| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Pharmacology and Therapeutics, Grace Cancer Drug Center, Roswell Park Cancer Institute, Buffalo, New York
Correspondence: Dr. Erica Berleth, Grace Cancer Drug Center, Roswell Park Cancer Institute, Elm and Carlton Streets, Buffalo, NY 14263. E-mail: erica.berleth{at}roswellpark.org
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: mammary • apoptosis • TNF • lymphoid • T cell
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
is one such factor. It has regulatory functions
throughout the body, some of which are connected with down-regulation
of cell numbers, whereas others are involved in cell growth,
maturation, and survival. The mechanism by which this exquisitely
selective reactivity to TNF-initiated signals is accomplished is not
understood. TNF is a primary mediator of immune regulation and inflammatory responses. TNF and both TNF receptors (TNFRs) are expressed on essentially all cells of the immune system. The study of signal cascades initiated by TNF-TNFR interaction has yielded considerable information on intermediary components of the signal pathways that lead to apoptosis, proliferation, maturation, and survival. TNF induces T-cell colony formation and activation of cytotoxic T cells but also is the causative agent in induction of apoptosis in mature peripheral T cells after activation [1 ]. Studies in TNF-null mice provide evidence that TNF is essential for T- and B-cell splenic organization [2 ] and that follicular dendritic cells have critical roles in this organization [3 ]. Dendritic-cell proliferation, maturation, activation, and survival are dependent on TNF, and mature dendritic cells have central roles in innate and adaptive (both T- and B-cell) immunity [3 ]. Although there is little evidence that TNF has a role in negative selection in the thymus, it has been linked to other forms of selective thymocyte deletion [4 ], and it also induces thymocyte proliferation [5 ].
Mammary epithelial cells (MECs) express TNF mRNA and protein [6 , 7 ] as well as mRNA for both TNFRs [7 ]; stromal breast cells may also express these proteins [8 , 9 ]. Moreover, Ip and coworkers observed in their in vitro rat MEC growth and differentiation model that TNF can induce extensive branching morphogenesis and can substitute fully for epidermal growth factor in the stimulation of growth of normal MECs but not in the functional differentiation as assessed by casein production [10 11 12 ]. Furthermore, in the presence of optimal epidermal growth factor, TNF inhibits this differentiation. It was also demonstrated that there is specific regulation of the expression of TNF in both epithelial and stromal components and of each of the TNFRs in MECs during the rat mammary gland developmental stages of puberty, pregnancy, lactation, and postweaning apoptotic regression [7; M. M. Ip, unpublished results]. Observations in TNF-null mice confirmed the requirement for TNF in normal mammary ductal proliferative branching morphogenesis during puberty (Ip, unpublished results). Some data indicate that human breast cancer tissue may express higher levels of TNF and TNFRs than does normal or benign mammary tissue [8 , 9 ]. Yet, it has also been reported that TNF has preferential cytotoxicity against malignant versus nonmalignant cells from human breast cancer biopsies [13 ].
A novel TNF-inhibitory protein (TIP)-B1, has been identified, purified,
and characterized from cytosolic extracts (100,000 x g
supernatant) of TNF-treated human fibroblasts, and a partial TIP-B1
cDNA clone, tip-sn, has been obtained [14
,
15
]. A schematic depiction of the information previously
reported [14
] on the molecular characterization of
TIP-B1 is shown in Figure 1
. The 
27-kDa TIP-B1 protein, with an isoelectric point of
4.5, was isolated from the cytosolic fraction of TNF-stimulated
human fibroblasts. After a partial enzymatic (Lys-C) hydrolysis of the
purified protein, four peptides were purified and sent for sequence
analysis. One peptide did not yield any sequence information; the other
three peptides were designated A, B and C. Amino acid sequences of the
indicated lengths were obtained and found to be unique. Using this
information and standard molecular biochemical techniques, a reverse
transcription (RT)-PCR product was produced and used to probe a cDNA
library from TNF-treated human fibroblasts. A 783-nucleotide clone was
identified, sequenced, and designated tip-sn. Database
searches indicated that tip-sn is homologous to 102 human
expressed-sequence tags; none contain any additional information 5' of
tip-sn. Clone tip-sn has an open reading frame
that predicts a 14-kDa protein including the three peptides in tandem.
After cloning of the tip-sn insert in an inducible
prokaryotic expression vector, a 14-kDa recombinant protein (rTIP-B1p)
was produced. The original protein, TIP-B1, is unique based on both the
sequence of the three internal peptides (comprising 51 amino acids) and
the nucleotide sequence of the tip-sn cDNA clone.
TNF-sensitive cells, when exposed to TIP-B1 prior to the addition of
TNF, are completely protected from TNF-induced lysis. Thus, the
addition of TIP-B1 to cells in culture for >4 h effectively made them
resistant to concentrations of TNF that would otherwise cause induction
of cytolysis (including apoptosis). Incubation of TIP-B1 with TNF for
up to 24 h, however, did not inhibit the ability of TNF to induce
cells to undergo lysis, and cells protected by TIP-B1 bind TNF to the
same extent as TNF-sensitive (i.e., unprotected) cells. These data,
together with results from protein database searches indicating that
TIP-B1 is unique, demonstrated that TIP-B1 was not a soluble TNF
receptor, nor an anti-TNF antibody, nor a protease which degrades TNF,
nor any of the other proteins previously reported to be involved in
resistance to TNF. Nevertheless, TIP-B1 isolated from the cytosol
(100,000 x g supernatant) functions when added
exogenously to cells [14
, 15
]. TIP-B1 has
been shown to be constitutively expressed by cultured human dermal
fibroblasts and a panel of tumor cell lines [14
,
16
], and its subcellular distribution in fibroblasts was
examined by two differential centrifugation/extraction methods
[16
]. Treatment with low, nontoxic concentrations of
TNF, which renders cells insensitive to cytotoxic concentrations of
TNF, induces increased expression of TIP-B1 in the fibroblasts
[16
] and in some tumor cells (E. S. Berleth,
unpublished data).
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cytotoxicity assay
Target MCF7 cells were seeded at 104 cells/50
µL/well, with 3 or 6 wells per treatment group, in 96-well plates
(Costar; VWR Scientific, West Chester, PA). The cells were allowed to
adhere, precultured for 18 h with medium only (control), with
added low concentrations (0 to 100 ng/mL) of recombinant human TNF (0.4
ng per U; Asahi Chemical Industry Co., Ltd., Tokyo, Japan), or with
purified TIP-B1 [14
] and then cultured in medium alone
or medium containing a high concentration of TNF (400 ng/mL) for
72 h. The relative density of the cells remaining in the wells was
determined in a standard 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl
tetrazolium bromide assay as described previously [17
].
When appropriate, the percent protection was calculated as follows:
percent protection = [(mean absorbance of test wells - mean
absorbance of TNF wells)/(mean absorbance of control wells - mean
absorbance of TNF wells)] x 100%. Test wells contained cells
cultured during the protection-induction period with medium containing
agents to be tested for a protective effect and then cultured with TNF
(400 ng/mL) for the 72-h cytolysis period. TNF wells contained cells
grown in the presence of medium for the protection-induction period and
in medium containing TNF (400 ng/mL) during the 72-h cytolysis period.
Control wells were those containing cells cultured in the presence of
medium only for both periods.
Preparation of cell samples for Western blotting
Rat inguinal mammary gland tissue (mammary gland six) was
obtained from female Sprague-Dawley CrL:CD(SD) IGS BR rats purchased
from Charles River (Wilmington, MA). Other rat tissues surveyed were
from female Fischer F344/NHsd rats purchased from Harlan Sprague Dawley
(Frederick, MD). One Western blot of rat tissue and one Western blot of
human tissue were purchased from Imgenex Corp. (San Diego, CA).
Whole-cell lysates
Fresh minced tissues obtained from rats, as well as isolated rat
mammary stromal cells (MS), were sonicated in sodium dodecyl
sulfate-polyacrylamide gel electrophoresis sample buffer without dye or
reducing agent but with the addition of 100µM phenylmethylsulfonyl
fluoride, 40µg/mL of leupeptin, and 1µg/mL of aprotinin. This
lysate was then boiled for 10 min and centrifuged (microcentrifuge,
13,000 x g) for 5 min, and the soluble fraction was
used for Western blot analyses. The MEC extract was prepared using
Trizol (GIBCO, Grand Island, NY) as described previously
[7
].
Cytosolic extract
Cytosolic extracts (100,000 x g supernatants)
from MCF7 cells were prepared as previously described for human
fibroblasts [14
].
Protein quantitation
Protein concentrations were determined using either the Bio-Rad
DRC reagent (Bio-Rad, Hercules, CA) or the bicinchoninic acid BCA
protein assay reagent (Pierce, Rockford, IL).
Preparation of polyclonal antisera
The preparation and characterization of antisera against
purified native TIP-B1 and rTIP-B1p have been described
[14
]. The same inoculation procedure used to generate
the rabbit anti-rTIP-B1p antiserum was used to generate a polyclonal
antiserum against a peptide encompassing the last 15 amino acids
encoded by tip-sn, i.e., the 15-amino-acid carboxy terminus
of the predicted rTIP-B1p protein (EAVEQNTLQEFLKLA)
[14
]. The peptide was synthesized and coupled to keyhole
limpet hemocyanin for injection at the Roswell Park Biopolymer
facility. This antiserum was affinity purified by standard peptide
affinity column techniques and was designated anti-carboxy-15-mer
antibody.
Western blot analysis
Western blot analyses were performed on protein samples (10 to
20 µg/sample) separated by sodium dodecyl sulfate-12.5%
polyacrylamide gel electrophoresis and were developed using one of the
two polyclonal antisera (1:1,000 dilution) or the affinity-purified
antibody (0.1 or 0.2 µg/mL) as indicated. Peptide competition
experiments used 0.25 to 0.5 mg of peptide in 4 mL of diluted antibody.
Films after chemiluminescent detection were scanned on a computing
densitometer model 300A (Molecular Dynamics, Sunnyvale, CA) using
ImageQuant (Molecular Dynamics) and PowerPoint (Microsoft) software for
scanning and presentation. Two blots, one of rat and the other of human
protein, were purchased from Imgenex Corp. (San Diego, CA) and
developed, as above, using 0.2 µg/mL anti-carboxy-15-mer antibody.
Northern blot analysis
The human tissue RNA blot was purchased from Invitrogen
(Carlsbad, CA) and was probed as per manufacturer’s instructions. The
Northern blot analysis of rodent RNA was performed as described in ref. 14
. The mouse J774A.1 macrophage cell line was cultured with 10 ng/mL
of lipopolysaccharide for 3 h, and total RNA was isolated by
Trizol. Trizol was also used to extract total RNA from freshly isolated
rat MECs [18
] and rat MS cells [19
] at
passage 3 to 4. In all cases the autoradiographs were scanned as
described above for Western blots.
Immunohistochemical analysis
Tissues were fixed in 3.7% (v/v) phosphate-buffered formalin
prior to dehydration and paraffin embedding. Sections (5 µm thick)
were deparaffinized and rehydrated, quenched in 3%
H2O2 for 30 min, and washed (all washes were in
phosphate-buffered saline). The sections were treated with 20 µg/mL
of proteinase K for 10 min, washed, blocked with 0.03% casein for 30
min, and incubated in anti-carboxy-15-mer antibody at 10 µg/mL or the
same concentration of control rabbit immunoglobulin G (IgG) (Jackson
ImmunoResearch Laboratories, West Grove, PA) in phosphate-buffered
saline for 1 h. The sections were washed and developed using a
biotinylated donkey anti-rabbit IgG (Jackson), streptavidin, and
diaminobenzidene chromogen. They were counterstained in hematoxylin.
They were dehydrated through a series of graded alcohols and cleared in
Histoclear, and coverslips were applied. All procedures were performed
at room temperature.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Protection of MCF7 cells from TNF-induced cytolysis
Protection by pretreatment with low, nontoxic concentrations of TNF
As has been reported for other cell lines [14
,
16
], MCF7 cells cultured in the presence of low
concentrations of TNF (5–100 ng/mL) for 18 h were protected, in a
concentration-dependent manner, from the cytolysis induced by a higher
concentration (400 ng/mL) of TNF, during a 72-h lytic assay (Fig. 2A
).
|
Western blot analyses of TIP-B1 protein expression in MCF7 cells
To determine whether the TIP-B1 mRNA in MCF7 cells was translated,
Western blot analysis was performed using MCF7 lysates. As indicated in
Materials and Methods, three polyclonal antisera have been generated:
anti-native TIP-B1, anti-rTIP-B1p, and anti-carboxy-15-mer antisera.
The anti-carboxy-15-mer antibody was subsequently affinity purified. As
shown in Figure 3
, antibodies from all three reacted with a 27-kDa protein in
lysates of MCF7 cell lines (lanes 1–3). A 29-kDa protein was also
detected by all three. Competition with the carboxy-15-mer peptide
ablated reactivity of the affinity-purified anti-carboxy-15-mer
antibody with the 27-kDa protein and partially blocked reactivity with
the 29-kDa protein (lane 4). The previously described
[14
] reactivity of the antisera raised against the
purified native TIP-B1 and against rTIP-B1p with a 34-kDa protein
(lanes 1 and 2) was not seen with the anti-carboxy-15-mer antibody
(lane 3). It is not clear whether the 34-kDa immunoreactive protein was
a posttranslationally modified TIP-B1, a TIP-B1-related protein, or an
unrelated protein. All subsequent data from immunohistochemical and
Western blot analyses, therefore, were generated using the
anti-carboxy-15-mer antibody, which did not react with a 34-kDa protein
in the extracts from MCF7 cells.
|
Northern blot hybridization of tip-sn insert to human
and rodent tissue RNA
As had been reported for MCF7 cells [14
], the
783-nucleotide clone tip-sn insert identified a 1.1-kb RNA
fragment in three rodent samples (Fig. 4A
): the mouse macrophage J774 cell line (lane 1) and rat mammary
epithelial (lane 2) and rat MS (lane 3) cells. Using a purchased RNA
blot (Invitrogen), it was found that the tip-sn probe also
hybridized to a 1.1-kb RNA in all human tissues examined (Fig. 4B)
except pancreas (lane 6). Hybridization with RNA from skeletal muscle
was very weak but detectable (lane 8).
|
39-kDa protein was detected
in the extracts of heart and kidney tissues (Fig. 5A
, lanes 1 and 4,
respectively) and faintly in those of several other tissues (i.e.,
lanes 2, 3, 5, 6, 7, 9, and 10).
|
22 kDa [Fig. 5B
and 5C
, lane 1 (both blots)]. No reactivity
was seen with proteins in the extract from small intestine tissue in
either case, and, although not shown in Figure 5A
, in a single
experiment, rat small intestine tissue was also negative for TIP-B1
immunoreactive proteins by our procedure. The Western blot of rat
proteins (Fig. 5C)
was purchased to allow direct comparison between
those blots prepared in the laboratory (Fig. 5A) and ones obtained from
the commercial source of the blot of human proteins. Similar patterns
of 27-kDa-protein reactivity were seen with blots of both rat kidney
and brain tissues (Fig. 5A and 5C
, respectively), suggesting that
comparable results were obtained. By our procedure, however, the
predominant immunoreactive protein in rat heart extract was 39 kDa
and in rat brain was 27 kDa, whereas the immunoreactive proteins from
the same tissues on the commercial blot were 27 and 22 kDa in size,
respectively. Such differences could arise from use of different
strains of rats or different protein extraction procedures. Since
details on the commercial product are limited, it is not possible to
resolve this issue here.
Cellular localization of TIP-B1
Because the cellular and subcellular (i.e., nuclear vs.
nonnuclear) localization of proteins can often provide clues about
function, immunohistochemical analysis of TIP-B1 expression was
performed on multiple rat tissues. In all cases, consecutive sections
were taken, one was stained with anti-carboxy-15-mer antibody, and the
other was stained with control rabbit IgG. Essentially no staining was
seen in any case on the sections stained with control rabbit IgG
(Fig. 6
). As shown below (Figs. 7
8
9
10)
, nuclear and nonnuclear staining
could be distinguished. In nonnuclear staining, the observed pattern
was almost exclusively diffuse, not punctate, staining. Diffuse
nonnuclear staining is referred to as cytoplasmic staining, although
this does not mean that the association of TIP-B1 with organelles such
as the endoplasmic reticulum, Golgi apparatus, or secretory granules
has been excluded.
|
|
|
|
|
Immunohistochemical analysis of rat spleen
Two cross sections of rat spleens are shown in Figure 8
at three different magnifications. The red pulp was strongly
stained (Fig. 8A
and 8D)
. In Figure 8A
, the white pulp (delineated by
black arrows) ensheathed the central arteriole, which in this case was
acentric due to the presence of a germinal center. With the exception
of the T-cell area of the periarteriolar lymphoid sheath (black
arrowheads), the white pulp, including the B-cell-rich area surrounding
the germinal center, was relatively unstained. In a
higher-magnification view (Fig. 8B)
, it was possible to see that the
staining of the T-cell-rich region (arrowheads) was diffuse and
heterogeneous. This periarteriolar area contained largely T cells and
interdigitating dendritic cells. On further magnification (Fig. 8C)
, it
appeared that the diffuse staining might represent staining of the
dendritic cells (arrowheads). At the boundary of the periarteriolar
region (arrows), lymphocytes whose nuclei were stained for TIP-B1
(double arrowheads) could be seen. The contrast between the high
staining in the red pulp and the moderate staining in the T-cell-rich
areas (Fig. 8D
, black arrowheads) in the otherwise low TIP-B1-stained
white pulp is also seen in Figures 8D
and 8E
. The very low TIP-B1
staining of the germinal center can be seen best in Figure 8F
.
Immunohistochemical analysis of rat lymph node
TIP-B1 expression in lymph node tissue is shown in Figure 9
. The subcapsular sinus (arrow, Fig. 9A
and 9B
) was highly stained,
but the lymphocytes in the cortex (arrowheads, Fig. 9A
and 9C
) were not.
On examination at higher magnification (Fig. 9B)
, it could be seen that
the sinusoidal histiocytes were highly stained. Circulating lymphocytes
in sections through the large blood vessels (arrows, Fig. 9C
and 9D
) had
higher TIP-B1 staining than did the adjacent lymph node-resident
lymphocytes (arrowheads). In the medullary sinuses (Fig. 9E
and 9F)
, the
histiocytes (arrows) were highly stained, but lymphocyte staining was
variable.
Immunohistochemical analysis of rat mammary gland
The sections shown in Figure 10
were from rat mammary glands at the following stages: day 28 of
age (A and B), day 14 of pregnancy (C and D), day 7 of lactation (E and
F), and day 4 of involution after removal of the pups at 21 days of age
(G and H). In a day-28 developing duct, strong cytoplasmic staining of
the luminal epithelial cells and basal myoepithelial cells were seen
(Fig. 10A
, arrowhead and black arrow, respectively). In contrast,
surrounding stromal cells were poorly stained (white arrow). In the
terminal-end bud (Fig. 10B)
, TIP-B1 was strongly stained in the cap
cell layer (Fig. 10B
, arrowhead), while staining of the body cells was
generally decreased (Fig. 10B
, arrow), with the exception of some cells
directly adjacent to forming lumen (Fig. 10B
10L)
. Glands from day-14
pregnant rats showed apical staining of ducts (Fig. 10C
, small arrow)
and alveolar luminal epithelium (Fig. 10D
, small white arrows).
Overall, TIP-B1 staining was decreased in the cytoplasm of ductal
epithelial cells (Fig. 10C
, large arrow) compared with the alveoli
(Fig. 10C , ALV) that they drained but increased in cells closest to the
alveoli. Thus, the transitional epithelium of the subtending duct
approaching an alveolus (Fig. 10C
, black arrowhead) had higher TIP-B1
staining than that of the ductal epithelial cells. Diffuse cytoplasmic
staining was seen throughout the alveolar epithelium of an individual
alveolus, delineated by black arrows in Figure 10D
. During lactation,
the lumen of the distended ducts was filled with secretions (Fig. 10E
,
arrow) which stained strongly for TIP-B1. The cells of the luminal
epithelium lining the duct showed heterogeneous nuclear staining for
TIP-B1; some nuclei showed high staining and adjacent nuclei showed no
staining (Fig. 10E
, black arrowheads and white arrowhead,
respectively). Alveoli of the lactating gland also showed heterogeneous
nuclear staining, with intensely stained nuclei and unstained nuclei
(Fig. 10F , black arrowhead and white arrowhead, respectively) present
within the same alveolus. TIP-B1 could also be seen staining the casein
micelle (Fig. 10E
, crescent shape) surrounding the milk fat droplet
(Fig. 10E , black arrow). At day 4 of involution, the ducts maintained
heterogeneous nuclear staining (Fig. 10G
, white and black arrowheads),
and variable staining of the luminal secretions. Mast cells (Fig. 10G ,
arrow), which were abundant in the stroma at this time point, were
unstained. An individual alveolus within a regressing alveolar lobule
at day 4 of involution showed variable nuclear staining (Fig. 10H
,
white and black arrowheads). The expanding interalveolar stroma
[containing fibroblasts and invading leukocytes (Fig. 10H
, white
arrows)] between regressing alveoli was relatively unstained by TIP-B1
antisera.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Since mammary gland epithelial development can be regulated by TNF [10 11 12 ] and human MCF7 mammary adenocarcinoma cells express TIP-B1 mRNA [14 ], MCF7 cells were used to verify the findings from studies of human fibroblasts. Indeed, similar to what had been seen with fibroblasts, MCF7 cells were insensitive to the cytolytic effects of TNF after pretreatment with either low, nontoxic concentrations of TNF or purified TIP-B1. Furthermore, extracts of MCF7 cell lines contained a 27-kDa protein that was immunoreactive with all three polyclonal antisera raised against TIP-B1. As was predicted, the addition of carboxy-15-mer peptide completely blocked the reactivity of the affinity-purified anti-carboxy-15-mer peptide antibody with this protein. These results established that previously reported findings with TIP-B1 were not limited to fibroblasts but were also germane to human breast cancer cells.
The survey of RNA from rodent cells and human tissues indicated that, with the exception of pancreas, a 1.1-kb mRNA for TIP-B1 was detectable in all samples tested. The signal intensity varied from sample to sample, but among the human samples tested it was highest in spleen and lung. High expression was also noted in cells of the rat mammary gland and in a mouse macrophage cell line. Thus, similar to the widespread expression of TIP-B1 mRNA in cancer cell lines [14 ], there was mRNA for TIP-B1 present in a variety of normal tissues, especially breast and lymphoid tissues. If the TIP-B1 protein was also expressed in these tissues, this would be consistent with the hypothesis that TIP-B1 has a role in tissues known to undergo remodeling.
In fact, a majority of the normal tissue extracts that we examined contained a 27-kDa protein that reacted with the affinity-purified anti-carboxy-15-mer peptide antibody. The most intense reactivity was with rat kidney extracts, which had both the 27-kDa TIP-B1 protein and a 39-kDa protein that also reacted strongly. The inclusion of the carboxy-15-mer peptide blocked the reactivity of the antibody with both proteins (E. Berleth, unpublished results), suggesting that the two proteins might be closely related. It is possible that the 39-kDa protein is either a precursor of the active 27-kDa TIP-B1 or a highly modified form of the 27-kDa protein or that both proteins are members of a protein family; data are not yet available to distinguish among these possibilities. However, supportive data include observations that the tip-sn cDNA insert hybridizes to three sizes of mRNA [14 ] and TIP-B1 undergoes posttranslational modification [16 ].
Tissue extracts, in addition to that from rat kidney, which showed high reactivity of the anti-TIP-B1-peptide antibody with the 27-kDa TIP-B1, were from rat lung, rat ovary plus uterus, human heart, and human kidney tissues. The major reactive protein in both human and rat kidney tissues was the 27-kDa TIP-B1. In contrast, the major reactive protein in rat heart tissue was a 39-kDa protein with only faint reactivity with 27-kDa TIP-B1, whereas in human heart extract the major reactive protein was also the 27-kDa TIP-B1. Because the blot of human tissue extracts was from a commercial source and certain information, therefore, was proprietary, to help us resolve this apparent discrepancy, a blot of rat tissue extracts was purchased from the same commercial source. The pattern of reactivity seen on the purchased rat blot was quite similar to that on the human blot. Reactivity with a 39-kDa species in rat heart tissue was not seen as it had been seen in rat heart lysates prepared by our procedure. Because the rat tissue survey has been repeated a number of times with consistent results, the lack of an immunoreactive 39-kDa protein on the commercial blots suggests that the procedure used to prepare these blots preferentially selects for the 27-kDa species and/or discriminates against the extraction of the 39-kDa species. In further support of this possibility, the proteins from the MECs were obtained by Trizol extraction, and, although there was detectable TIP-B1 mRNA in the RNA and TIP-B1 protein was detected in MECs by immunohistochemistry, there was no immunoreactive protein in the Trizol extract detected by Western blotting. In contrast, the MS cell protein extracts were prepared with a standard sample buffer preparation, which included a number of protease inhibitors, and a clear 27-kDa TIP-B1-immunoreactive band was seen. It is not yet clear whether the difference was due to the lack of protease inhibitors in the Trizol extractions or the fact that TIP-B1 proteins partitioned into the aqueous RNA layer during the Trizol extraction.
Immunohistochemical evaluations confirmed and extended the information obtained by Western blot analyses. In kidney tissue, TIP-B1 staining was found to be selectively associated with the epithelium of the collecting ducts rather than the glomeruli. In addition, the distal tubules had less intense staining than the adjacent proximal tubules. This staining was similar to that seen for TNF in the kidney [20 ]. In contrast to the kidney, the liver was fairly homogeneous in TIP-B1 staining throughout the epithelium, perhaps reflecting the less compartmentalized nature of this tissue.
In the ovary, distinct patterns of TIP-B1 staining were observed in follicles at different stages of maturation. Specifically, TIP-B1 staining appeared to be more highly expressed during the development and maturation of the primary follicle, secreted within the lumen of the secondary follicle, and subsequently down-regulated, culminating in low expression of TIP-B1 protein in the corpus luteum and corpus albicans. These results indicate that TIP-B1 is modulated during the hormone-dependent proliferation and regression of the granulosa cells. Like the ovary, the uterus is a complex tissue that undergoes hormone-dependent remodeling. In the uterus, TIP-B1 was strongly stained in the uterine epithelium and smooth muscle but not in the intervening stroma. It is interesting that TNF has also been localized to the uterine epithelium; moreover, TNF immunoreactivity changes during estrous cycle progression [21 ]. It remains to be determined but it is tempting to speculate that similar, perhaps parallel changes in TIP-B1 protein expression might occur in periods of uterine epithelial proliferation, secretory differentiation, and regression during the estrous cycle.
The mammary gland also undergoes extensive, hormonally regulated epithelial remodeling during the postnatal developmental states of puberty, pregnancy, lactation, and postweaning apoptotic regression, as well as during the estrous cycle [reviewed in ref. 22 ]. The influence of TNF on remodeling, together with its differential expression during these stages, is discussed above. Therefore, it was of particular interest to determine the expression and localization of TIP-B1 at various developmental stages of the gland, to establish whether TIP-B1 might play a role in modulating the responsiveness of the mammary epithelium to TNF-induced cytotoxicity or proliferation.
At day 28 of age, prior to the onset of estrous cycling, the rat mammary gland undergoes ductal elongation and the development of primitive alveolar lobules. TIP-B1 cytoplasmic staining was high in luminal epithelium and basal myoepithelium of the forming ducts and, generally, in proliferative and invasive cell populations of the cap cell layer. It was also high in the body cells that directly abut the lumen but not in the other body cells. The high expression of TIP-B1 in the body cells near the lumen was unexpected because formation of the lumen occurs by apoptosis of the body cells. In this regard, it appears that TIP-B1 can be expressed by cells which do not survive.
In comparison to the luminal-epithelium ducts of rats at day 28 of age, TIP-B1 expression in those from rats at day 14 of pregnancy was relatively low, also compared with staining in the now abundant alveoli. In both ducts and alveoli, TIP-B1 was enriched at the apical face of the luminal epithelium, which was similar to the TNF-staining result described for the uterine epithelium during estrous cycling [21 ] and for the rat mammary gland during pregnancy [M. M. Ip, unpublished results].
During lactation, distinct nuclear staining, which was heterogeneous from cell to cell within the same ductal and alveolar units, was seen. It was intriguing that intense TIP-B1 staining was seen in secretions present in the ducts and was found specifically associated with the crescent-shaped, proteinaceous portion of the casein micelle surrounding the milk fat droplet. Its presence in the milk suggests that newborns exposed to environmental pathogens that induce TNF might need TIPs from their mother’s milk to protect them, similar to their need for antibodies.
The nuclear-staining pattern of lactation was retained for 4 days after weaning, a time when regression of the mammary epithelium approaches its peak in the rat. The retention of high TIP-B1 staining in regressing, apoptosing epithelium suggests that TIP-B1 can be associated with cell populations that are undergoing selective death by apoptosis. This association and TIP-B1’s ability to inhibit TNF-induced apoptosis might initially seem contradictory. However, one explanation consistent with these data is that nuclear sequestration of TIP-B1 might render it unavailable to inhibit apoptosis. Alternate hypotheses are that TIP-B1 might be capable of inhibiting only TNF-induced apoptosis and that cell death in this instance is not triggered by TNF, or that nuclear-localized TIP-B1 has functions other than inhibition of TNF-induced apoptosis. Nevertheless, TIP-B1 staining was high in the regressing, apoptosing mammary epithelium, and further studies are needed to understand the association of TIP-B1 with this cell population when it is undergoing selective death by apoptosis.
The staining pattern for TIP-B1 in the thymus was heterogeneous. The medullar region, containing mature T lymphocytes as well as thymic epithelium, was strongly stained for TIP-B1. In contrast, the lymphocytes in the surrounding cortex showed no staining for TIP-B1. The intensity of staining for some large cells in the cortical region, which might represent histiocytes or thymic epithelial cells (which are less abundant in the cortex), was similar to that of the medulla. The preferential staining of the thymic medulla for TIP-B1 suggests that this protein might play a role in modulating the responses of the mature T lymphocytes (and associated epithelium) to TNF. The lack of expression of TIP-B1 protein in immature cortical lymphocytes, the majority of which are slated to die by selection, suggests that these cells lack the TNF-protective effects of TIP-B1.
Because of the association of high TIP-B1 staining with mature T lymphocytes in the thymic medulla, the expression and localization of TIP-B1 were also examined in secondary lymphoid tissues. In the spleen, TIP-B1 was generally expressed more in the red pulp than in the white pulp surrounding the central arteriole. However, heterogeneity of TIP-B1 staining within the white pulp was also observed in that the T-cell-rich zone directly adjacent to the central arteriole was specifically stained, whereas the B-cell-rich regions of the primary follicle, marginal zones, and germinal centers showed weak or no staining. TIP-B1 staining of the lymph node revealed a similar heterogeneity of expression. The compartmentalization of expression of a TNF-modulating protein to distinct functional zones within the thymus, spleen, and lymph node might play a role in regulating the distinctive responses of these cells to TNF or other cytokines. Alternatively or in addition, the expression of TIP-B1 in long-lived dendritic cell populations, most likely the cells stained in the T-cell-rich zones of the spleen, might have a function in their resistance to cell death during inflammatory immune responses. The histiocytes of the subcapsular and the medullary sinuses of the lymph node, which also are intensely stained, might have a similar requirement for TIP-B1.
It was possible to also evaluate the relative TIP-B1 expression in various lymphocytes within the lymph node. More than 50% of the recirculating peripheral-blood lymphocytes seen in cross-sections of blood vessels showed intense nuclear and/or cytoplasmic staining. The percentage of lymphatic lymphocytes with intense staining was somewhat lower, and little or no staining was seen in the lymph node plasma cells or in the resident lymphocytes. The high expression of TIP-B1 protein in activated recirculating lymphocytes and to a lesser extent in the lymphatic lymphocytes but not in the lymph node resting lymphocytes or in the relatively short-lived plasma cells suggested that TIP-B1 may play a role in long-lived functional lymphocyte populations.
In summary this study has provided new information about the extremely selective tissue distribution of the unique TIP-B1 protein. TIP-B1 was found, in many cases, to be associated with tissues and with specific cells within those tissues known to be involved in TNF tissue-specific regulation. Given that cells cultured with TIP-B1 were protected from TNF-induced apoptosis, the association between TIP-B1 and TNF expression, together with the heterogeneous subcellular distribution of TIP-B1 seen in many cases, is intriguing. Thus, the interdigitating dendritic cells were known to have a critical role in TNF-dependent T- and B-cell splenic organization, and it would be expected that they would need to be protected from untoward TNF effects. They were found to express TIP-B1, and the staining was diffuse throughout the cytoplasm. Similarly, there was diffuse cytoplasmic staining in the developing mammary gland, where TNF induced growth and proliferation and the cells needed to be protected from TNF lytic effects. However, in the lactating and involuting mammary gland in which proliferation was reduced, TIP-B1 was found in the nucleus. It is interesting that there was high TIP-B1 staining of the mature peripheral-blood T cells relative to that seen in the resident lymphoid organ lymphocytes. TNF is known to be a causative agent in induction of apoptosis of mature, activated peripheral T cells [1 ], but the premature elimination of these cells must be avoided. There was both cytoplasmic and nuclear TIP-B1 staining of the mature blood lymphocytes. This is consistent with some of these cells still being needed and others no longer needed. Overall, this study identified a broad but highly specific TIP-B1 distribution, consistent with a major role for TIP-B1 in regulation of the sensitivity of diverse cells to TNF.
|
ACKNOWLEDGEMENTS |
|---|
The authors acknowledge Drs. Alicia Henn, Linda Varela, Karoly Toth, and Jennifer Black, who contributed to these studies and/or to preliminary studies. We also thank Mrs. Susan Staples, Mrs. Cheryl Eppolito, Mr. Eric Searls, and Mrs. Mary Vaughan for their invaluable assistance.
Received December 20, 2000; revised February 26, 2001; accepted February 28, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
inhibitory protein, TIP-B1 Int. J. Immunopharmacol. 22,1137-1142[Medline]
gene expression in the tissues of normal mice Cytokine 4,340-346[Medline]
messenger ribonucleic acid and protein: localization and regulation by estradiol and progesterone Endocrinology 135,2780-2789[Abstract]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, mouse J774A.1 macrophage
cell line after LPS (10-ng/mL, 3-h) stimulation; ME, rat mammary
epithelial cells; MS, rat mammary stromal cells. (B)A
membrane was purchased (Invitrogen) on which 20 µg of RNA from each
of eight human tissues had been transferred after electrophoretic
separation. This membrane was probed with a 32P-labeled
783-nucleotide tip-sn cDNA insert. Lanes with human RNA are
as follows: H, heart; Br, brain; K, kidney; Lv, liver; Lu, lung; P,
pancreas; S, spleen; and M, skeletal muscle. These data are the results
of a single experiment.
