442
T . G . HSU
expresses as if it were in a homozygous or a hemizygous state. However,
the gene is not always involved in the condensed, heteropyknotic
neighborhood. In another group of cells, the wild-type allele does not
get into the tight coils; it produces the necessary messenger R N A ,
which eventually leads to the wild-type phenotype. The final phenotype is, therefore, a mixture of patches of both the dominant and the
recessive phenotypes, or variegation. This is precisely the reason why
individuals showing V-type position effect must be heterozygous, and
the gene inactivated must be the dominant allele. The fact that phenotypes showing variegation must be localized in effect is probably
because of the lack of diffusible substance. Genes controlling gross
developmental processes or of a systemic nature, if inactivated, would
result either in lethality, or incomplete dominance, or complete
dominance through the expression of the genes not inactivated.
Available information on the variegated-type of position effect
seems to substantiate this view further. For instance, when a segment
of a chromosome is translocated or inverted to the proximity of heterochromatin, the gene nearest the heterochromatin is usually the one
that shows the most severe variegated effect, the one next to it less
so, and so on down the genetic map (e.g. Demerec and Slizynska, 1937).
In an inversion on the X chromosome (Lewis, 1950), a number of wildtype genes have been moved close to the right end of the X chromosome
and situated next to the heterochromatin. In the heterozygotes, each
of the six genes produced a variegated phenotype except one, peb
+
,
the furthest from the heterochromatin, whose expression was unaffected.
Perhaps its location was so far away from the heterochromatin that
it was never involved in the pyknotic segment, even at the periphery
of it. O f particular interest is that if further chromosomal rearrangements moved the dominant genes away from the juxtaposition of the
heterochromatin, the variegation effect would be lost (Lewis, 1950;
Hannah, 1951). According to our concept of gene inactivation, this
must be so, unless the second rearrangement carries a piece of heterochromatin with it.
Admittedly, we may be making a sweeping conclusion about the
physiological activities of heterochromatin, but if this hypothesis
stands, we may re-examine with profit a number of genetic phenomena.
In presenting data on variegation in Drosophila, Baker and Spofford
(1959) commented that "the subject is intriguing to a student of
developmental biology because of its relevance to the problem of
differentiation of cellular function". More explicitly, Schultz (1952)
theorized:
. .embryonic development is so conditioned that only
the specific genes required for function in a particular type of cell are
in active state; the others are in the 'heterochromatic' state, capable of
T . G . HSU
expresses as if it were in a homozygous or a hemizygous state. However,
the gene is not always involved in the condensed, heteropyknotic
neighborhood. In another group of cells, the wild-type allele does not
get into the tight coils; it produces the necessary messenger R N A ,
which eventually leads to the wild-type phenotype. The final phenotype is, therefore, a mixture of patches of both the dominant and the
recessive phenotypes, or variegation. This is precisely the reason why
individuals showing V-type position effect must be heterozygous, and
the gene inactivated must be the dominant allele. The fact that phenotypes showing variegation must be localized in effect is probably
because of the lack of diffusible substance. Genes controlling gross
developmental processes or of a systemic nature, if inactivated, would
result either in lethality, or incomplete dominance, or complete
dominance through the expression of the genes not inactivated.
Available information on the variegated-type of position effect
seems to substantiate this view further. For instance, when a segment
of a chromosome is translocated or inverted to the proximity of heterochromatin, the gene nearest the heterochromatin is usually the one
that shows the most severe variegated effect, the one next to it less
so, and so on down the genetic map (e.g. Demerec and Slizynska, 1937).
In an inversion on the X chromosome (Lewis, 1950), a number of wildtype genes have been moved close to the right end of the X chromosome
and situated next to the heterochromatin. In the heterozygotes, each
of the six genes produced a variegated phenotype except one, peb
+
,
the furthest from the heterochromatin, whose expression was unaffected.
Perhaps its location was so far away from the heterochromatin that
it was never involved in the pyknotic segment, even at the periphery
of it. O f particular interest is that if further chromosomal rearrangements moved the dominant genes away from the juxtaposition of the
heterochromatin, the variegation effect would be lost (Lewis, 1950;
Hannah, 1951). According to our concept of gene inactivation, this
must be so, unless the second rearrangement carries a piece of heterochromatin with it.
Admittedly, we may be making a sweeping conclusion about the
physiological activities of heterochromatin, but if this hypothesis
stands, we may re-examine with profit a number of genetic phenomena.
In presenting data on variegation in Drosophila, Baker and Spofford
(1959) commented that "the subject is intriguing to a student of
developmental biology because of its relevance to the problem of
differentiation of cellular function". More explicitly, Schultz (1952)
theorized:
. .embryonic development is so conditioned that only
the specific genes required for function in a particular type of cell are
in active state; the others are in the 'heterochromatic' state, capable of