:
429433.
29. Page DC, Mosher R, Simpson EM, Fisher EMC, Mardon G, Pollack
J, McGillivray B, de la Chapelle A, Brown LG. The sex determining
region of the human Y chromosome encodes a finger protein. Cell
1987; 51:10911104.
30. Sinclair AH, Berta P, Palmer MS, Hawkins JR, Griffiths BL, Smith
MJ, Foster JW, Frischauf AM, Lovell-Badge R, Goodfellow PN. A
gene from the human sex-determining region encodes a protein with
homology to a conserved DNA-binding motif. Nature 1990; 346:240
244.
31. Hawkins JR, Taylor A, Berta P, Levillier SJ, van der Auwera B, Goodfellow PN. Mutational analysis of SRY: nonsense and missense mutations in XY sex reversal. Hum Genet 1992; 88:471474.
32. Koopman P, Gubbay J, Vivian N, Goodfellow P, Lovell-Badge R.
Male development of chromosomally female mice transgenic for Sry.
Nature 1991; 351:117121.
33. Foster JW, Brennan FE, Hampikian GK, Goodfellow PN, Sinclair AH,
Lovell-Badge R, Selwood L, Renfree MB, Graves JAM. Evolution of
sex determination and the Y chromosome: SRY-related sequences in
marsupials. Nature 1992; 359:531533.
34. Foster JW, Graves JAM. An SRY-related sequence on the marsupial
X chromosome: implications for the evolution of the mammalian testis-determining gene. Proc Natl Acad Sci U S A 1994; 91:19271931.
35. Collignon J, Sockanathan S, Hacker A, Cohen-Tannoudji M, Norris
D, Rastan S, Stevanovic M, Goodfellow PN, Lovell-Badge R. A com-

676

GRAVES

parison of the properties of Sox3 with Sry and two related genes,
Sox1 and Sox2. Development 1996; 122:509520.
36. Bowles J, Cooper L, Berkman J, Koopman P. Sry requires a CAG
repeat domain for male sex determination in Mus musculus. Nat Genet
1999; 22:405408.
37. Graves JAM. Interactions between SRY and SOX genes in mammalian sex determination. Bioessays 1998; 20:264269.
38. Foster JW, Dominguez-Steglich M, Guioli S, Kwok C, Weller PA,
Stevanovic M, Weissenbach J, Mansour S, Young ID, Goodfellow PN,
Brook JD, Schafer AJ. Campomelic dysplasia and autosomal sex re-

versal caused by mutations in an SRY-related gene. Nature 1994; 372:

525530.
39. Wagner T, Wirth J, Meyer J, Zabel B, Held M, Zimmer J, Pasantes J,
Bricarelli FD, Keutel J, Hustert E, Wolf U, Tommerup N, Schempp
W, Scherer G. Autosomal sex reversal and campomelic dysplasia are
caused by mutations in and around the SRY-related gene SOX9. Cell
1994; 79:11111120.
40. Western PS, Harry JL, Graves JAM, Sinclair AH. Temperature dependent sex determination: upregulation of SOX9 expression after
commitment to male development. Dev Dyn 1999; 214:171177.