| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cell Growth & Differentiation |
Articles |
Departamento de Genética y Fisiología Molecular, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Cuernavaca Mor. 62250 [D. E-A., F. R-T., C. V., J. S-O., L. C.]; Centro de Investigación en Ciencia Aplicada y Tecnología Avanzada, Mexico, D.F. 11500 [P. G., L. C.]; Departamento de Genética, Centro de Investigación y Estudios Avanzados del Instituto Politécnico Nacional, Mexico, D.F. 07360 [P. C., A. M., P. G.]; and Departamento de Virus y Cáncer, Instituto Nacional de Salud Pública, Cuernavaca, Mor. 62508 [L. G-X.], Mexico
Abstract
Hair follicle growth cycle proceeds through a series of stages in which strict control of cell proliferation, differentiation, and cell death occurs. Transgenic mice expressing human papillomavirus type 16 E6/E7 papillomavirus oncogenes in the outer root sheath (ORS) display a fur phenotype characterized by lower hair density and the ability to regenerate hair much faster than wild-type mice. Regenerating hair follicles of transgenic mice show a longer growth phase (anagen), and although bulb regression (catagen) occurs, rest at telogen was not observed. No abnormalities were detected during the first cycle of hair follicle growth, but by the second cycle, initiation of catagen was delayed, and rest at telogen was again not attained, even in the presence of estradiol, a telogen resting signal. In conclusion, expression of E6/E7 in the ORS delays entrance to catagen and makes cells of the ORS insensitive to telogen resting signals bearing to a continuous hair follicle cycling in transgenic mice.
Introduction
Proliferation and death are essential cellular processes during the development of different animal structures. Proliferation is required to produce as many cells as needed to form the tissues of a specific structure, and its inhibition allows terminal differentiation. On the other hand, cell death is essential both to eliminate cells within tissues that have to achieve the right size and form and to eliminate unnecessary tissues in the mature structure. In recent years, many signals that control proliferation, differentiation, and cell death have been characterized, several of which are shared, showing the intimate interactions between these processes (1 , 2) .
Hair follicle development (reviewed in Ref. 3 ) is an appropriate model to study how proliferation, differentiation, and cell death coordinate to give rise to a complex structure. Hair follicles originate late in embryogenesis from the epidermis by a process involving thickening and down-growth of the epithelial tissue at specific sites; the dermal mesenchyme tissue underneath, whose cells aggregate to form the dermal papilla, determines these sites. The plug of epithelial tissue formed, which is always in intimate contact with the dermal papilla, is initially constituted of what is known as the ORS,5 which will later differentiate into the cells of the hair matrix at the base of the follicle. Hair follicle growth results from differentiation and migration of cells from the hair matrix, giving rise to the inner root sheath and the shaft. After growth, the base of a hair follicle degenerates up to one-third of its maximum size. The follicle at this stage is composed of nonproliferating cells, which include the putative stem cells responsible for hair regeneration in the adult.
The hair growth cycle appears to start from a population of stem cells located in a special part of the ORS, just below the bulge (4 , 5) . Every cycle is characterized by three phases: (a) anagen, the stage of active cell proliferation during which the follicle is regenerated, and a new hair is produced; (b) catagen, the stage during which hair elongation ceases, and active cell death causes hair follicle regression; and (c) telogen, the stage during which the follicle is at rest, and cell activity is minimal. At this latter stage, the stem cells are ready to start a new cycle.
Experimental modifications of hair follicle development may contribute to the understanding of how different regulatory pathways act within the epidermis. To modify the patterns of proliferation, differentiation, and cell death in a variety of renewable stratified epithelia, including the ORS of hair follicles, we have used the K6 promoter to direct the expression of HPV-16 E6/E7 oncogenes. E7 binds and inactivates molecules, such as pRB, that are key regulators of the cell cycle and thus promotes cell proliferation (6 , 7) . On the other hand, E6 induces p53 degradation through a coupling molecule (6 , 7) . This tumor suppressor protein negatively regulates cell cycle progression by transcriptional activation of the gene coding for p21, a cyclin-dependent kinase inhibitory protein, and it also participates in the control of apoptosis (8) . Transgenic mice expressing E6/E7 oncogenes under the control of the K6 promoter show lower hair density by 2 months of age. However, hair growth after depilation was clearly faster in transgenic animals than in wild-type animals. Histological analysis shows that the hair follicle growth cycle in transgenic mice had a longer growth phase and lacked a resting phase. Furthermore, hair follicles of transgenic mice did not respond to topical applications of estradiol, a molecule capable of arresting follicles at telogen (9) . Therefore, the hair follicles of transgenic mice proceed through all stages of the growth cycle, with an extended anagen phase, and bypass the resting at telogen.
Results
Expression of E6/E7 Oncogenes in Transgenic Mice.
We used 9.8 kb of the bK6 promoter to drive expression of
E6/E7 coding sequences of HPV-16 in transgenic mice (Fig. 1)
. Five founder mice were generated, four of which showed a similar
phenotype in their fur characterized by a low hair density (see below),
which was obvious from 2 months of age. Two founders died unexpectedly;
therefore, progeny was only produced from the remaining two animals,
giving rise to the lines named Tg(bK6-E6/E7)M8 and
Tg(bK6-E6/E7)H1. Matings between wild-type CD-1 and
hemizygous transgenic mice from these two lines produced litters of
normal size, comparable that of the CD-1 wild-type strain (about 10
pups/litter), and transgene from both lines segregated in a Mendelian
fashion (Table 1)
; therefore, few or no transgenic mouse died during embryonic life, and
a single integration site was expected. Sex distribution was determined
after weaning, and among the transgenic mice produced, sex distribution
was normal within line Tg(bK6-E6/E7)M8 but abnormal within
line Tg(bK6-E6/E7)H1, in which females prevailed (Table 1)
;
in this latter case, some males may be within the population of mice
that died before weaning (see below).
|
|
. Males were predominately affected, leaving almost only
females among the group of animals more than 1 year of age
(Table 1)
. The cause of death appeared to be related to a progressive
disease associated with kidney degeneration (data not shown). More
studies are required to understand the cause of death; however, it is
unlikely that this interfered with the studies described below.
Northern blot analysis of total RNA from different tissues of mice from
line Tg(bK6-E6/E7)M8 (Fig. 2A)
and Tg(bK6-E6/E7)H1 (data not shown) was used
to determine the transgene expression pattern. In both lines,
expression was observed in tissues expected to express K6
and was highest in skin and tongue and absent in liver (Fig. 2A)
. Accordingly, combining a specific immunoprecipitation
and Western blotting, we were also able to detect the E7 protein in the
skin of transgenic mice (Fig. 2B)
. Within the hair follicle,
E6/E7 mRNA (Fig. 2C)
and the proteins encoded
(Fig. 4)
were detected in the ORS, in which K6 is normally present. The
independent transgenic mouse line with a normal fur phenotype mentioned
above did not express the transgene (data not shown), supporting the
notion that abnormal fur phenotype is due to transgene expression.
|
|
|
; E6*I transcript, considered to be the main
product from the papillomavirus (10)
, was the most
abundant. A semiquantitative RT-PCR protocol was implemented to
determine the E6/E7 mRNA levels in skin and tongue samples
of mice from different postnatal ages. E6/E7 mRNA was not
detected in newborn transgenic mice but started to be detected 1 week
after birth in both tongue and skin and remained almost constant later
in life (Fig. 3B)
; these data are in agreement with the
expected expression pattern of the mouse K6 gene in these
tissues.
Using immunohistochemistry, we determined the location of E6 and E7
proteins in the skin of transgenic mice. Fifteen days after birth, E6
protein started to be clearly detected in the ORS of hair follicles
(Fig. 4A)
, although not many positive cells were generally found. At
30 days after birth, when hair follicles are expected to be in the
early phase of the second growth cycle (see below), E6 protein was
detected mainly in the "new" growing follicle, with positive cells
apparently coming from around the base of the "old" follicle (Fig. 4, B and C)
. By 40 days after birth, many E6- and
E7-positive cells were located just below the sebaceous gland in the
region known as the bulge, which is characterized by the
attachment of a muscle fiber to the follicle (Fig. 4, D and G)
, although the viral proteins were also distributed in the
cells of the ORS along the whole length of the hair follicle (Fig. 4, E and H)
. During this growing phase, E6 and E7
proteins were also abundantly detected in the interfollicular epidermis
(Fig. 4, B and G)
. In control animals
(i.e., wild-type CD-1 mice), immunoreactivity was regularly
found in the sebaceous gland (Fig. 4, F and I)
;
therefore, the signal in this region in samples from transgenic mice
was not considered relevant.
Fur Phenotype in Tg(bK6-E6/E7)M8 Transgenic Mice.
Fur phenotype was indistinguishable between wild-type and transgenic
mice until the age of 1 month. After this, lower hair density started
to be noticeable in transgenic mice and became obvious in 2-month-old
mice. This phenotype was always more marked in mature females than
males (Fig. 5A)
and was maintained for the remainder of the animal’s
life, with an increasing degree of manifestation in older animals and
in pregnant females. Some differences in fur phenotype were observed
between lines Tg(bK6-E6/E7)M8 and
Tg(bK6-E6/E7)H1. In mice from line
Tg(bK6-E6/E7)M8, hair density was seen as a gradient, with
waves of less hair density moving rostral to caudal (Fig. 5
A,
left). In mice from line Tg(bK6-E6/E7)H1, this gradient
was not evident, and a patchy phenotype was seen instead (Fig. 5
A, right); in this case, each patch could represent
differences in hair density or growing stage.
|
. On the other hand, by 10 days,
stained hairs were completely lost in the clipped region and almost
completely lost in the unclipped region of transgenic mice, but
not in control ones (Fig. 5B)
, suggesting that the hair
follicles of transgenic mice were not at the resting stage expected in
2-month-old mice. These data are consistent with a continuous growth of
hairs in transgenic mice; therefore, we were persuaded to analyze this
possibility in more detail.
Hemizygous mice from line Tg(bK6-E6/E7)M8 were selected for
the studies described below because of their more even fur phenotype in
defined areas of the skin and longer survival than
Tg(bK6-E6/E7)H1 animals, thus allowing production of the
animals required and decreasing the probability of death during the
experiments. In the present work, we show results obtained mainly with
females because of their longer survival, but very similar data have
been achieved with males, particularly in relation to the hair growth
(see Figs. 5
and 11
).
|
. The hair follicles of transgenic and control mice behaved similarly
through the anagen phase; however, follicles of control mice started
follicle degeneration by about 18 days and reached the resting phase by
24 days, whereas follicles of transgenic mice showed no signs of
degeneration until 30 days after depilation, with few follicles with
telogen appearance. Forty-five days after depilation, wild-type
follicles were still in telogen, whereas transgenic ones were again in
the anagen phase (data not shown). Therefore, rather than affecting the
whole hair cycle, E6/E7 transgenes appear to extend the
anagen phase and prevent hairs from remaining in the resting phase. No
abnormality was observed in follicle morphology through the different
developmental stages.
|
. The first growth cycle of transgenic mice hair follicles proceeded
normally, reaching the telogen phase at about the expected time.
However, it is important to note that at 20 days after birth, wild-type
hair follicles appeared at a more advanced stage of catagen than those
of transgenic mice, suggesting that the first anagen phase was slightly
longer in transgenic hair follicles than in wild-type hair follicles
(see below). The second growth cycle appeared to proceed normally until
the end of the anagen phase, when hair follicles of wild-type mice
initiated the catagen and reached the telogen phase, but the follicles
of transgenic mice remained in the anagen phase or were just initiating
catagen. These results are in agreement with the fact that marked
differences in phenotype started to appear after the first hair growth
cycle.
|
; a similar result was obtained in the hair follicles of mice 35 days
after birth, when they were also in anagen but in the second hair
growth cycle (Fig. 8)
. Thirty days after birth, the initial stages of
anagen of the second cycle were occurring, and many PCNA-positive cells
were found in the growing hair follicle. By day 40, follicles with
catagen morphology were observed, but some PCNA-positive cells were
still detected (Fig. 8)
. In transgenic mice, hair follicles at 5 and 35
days after birth were at a similar stage of anagen as wild-type mice;
the location of PCNA-positive cells in these follicles was similar to
that observed in wild-type mice (Fig. 8)
. In contrast, at 30 days after
birth, anagen in transgenic hair follicles was at an earlier stage than
that in wild-type ones, showing abundant proliferating cells (Fig. 8)
.
This observation is in agreement with the histological data described
above, indicating that anagen of the first cycle was longer for
transgenic follicles, resulting in a delay in the initiation of the
second cycle. Additionally, at 40 days after birth, when hair follicles
of wild-type animals were initiating catagen, the PCNA staining pattern
in transgenic follicles corresponded to that observed in the normal
anagen phase (Fig. 8)
. Proliferation in hair follicles was also
evaluated at several time points after depilation (data not shown). In
general, as occurred in the experiments described above, the PCNA
staining pattern observed during the initial stages of anagen (6–12
days after depilation) was essentially similar between wild-type and
transgenic follicles; however, in agreement with an extended anagen
phase, the PCNA staining in the hair matrix and the ORS characteristic
of follicles in anagen was preserved in transgenic follicles, even when
wild-type follicles had reached the telogen phase. It is
important to note that although we were not able to detect a
significantly larger number of PCNA-positive cells in transgenic hair
follicles as compared with wild-type ones at a single similar stage,
more PCNA-positive cells are obviously produced in transgenic hair
follicles than in wild-type hair follicles because the former have a
prolonged anagen phase.
|
. During the anagen phase, cell death was detected in the
most internal cell layer of the inner root sheath and is probably
related to the formation of the hair shaft (Fig. 9C)
. Cell
death in transgenic mice hair follicles at the time of catagen
initiation in wild-type mice (18 days after depilation) was located in
regions identical to those of the hair follicles of wild-type mice in
anagen (Fig. 9, B and D)
. This pattern did not
change dramatically until the time at which catagen-like follicles were
observed in transgenic mice. At this later stage, massive cell death
was found in some transgenic mouse follicles with telogen-like
morphology, a pattern never seen in follicles of wild-type mice (Fig. 9E)
. At the beginning of a new cycle, cell death was not
detected, and only proliferating cells were observed.
|
. As expected, hair growth was blocked by 17ß-estradiol in wild-type
mice, an effect that was obvious 39 days after treatment. In contrast,
the hairs of transgenic mice grew much faster, and no response to
17ß-estradiol was observed. These data are in agreement with
E6/E7 transgenes being able to skip a hair follicle resting
signal at telogen.
|
. Differences in hair length were observed (0.96 ± 0.12 cm in
wild-type mice versus 0.72 ± 0.09 cm in transgenic
mice) that might contribute to the fur aspect.
|
or second telogen phase (data not shown). Therefore, lower hair
density may be due to increased shedding occurring mainly at the
telogen phase. Discussion
We have expressed HPV-16 E6/E7 oncogenes in the ORS of
hair follicles using the K6 promoter. K6 is
expressed in several stratified epithelia such as those of the skin,
tongue, esophagus, forestomach, and cervix (11)
. In the
skin, K6 expression is restricted mainly to the ORS of the
hair follicle (12)
, and the protein product has been
initially observed in this tissue by 2 weeks after birth
(13)
, although we detected significant K6
promoter activity before this time (Fig. 3)
. Expression in
suprabasal layers of the epidermis is detected only after injury or in
response to activators such as phorbol esters or retinoids
(14, 15, 16, 17, 18, 19)
. In all cases, expression of K6 is
coincident with hyperproliferative cells and is also
up-regulated in tumor cells derived from epithelial tissues (20
, 21)
. Therefore, the K6 promoter is a suitable tool to
direct expression of genes to proliferative layers of epithelia in
transgenic mice (22, 23, 24, 25, 26)
. Thus, the effects of a transgene
with its expression directed by the K6 promoter on hair
growth should be caused by direct activity of transgene in cells of the
ORS and not indirectly through its activity in cells of the epidermis.
In this sense, use of the K6 promoter may be more
advantageous than other promoters that share constitutive specificity
to the ORS and suprabasal layers of the epidermis (e.g.,
K14 promoter; Ref. 27
).
It has been proposed that every hair regeneration cycle starts from the
activation of a slowly dividing cell population of stem cells located
in the ORS (4
, 5)
. Electron microscopy data suggest that
these cells do not contain intermediate filaments (4
, 5)
.
If the hypothesis for hair regeneration involving these stem cells is
correct, then our data indicate that the K6 promoter should be active
in this slowly dividing cell population. Interestingly, E6 and E7
antibodies detected cells that are located in the zone proposed to
contain this stem cell population in the growing follicle (Fig. 4)
.
Furthermore, at the initial growth stages, oncoproteins are abundantly
present in the newly forming follicle, with positive cells apparently
coming from a region close to the base of the "parent" follicle
(Fig. 4, B and C)
. Although the viral oncogenes
may not be abundantly expressed in the quiescent stem cells, oncogene
expression appeared to be induced when follicle regeneration was
initiated. Therefore, E6 and E7 oncogenes in our
transgenic mice may be acting directly in the stem cell population
responsible for hair follicle regeneration. However, it still remains
to be demonstrated whether the more differentiated cells of the ORS,
which express K6, are capable of initiating follicle regeneration.
E6 and E7 are both able to immortalize cells in culture (28, 29, 30, 31) , but the oncogenes have different effects in transgenic mice. For example, in the eyes of transgenic mice with directed expression of E7, E6, or both, E7 can promote apoptosis of lens (32) and photoreceptor (33) cells, an effect that is at least partially dependent on p53 expression. E6 expression, on the other hand, reduces natural cell death in developing lenses and inhibits terminal differentiation of lens fiber cells (32) . However, E7 can promote proliferation and tumor formation when p53 is absent (33 , 34) or E6 is coexpressed (32) , probably by reducing its apoptotic activity. As described below, most of the effects on hair follicle growth observed in the present work can be interpreted as being due to an interference by E6/E7 oncogenes with the control of cell proliferation rather than to a prevention or promotion of cell death. Therefore, our data are in agreement with the expression of both oncogenes in transgenic mouse hair follicles.
In general, cell differentiation during hair follicle regeneration
appears to be unaffected in our transgenic mice. The previous
conclusion is derived from the presence of the major hair types with
typical morphology (data not shown) and, in general, normal proportion
awl/auchene versus zigzag in transgenic mice (Table 2)
.
Furthermore, the expression pattern of cytokeratins in transgenic hair
follicles displayed using a broad spectrum antibody resembled the
pattern observed in wild-type animals (data not shown). Similarly, the
patterns of proliferation (Fig. 7)
and cell death (Fig. 8)
were not
greatly affected, and hair follicle morphogenesis along the growth
cycle appeared normal. Therefore, the effects of E6/E7
oncogenes should be regulatory, which could change the initiation or
end of phases during the hair growth cycle. Nonetheless, shorter hair
in transgenic mice might be due to delayed differentiation resulting
from the promotion of cell proliferation by the oncogenes during the
extended anagen phase.
From the growth regulatory point of view, the hair follicle has two relevant tissues, the ORS and the dermal papilla (3) . The ORS is an epithelial layer of cells in continuity with the epidermis and may be equivalent to the epithelial tissue invaginated early during primary hair follicle formation. It is from the ORS that the hair matrix forms and other tissues, such as those giving rise to the hair shaft, are subsequently originated. At the telogen phase, the ORS is the most important tissue left after catagen, because the putative stem cells reside in this compartment.
The dermal papilla and its precursor cells are key tissues producing relevant regulatory molecules. For instance, hair formation is initiated by a mesenchymal signal (first dermal message) acting on the epidermis, and proliferation of hair matrix cells is stimulated later by a signal from the dermal papilla (second dermal message). These two kinds of signals seem to be used continuously during cycling, with the first signal allowing the follicle to exit from telogen and begin the formation of a new hair follicle and the second signal regulating the anagen phase. Therefore, cross-talk between the dermal papilla and the ORS should be expected during hair follicle development. Disturbances in this communication should result in serious consequences on hair growth, as might be happening in the follicles of our bK6-E6/E7 transgenic mice (see below).
Two critical regulatory points can be viewed during the hair follicle growth cycle: (a) initiation of growth; and (b) initiation of regression (also the end of the growth phase). Initiation of growth could result from both a decrease in negative signals and an increase in positive signals. Resting at telogen would be expected to be due to a higher concentration of negative signals and a lower concentration of positive signals. Cells at telogen are in a quiescent state; consequently, a transition from stages G1 to G0 of the cell cycle would be expected to establish the resting phase (35) . Regression could also require two types of signals, one inhibiting proliferation and/or differentiation of the ORS and hair matrix cells, and the other activating cell death in the lower region of the grown hair follicle.
Our results suggest that E6/E7 oncogenes were able to
interfere with the G1
G0
transition in cells of the ORS at telogen, thus explaining why these
cells could not respond to negative signals (i.e.,
estradiol) and why the long resting phase is never established; it is
likely that in the presence of these oncogenes, positive signals to
initiate growth (e.g., a first dermal-like signal) are not
required. Similarly, our data indicate that E6/E7 oncogenes,
by blocking the response to inhibitors of cell proliferation and
consequently allowing the G1S transition of
the cell cycle in cells of the ORS (e.g., as the second
dermal-like signal may be doing), were able to extend the growth phase
(i. e., anagen), thus delaying the initiation of follicle
regression. We propose that a balance between signals promoting cell
proliferation and cell death defines the time at which catagen starts.
How could the effects on the hair follicle cycle by the E6
and E7 oncogenes produce lower hair density and more rapid
hair growth? Cell death at the time of follicle degeneration may have
been affected by the E6/E7 transgene as shown (Fig. 9E)
, but fur phenotype is more likely to be associated with
the uncovered inability of hair follicles of transgenic mice to enter
the long-lasting resting phase characteristic after the second cycle of
hair growth. Under this view, the apparent faster hair growth in
transgenic mice as compared with wild-type mice after depilation could
result from incomplete plucking, because the hairs of some already
growing follicles might be under the skin surface at the time of
plucking in transgenic mice. Alternatively, because transgenic hair
follicles were more likely in the anagen phase at the time of plucking,
hair differentiation could occur earlier in transgenic mice follicles
than it does in wild-type mice follicles, which were at telogen at the
time of plucking. On the other hand, lower hair density in transgenic
mice could be due to easier shedding, because hair holding by the club,
a structure specifically formed at the telogen phase, could never be
established in transgenic mice. This latter hypothesis is in agreement
with the easier hair plucking precisely at the telogen phase in
transgenic mice as compared with wild-type animals (Fig. 11)
. Recently,
mutant mice were generated that are unable to form the club and then
attach the hair to the follicle (36)
. These mice lose
their hair in bands moving in an antero-posterior direction just after
the first growth cycle. The mice characterized in the present work
appear to behave similarly to those mutant mice in some respects,
although it is not expected they would lose their hair completely due
to the continuous growth promoted by the oncogenes; bands of growth
were not regularly seen, but sometimes they were conspicuous
(Fig. 5A)
. Grooming could influence this atypical hair
shedding but could not define the cause of the fur phenotype because it
was not identical in the two transgenic lines analyzed.
Some signaling molecules acting during hair follicle formation and cycling have been characterized as peptide growth factors or molecules such as steroid hormones or retinoic acid (37 , 38) . Nonetheless, a specific role for these molecules has been established in only a few cases. Steroid hormones, such as estradiol and testosterone, have been shown to act indirectly through the dermal papilla to arrest hair follicles at telogen (9 , 39) . Because these molecules act negatively on growth, it is possible that they repress the expression of those positive signals synthesized by the dermal papilla. In our transgenic mice, we propose that the inability to respond to estradiol is because, due to the expression of E6 and E7 oncogenes, initiation of growth is independent of these positive signals, and consequently arrest at telogen is never established.
On the basis of the expression patterns observed in hair follicles, candidate molecules regulating growth could be members of the transforming growth factor ß (40 , 41) and the FGF (42 , 43) gene families. Overexpression of members of these families blocks hair growth and/or formation (44 , 45) . Relevant to the present work is the phenotype observed in mice lacking fgf-5, which have a phenotype identical to the angora strain of mice and characterized by abnormally long hair (46) . Histological analysis of these mice revealed an extended anagen phase, which the authors interpreted as the cause of the long hair. Interestingly, our transgenic mice also have an extended anagen phase, but hair length was even shorter than normal. Therefore, an extended anagen phase seems insufficient to cause an increase in the hair length; consequently, fgf-5 might have an additional function in controlling hair growth. Alternatively, we might not have observed long hairs because hair growth is delayed by the effects of oncogenes on cell differentiation, as mentioned above.
A contrasting phenotype to the one displayed by our bK6-E6/E7 transgenic mice was the presented in a recent report (47) . Authors show that transgenic mice that express bcl-xL in the ORS have hair follicle growth cycles with a shorter anagen phase and a longer telogen phase when compared with those of wild-type mice. Although the authors attribute this hair growth cycle behavior to increased survival of ORS cells able to express fgf-5, they do not show a direct comparative measurement of cell death. Furthermore, crosses between K14-bcl-xL transgenic and fgf-5-deficient mice display a similar effect on hair growth cycle when compared with the original fgf-5-deficient mice, hampering an obvious conclusion from these studies. Recent reports have shown that Bcl-2 family members, in addition to their antiapoptotic function, have the ability to delay entry to the S phase of cell cycle (48 , 49) . Therefore, it is possible that the antiproliferative activity of Bcl-xL in hair follicles of the K14-Bcl-xL transgenic mice shortens anagen and makes the exit from telogen difficult. This latter interpretation would be in agreement with the data presented here, which highlight the critical role of cell cycle regulators as major determinants of hair growth cycle dynamics.
Materials and Methods
Animal Procedures.
Mouse strain CD-1 was used in this study. Transgenic mice were produced
by pronuclear injection of recombinant DNA following standard protocols
(50)
. Depilation was performed by wax stripping of about 3
cm2 of mid-dorsal skin on anesthetized mice. A
regular permanent hair stain was used to determine hair growth and
shedding; in some cases, hair staining was performed after clipping.
Estradiol (50 µM; Sigma) was topically applied (50
µl/cm2) every 2 days on a dorsal skin area with
hair previously clipped and stained. Hair types were analyzed following
the characteristics described by Sundberg and Hogan (51)
.
Construction of Recombinant DNA.
E6/E7 open reading frames were amplified by PCR using as a
template a plasmid carrying the complete genome of HPV-16. A
ClaI recognition site was added to the 5' end of the
specific E6/E7 sequences (nucleotides 80–101 and 864–886 for E6/E7–5' and E6/E7–3' oligonucleotides, respectively;
GenBank accession number U89348). The oligonucleotides had the
following specific sequence: (a) E6/E7–5',
5'-CTCATCGATTTTTATGCACCAAAAGAGAACTG-3'; and (b) E6/E7–3',
5'-CTCATCGATTACCTGCAGGATCAGCCATG-3'. The PCR product cut with
ClaI was inserted into the ClaI site of
pBluescript KS+ (Stratagene), and the E6/E7 fragment of
several resulting recombinant plasmids was sequenced. One plasmid
(pKS-E6/E7) with the correct desired sequence was selected to be used
as a source of the E6/E7 ClaI fragment. This
latter fragment was inserted into the unique ClaI site of an
expression vector carrying 9.5 kb of the bK6 (also called
CIV*) promoter and 5' upstream regulatory sequences (44)
.
The resulting plasmid (bK6-E6/E7) was the source of the
NotI 13.8-kb vector-free fragment used for pronuclear
injection.
RNA Purification, Northern Blot, cDNA Synthesis, and PCR
Procedures.
RNA was purified from tissue of either wild-type or transgenic animals
using a modified version of the method published by Chomczynski and
Sacchi (52)
. Briefly, transgenic or CD-1 wild-type mice
were sacrificed by cervical dislocation, and multiple organs and
tissues were dissected and frozen immediately. Tissue fragments were
crushed in liquid nitrogen, lysed cells were mixed with 1 ml of 6
M urea/3 M LiCl solution, incubated at -20
°C overnight, and centrifuged at 10,000 x g for 10
min, and the pellet was resuspended in 500 µl of GuSCN solution [4
M guanidinium isothiocyanate, 25
mM sodium citrate (pH 7.0), 0.5% sarcosyl, and
0.1 M 2-mercaptoethanol]. Two
M sodium acetate (1:10; pH 4.0), phenol (1:1),
and chloroform (1:20) were then added. Chloroform is a 24:1 mixture of
chloroform and isoamyl alcohol. The suspension was vortexed for 1 min,
kept on ice for 15 min, and centrifuged at 10,000 x g
for 20 min. The aqueous phase was collected and precipitated with an
equal volume of isopropanol at -20°C. The pellet was resuspended in
150 µl of GuSCN and reprecipitated. The RNA was washed with 75%
ethanol and dissolved in 100 µl of water. For Northern blot analysis,
samples (10 µg of total RNA) were loaded on a 1%
agarose-formaldehyde gel, electrophoresed, transferred to
Hybond-N+ membrane (Amersham), and hybridized
with a 32P-labeled random-primed DNA probe
corresponding to the E6/E7 coding region using standard
protocols. For RT-PCR, 1 µg of total RNA was used for cDNA synthesis
using Moloney murine leukemia virus reverse transcriptase (Life
Technologies, Inc.) in a 30-µl final reaction volume according to the
manufacturer’s recommendations. Three µl of the cDNA reaction
mixture were routinely used in a PCR reaction containing 2.5
units of Taq DNA polymerase (Roche), 200 µM
deoxyribonucleoside-5'-triphosphates, and 1.86 mM
MgCl2 in the buffer provided and 30 pmol each of
forward and reverse primers; all reactions were carried out in a
programmable heating block (Hybaid). The PCR protocols for HPRT (a
housekeeping enzyme), E6, and E6/E7 were 93°C for 1 min,
57°C for 1 min, 72°C for 1 min for 20, 25, and 30 cycles,
respectively; in all cases, PCR was ended with 10 min at 72°C.
Oligonucleotide primers were as follows: (a) HPRT,
CCTGCTGGATTACATTAAAGCACTG (forward) and GTCAAGGGCATATCCAACAACAAAC
(reverse); (b) E6 and E6/E7,
TTTTATGCACCAAAAGAGAACTG (forward); (c) E6,
GTATCTCCATGCATGATTACAG (reverse); and (d) E6/E7,
TACCTGCAGGATCAGCCATG (reverse). All PCR products were of the expected
size, and their identity was confirmed by digestion with several
restriction endonucleases.
In Situ Hybridization.
Skin fragments of transgenic mice were fixed in 4% paraformaldehyde in
PBS (pH 7.0) for 18–24 h at 4°C, dehydrated in ethanol, and embedded
in paraffin; 5-µm sections were mounted on
poly-L-lysine-treated slides. After paraffin removal and
rehydration, the sections were pretreated with proteinase K (1 µg/ml)
for 7 min at room temperature and washed sequentially in PBS, 0.2%
glycine in PBS, PBS, TEA (0.1 M triethanolamine and 0.25%
acetic anhydride), and PBS, for 3 min each. Sections were dehydrated in
ethanol and dried for 1–2 h at room temperature. Prehybridization was
performed by incubation in 100 µl of hybridization solution [4x SSC
(1x SSC = 150 mM NaCl, 15 mM sodium
citrate (pH 7.0), 10% dextran sulfate, 1x Denhardt’s solution
(0.02% Ficoll 400, 0.02% polyvinylpyrrolidone, and 0.02% BSA), 2
mM EDTA, 50% deionized formamide, and 0.5 mg/ml herring
sperm DNA] at 42°C for 2 h. E6/E7 riboprobe was
synthesized from the pKS-E6/E7 plasmid using DIG-labeled UTP
and T7 RNA polymerase (sense) or SP6 RNA polymerase (antisense). Two to
5 µl of DIG-labeled RNA (200 ng/ml, final concentration) were added
to 100 µl of hybridization solution, loaded on the sample, and
incubated at 42°C for 16 h. The slides were washed once in 2x
SSC at 37°C for 5 min, washed three times in 60% formamide/0.2x SSC
at 37°C for 5 min, and washed twice in 2x SSC at room temperature
for 5 min. Detection of hybridized DIG-labeled RNA was performed with
an alkaline phosphatase-conjugated anti-DIG antibody (Roche) and
developed with the nitroblue tetrazolium/5-bromo-4-chloro-3-indolyl
phosphate substrate solution in the presence of 2 mM fresh
levamisole for 16 h in the dark at room temperature.
Immunoprecipitation/Western Blot and Immunohistochemistry
Procedures.
Detection of E7 protein was performed by a combined
immunoprecipitation/Western blot procedure. Total protein from mouse
skin was extracted in radioimmunoprecipitation assay buffer as
described previously by Arbeit et. al. (53)
.
Briefly, 500 µg of total protein were immunoprecipitated with rabbit
polyclonal antibody C6563 (generated in our laboratory).
Immunoprecipitated protein was separated in a 12% SDS-PAGE gel and
transferred to Immobilon P membranes (polyvinylidene difluoride; 0.45
µm; Millipore, Bedford, MA) as described by Towbin et al.
(54)
. Filters were blocked with Tris-buffered saline [20
mM Tris-HCl (pH 7.4) and 150
mM NaCl] containing 5% skim milk for 2 h
at 4°C and then incubated with the same rabbit polyclonal antibody
used for the immunoprecipitation. Specific antibody binding was
detected by using horseradish peroxidase-linked goat antirabbit IgG
(Dako) and visualized by the enhanced chemiluminescence detection
system (Amersham Pharmacia Biotech). For detection of viral proteins by
immunohistochemistry, HPV-16 E7 and HPV-16 E6 monoclonal antibodies
(1:100 dilution, Santa Cruz Biotechnology) were used. Tissues were
fixed and embedded in paraffin as indicated for in situ
hybridizations. For immunohistochemistry, the antibody was incubated
for 2 h with the sample and detected using the Vectastain Elite
ABC Kit (Vector Laboratories catalogue number PK 6102) according to
manufacturer’s instructions.
Histological Analysis.
Tissues were fixed in 4% paraformaldehyde and embedded in paraffin.
Sections of 5–10 µm were stained with H&E according to standard
procedures. Photography was performed on an Axioscope microscope
(Zeiss, Germany) using Kodacolor 100 film.
Detection of Cell Proliferation and Death.
Four percent paraformaldehyde-fixed, paraffin-embedded, 10-µm
sections of skin fragments were used to determine the pattern of cell
proliferation and death. An antibody specific for PCNA
(Vector-Novocastra) was used to visualize proliferating cells following
a standard immunohistochemistry procedure. The terminal
deoxynucleotidyl transferase-mediated nick end labeling
(55)
method for detection of fragmented DNA using
fluoresceinated dUTP (Roche) was performed according to the
manufacturer’s instructions. Serial sections were analyzed by
Nomarski’s optic using a Diaphot Nikon microscope or by confocal
microscopy using a MRC-600 confocal laser scanning system equipped with
a krypton/argon laser (Bio-Rad) coupled to an Axioscope microscope.
Representative prints are shown.
Acknowledgments
We are grateful to Elizabeth Mata, Sergio González, Graciela Cabeza, Paul Gaytán, Eugenio López, and Xochitl Alvarado for technical assistance; Melanie Simpson for careful reading of the manuscript; and Rob Smart for advice on hair type analysis.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 Supported by the Consejo Nacional de Ciencia y
Tecnología (Grants 1163 M/9209 and 31730-N), Programa de
Naciones Unidas para el Desarrollo (pnud/mex/93/019) and
Dirección General de Asuntos del Personal Académico de la
Universidad Nacional Autónoma de México (IN 208697).
P. G. acknowledges The International Union Against Cancer for a
sabbatical fellowship sponsored by Novartis. 
2 These authors contributed equally to this
work.
3 Present address: Instituto de Fisiología
Celular, Universidad Nacional Autónoma de México, Ciudad
Universitaria, México, D.F., México.
4 To whom requests for reprints should be
addressed, at Departmento de Genética y Fisiología
Molecular, Instituto de Biotecnología, Universidad Nacional
Autónoma de México, AP 510-3, Cuernavaca, Mor. 62250,
Mexico. Phone: 52-5-622-7636/7631; Fax: 52-73-17-2388;
E-mail: covs{at}ibt.unam.mx
5 The abbreviations used are: ORS, outer
root sheath; HPV, human papillomavirus; K6, keratin 6; bK6, bovine K6;
RT-PCR, reverse transcription-PCR; PCNA, proliferating cell nuclear
antigen; FGF, fibroblast growth factor; HPRT, hypoxanthine
phosphoribosyl transferase; DIG, digoxygenin.
Received for publication 5/18/00. Revision received 6/13/00. Accepted for publication 8/31/00.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cell Growth & Differentiation |
