| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
From the Pathologisches Institut,*
the Institut
für Klinische Biochemie,
and the
Institut für Zellbiologie,§
Universität Bonn, Bonn; and the Abteilung für
Cytopathologie,
Deutsches
Krebsforschungszentrum, Heidelberg, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
One type of preneoplastic liver focus, namely the amphophilic cell focus, which is well known to occur in different models of hepatocarcinogenesis in different species7,8 and which is one of the most often observed types of focal lesion in human liver cirrhosis,9 was not induced by the islet transplants.6 This type of focus of altered hepatocytes is characterized by amphophilic (basophilic and eosinophilic) cytoplasm due to a proliferation of mitochondria, which are closely associated with rough endoplasmic reticulum, and a strong decrease in glycogen.7 Alterations of enzyme activities in this type of preneoplastic liver focus show an energy-wasting metabolism and have been suggested to reflect thyromimetic effects of the responsible oncogenic agents.10,11 We thus decided to investigate whether intrahepatic thyroid follicle transplants are able to induce amphophilic cell foci in thyroidectomized rats.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Male Lewis rats (inbred strain) weighing about 200 g were used. In the main group (MG), the animals were thyroidectomized; 2 weeks later thyroid tissue pieces were transplanted via the portal vein into the right part of the liver. The animals of control group 1 (CG1) were not thyroidectomized; thyroid tissue pieces were transplanted via the portal vein into the right part of the liver. Control group 2 animals (CG2) were thyroidectomized but did not receive a transplant. The thyroidectomy was done under an anesthesia with diethylether. Housing and treatment of the animals were in line with the guidelines of the Society for Laboratory Animals Service (GV-Solas) and the strict German animal protection law.
In preliminary experiments, it was found that thyroid follicles or small pieces of thyroid tissue dissected by razor blades were too fragile for an effective transplantation. The rigidity of intrafollicular thyroglobulin was suspected to be the main problem. Therefore a pretreatment of donor animals was necessary to get follicles almost depleted of thyroglobulin. Donor animals received an iodine-poor diet (<0.04 µg iodine/g) with 0.05% propylthiouracil and tap water for 10 days, followed by an iodine-poor diet and H2O containing 1% KClO4 for 2 days; and finally an iodine-poor diet and H2O for 3 days. The following day the donor animals were killed, and the thyroid gland was prepared. The pretreatment resulted in an about fourfold enlargement of the gland with almost empty follicular lumina. The glands were washed with Hanks’ solution (pH 7.2) and dissected with razor blades into small cubes of about 0.2 mm3. One gland was used for two recipient animals. The small cubes were dispersed in ice-cold Hanks’ solution (pH 7.2) with 1% Dextran (Mr 35,000–50,000; Merck, Darmstadt, Germany) and were aspirated with 500 µl of this mixture into a syringe with a needle of 23G x 1''. The addition of dextran reduced the attachment of the tissue cubes to each other and to the wall of the syringe. The recipient animals were anesthetized with diethyl ether, and the thyroid tissue pieces were transplanted only into the right part of the liver as follows. The branch of the portal vein that supplies the left lobe and the left part of the middle lobe was occluded by a vessel clamp. After infusion of the thyroid transplants into the portal vein, the clamp was removed (maximal time of ischemia, 1 minute). With this procedure it was possible to infuse the thyroid tissue pieces only into the right part of the liver, ie, the right lobe, the caudal lobes, and the right part of the middle lobe (the border between the right part and the left part of the middle lobe is marked by the falciform ligament). Thus, the left part of the middle lobe and the left lobe could be taken as an internal control in the experiments of the MG and the CG1.
At 1 week, 3 weeks, and 3 months after transplantation, six animals of each experimental group were killed between 10:00 a.m. and 11:30 a.m. An additional three animals of the MG were killed 18 months after transplantation. Additional MG animals were killed at days 3 (three animals) and 4 (three animals) without 5-bromo-2'-desoxyuridine (BrdU) application and without serum sampling (see below). Two hours before killing, another MG animal at 2 months after transplantation received 200 µCi 125I intravenously, and the intrahepatic thyroid transplants were prepared for autoradiography.
BrdU Application by the Method of Eldridge et al12
Seven days before killing, half of the animals to be killed at 1 week, 3 weeks, and 3 months and all three animals to be killed at 18 months were anesthetized, and osmotic pumps (Alzet model 2ML1, Alza Corp., Palo Alto, CA) filled with 40 mg of BrdU (Sigma, Heidelberg, Germany) were surgically implanted subcutaneously over the dorsal thoracolumbal area. These pumps continuously delivered BrdU until the animals were sacrificed. The remaining animals to be killed at 1 week, 3 weeks, and 3 months received a single dose of 50 mg BrdU/kg body weight intraperitoneally 1 hour before sacrifice.
Preparation of Tissues
Slices from the right part and from the left part of the middle lobe of the liver were snap-frozen and were used for enzyme histochemistry (see below). After removing the middle lobe of the liver, the animals were perfused with a mixture of 0.5% glutaraldehyde and 3% formaldehyde as described earlier.4 Immediately after perfusion, about 40 slices per animal were cut from the fixed liver lobes. These slices were transferred into phosphate-buffered saline (PBS) and were systematically examined with a stereomicroscope. With some experience it was possible to identify transplanted thyroid tissue pieces in these unstained liver slices as well as after embedding (see Results). Corresponding slices of the same transplants were embedded in Epon and in paraffin. Additionally, pituitary gland, kidney, adrenal glands, lung, heart, spleen, and pancreas were embedded in paraffin. Four horizontal sections of the upper trachea and larynx together with surrounding tissue were embedded in paraffin to check the area of thyroidectomy (completeness of thyroidectomy) of the MG and the CG2 or the thyroid gland of CG1.
From the paraffin-embedded specimen, serial sections of 2 to 3 µm in thickness were stained with hematoxylin and eosin (H&E) and with the periodic acid-Schiff reaction (PAS). Additional sections were used for immunohistochemistry. In addition, 10 small cubes (1 mm3) were cut from each liver and were embedded in Epon. Semithin sections of the Epon-embedded specimens were stained by the method of Richardson et al.13 Thin sections for electron microscopy were stained with uranyl acetate and lead citrate and were examined with a Phillips CM10 electron microscope (Einthoven, The Netherlands).
Immunohistochemistry
After examination of the H&E and PAS stains, appropriate sections
were selected and the corresponding sections were processed for
immunohistochemistry. Immunostains of the liver for BrdU (monoclonal
primary antibody from DAKO, Hamburg, Germany; dilution 1:100) and
transforming growth factor-
(monoclonal primary antibody from
Oncogene Science, Cambridge, MA; final antibody concentration 10
µg/ml) with antigen retrieval steps were performed as described
earlier.4,6
Glutathion S-transferase placental form
(GST-P; polyclonal primary antibody from Biogenex, San Ramon, CA;
dilution 1:100) was analyzed without antigen retrieval steps,
using the LSAB+-Kit (DAKO) and the
DAB+-Kit (DAKO). Immunostains of the hypophysis
for thyroid-stimulating hormone (TSH) were done with a monoclonal
antibody from DAKO (dilution 1:50) using the
LSAB+ and the DAB+ Kits.
Sections were counterstained with hematoxylin, dehydrated, and
coverslipped with Pertex (Medite, Burgdorf, Germany).
Anti-T3 and anti-T4 (polyclonal primary antibodies from ICN Biomedicals, Eschwege, Germany) were applied at a dilution of 1:250, anti-TG (polyclonal antibody, Institute of Cell Biology, University of Bonn, Bonn, Germany) and anti-calcitonin (monoclonal antibody from DAKO) were used at a dilution of 1:50. Secondary fluorescein isothiocyanate-conjugated goat anti-rabbit or goat anti-mouse antibodies (Sigma Chemical Co., Heidelberg, Germany) were used at a dilution of 1:50. In controls primary antibodies were omitted.
Determination of Apoptotic, Mitotic, and BrdU-Labeling Indices and Statistical Analysis
H&E-stained paraffin sections were viewed systematically for apoptotic and mitotic cells at a magnification of x400. Mitotic and apoptotic indices (MI and AI) were calculated as the number of mitotic figures and of apoptotic bodies, respectively, per 1000 hepatocytic nuclei. Paraffin sections immunostained for BrdU were examined at a magnification of x400. BrdU labeling indices (BrdU-LI) were calculated as the number of BrdU-labeled nuclei per 1000 hepatocytic nuclei. 5000–15,000 hepatocytic nuclei were counted per animal.
For the MG and the CG1, all indices were calculated separately for the liver acini downstream of the transplanted thyroid follicles in the right part of the liver, and in the left part of the liver which was free of transplants (intraindividual control). When no alterations were identifiable downstream of the transplants (MG at 1 week; CG1 at all times), the indices for the right part of the liver were calculated with the hepatocytes at a distance of about 1 mm from the transplants. For CG2 the indices were calculated only for the right lobe.
The body weight, AI, MI, BrdU-LI at 1 hour, BrdU-LI at 7
days, serum T3, serum thyroxin (T4), and serum TSH of the different
animal groups and the different times after thyroid follicle
transplantation (Table 1)
were compared
with the Wilcoxon-Mann-Whitney test. Significance was accepted when
P < 0.05.
|
Two hours before sacrifice, 200 µCi of 125I were injected into the tail vein of an animal from the MG at 2 months after transplantation. The liver tissue was fixed in 4% formaldehyde and embedded into paraffin. Autoradiography was done with 3-µm-thick paraffin sections and viewed with dark-field illumination at a Zeiss Axiophot microscope (Oberkochem, Germany).
Serum Free T3, Free T4, and TSH
Serum samples were taken from aortal blood at sacrifice. Free T3 (fT3) and free T4 (fT4) were measured by an electrochemiluminescence immunoassay with an Elecsys analyzer (Boehringer Mannheim, Mannheim, Germany). TSH was measured with the rat TSH 125I assay system from Amersham (Amersham, Buckinghamshire, UK).
Histochemistry of Enzymes and Metabolic Products
Pieces from snap-frozen liver tissue of five rats were
frozen onto the same tissue holder, and serial sections of all pieces
were cut simultaneously in a cryostat (Jung, Nussloch,
Germany).14
The sections were mounted onto the same slide
or the same membrane and incubated by the respective histochemical
reaction. With this technique it was possible not only to produce
sections of the same thickness but also to treat them
simultaneouslyunder identical conditions for the specific histochemical
assays. The liver and the thyroid gland of one completely untreated rat
were included as a normal control. The following enzymes were
investigated: glycogen synthase (SYN), glycogen phosphorylase (PHO),
glucose-6-phosphatase (G6Pase), glucose-6-phosphate dehydrogenase
(G6PDH), pyruvate kinase (PK), succinate dehydrogenase (SDH), malic
enzyme (ME), mitochondrial glycerol-3-phosphate dehydrogenase (mG3PDH),
cytochrome c-oxidase (COX), acid phosphatase (AP), and
-glutamyltransferase (GGT). Incubation conditions were essentially
as previously described.11,15-17
Furthermore, serial
cryostat sections were stained for basophilia with toluidine blue, for
the presence of neutral lipids with Fettrot B, and for the presence of
glycogen with the PAS reaction.
The intensities of the histochemical parameters in the serial sections were estimated semiquantitatively using five grades (no change, increase, strong increase, decrease, strong decrease) as compared with the reaction in the unaltered tissue of the same specimen (internal control).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The body weight gains are shown in Table 1
. At 1 week and 3 weeks
after transplantation, MG and CG1 were not different from CG2,
obviously because CG2 had no abdominal surgery. At 3 months, the weight
gain of CG1 was higher compared with CG2. The histological examination
of the heart, lung, kidney, adrenal gland, spleen, and pancreas did not
reveal any unusual finding. The immunohistochemically TSH-positive
cells of the hypophysis were larger in the thyroidectomized animals (MG
and CG2) than in the nonthyroidectomized animals (CG1).
As expected, the MG and CG2 animals had significantly decreased fT3 and
fT4 and significantly increased TSH serum values compared with the CG1
animals (Table 2)
. It was surprising that
no animal of the CG2 had completely negative fT3 or fT4 values. We made
four-step sections at the anatomical site of the excised thyroid, but
we did not find a thyroid remnant in any case. Between 3 and 18 months
after transplantation, the fT3 and fT4 values of the MG increased, and
the TSH decreased. At 18 months, the serum values (and the hypophyses)
of the MG did not differ from the CG1 at 3 months.
|
Stereomicroscopic examination of the unstained liver slices was a
great help for finding the transplanted follicles for light and
electron microscopical investigations (Figure 1, a and b)
. The transplants were
vascularized (Figure 1, c and d)
during the first 3 weeks, after which
the follicles were surrounded by regularly fenestrated capillaries
(Figure 2)
. In contrast, the transplants
of the CG1 rats fibrosed during the first 3 weeks (Figure 1, e and f)
,
probably due to low ("normal") TSH serum levels.
|
|
, and positive
immunohistochemical stainings for TG, T4, or T3 (Figure 3, A–C)
during the whole experiment.
Only a few calcitonin-positive C cells were found by means of
immunohistochemistry. In many specimens investigated electron
microscopically, no C cell was present. The autoradiography after
125I application at 2 months after
transplantation showed an iodine uptake by the intrahepatic thyroid
follicles (Figure 4)
. The follicular
epithelial cells of the MG exhibited a strong proliferative activity
(Figure 5, a and b)
. A typical follicle
transplant at 3 months is shown in Figure 6
. At 18 months most of the transplants
did not differ from those at 3 months, but single transplants grew up
to 2 mm in size (Figure 7, a
-c), and were
then still showing mitotic figures.
|
|
|
|
|
At 3 days, 4 days, and 1 week after transplantation, no
morphological alterations of the liver tissue surrounding the thyroid
follicles were observed in the MG. However, in the MG and the CG1, the
hepatocytes in the neighborhood of the transplants showed a slight
increase in the MI and the BrdU-LI-1h as compared with the CG2.
Furthermore the BrdU-LI-1h of the MG was increased in the right part of
the liver compared with the left part (Table 3)
.
|
; Figure 5, b and c
).
The 10-fold increase in the AI mean value in relation to the mean value
of the left part was not significant (due to a high SEM).
At 3 as well as at 18 months after transplantation, the liver acini
downstream of the transplanted follicles of the MG were clearly
demarcated from the surrounding unaffected acini by a strong increase
in cytoplasmic basophilia and eosinophilia and a negative PAS staining
(Figure 6
; Figure 7, b
-e and g). These altered light microscopic
staining properties resulted from an increase in rough endoplasmic
reticulum, a multiplication of mitochondria, and a depletion of
glycogen, as shown by electron microscopy (Figure 8)
. The number of peroxisomes was not
increased in the cytoplasm of the altered hepatocytes. Within some
altered liver acini, single hepatocytes showed intracytoplasmic areas
of glycogen accumulation, whereas in other areas of the same cells
glycogen was depleted but mitochondria and rough endoplasmic reticulum
were increased (Figure 7h)
. It is not clear whether these cells are
intermediate cells between unaltered and altered hepatocytes or they
result from secondary events in the altered-cell population. The nuclei
of the hepatocytes downstream of the transplanted follicles were
enlarged and hyperchromatic, and they showed multiple condensations of
their chromatin and several nucleoli (Figure 7i)
. Many mitotic figures,
some of which were pathological (Figure 9)
, but also many hepatocytic apoptotic
bodies, were found in the same altered liver acini (Figure 7g)
, which
is reflected by the fact that the MI, AI, and BrdU-LI (1h and 7d) at 3
months were strongly increased compared with 1) the left liver part of
the same animals, 2) the MG at 1 week, and 3) CG1 and CG2. Furthermore
the AI and the BrdU-LI 1h were significantly increased compared with
the MG at 3 weeks. At 18 months, the MI was significantly higher than
at 3 months (Table 3)
. Single altered acini at 18 months showed an
amphophilic cytoplasm with basophilic stripes (Figure 10)
. This was a result of a highly
organized rough endoplasmic reticulum, whereas the number of
mitochondria was increased, and glycogen was depleted. This type of
alteration was called amphophilic-tigroid cell focus. The livers of the
animals of CG1 and the CG2 did not show any of the alterations observed
in the MG.
|
|
|
The enzyme activities of the transplanted follicles did not differ from the normal control gland. In the unaltered liver acini, the histochemical pattern of the liver parenchyma reflected the well-known metabolic zonation showing gradients in enzyme activities from zone 1 to zone 3 of the acini. Thus, the reactions of the histochemical assay for G6Pase, G6PDH, SDH, AP, mG3PDH, COX, PHO, and SYN were stronger in zone 1 than in zone 3, whereas those for PK and the PAS stain were more pronounced in zone 3. ME and mG3PDH were more intense in zones 1 and 3 and less intense in zone 2. GGT was positive in bile ducts and some hepatocytes of zone 1.
The histochemical patterns of the altered liver acini at 3 and 18
months after follicle transplantation are summarized in Table 4
, and examples are shown in Figures 11 and 12
. The altered acini were
characterized by increased activities of G6PDH, ME, G6Pase, mG3PDH,
COX, AP (Figures 11, d–i
; and 12, b, e, f) and SDH. The activities of
PHO and SYN were strongly reduced (Figures 11c and 12
, c and d). PK was
not changed. GGT was seen orthotopically only in bile ducts, but was
not expressed in the altered acini.
|
|
|
and GST-P were
negative in the altered acini. At 18 months after follicle
transplantation, one animal exhibited a few very small foci of
glycogen-storing hepatocytes, which were seen in paraffin sections of
both parts of the liver and were not topographically related to the
transplants. These foci were positive for transforming growth
factor- and GST-P. In the cryostat sections of the middle lobe of
this animal, the incidental single glycogen-storing foci were not
present. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The hepatocytes of the liver acini downstream of the transplants were altered in their morphology, cell turnover, and enzymic pattern. The increase in hepatocellular proliferation and, at the same time, in their apoptotic elimination were the first alterations that could be detected after thyroid follicle transplantation. This supports earlier findings that T3 is a strong mitogen for hepatocytes inducing so-called direct hyperplasia and polyploidization in vivo.1,2,19-21 However, it cannot be ruled out that additional factors secreted by the thyroid tissue may contribute to the effects observed. Interestingly, the AI was simultaneously increased in the altered liver acini, which is a common phenomenon of preneoplastic liver foci including amphophilic cell foci22-24 that was also observed in glycogen-storing altered liver acini after pancreatic islet transplantation.4
The altered hepatocytes did not grow beyond the borders of their acini, where they would be out of the influence of the thyroid hormones. It is unlikely that the altered hepatocytes undergo a complete reversion to "normal" after their outgrowth of the acinar borders and the influence of the hormones because there was no indication of any transitional changes in the adjacent acini. An immediate change under the microenvironmental conditions of the adjacent acini is not plausible, because the development of the alterations needed such a long time.
An intriguing question is whether the altered acini represent clonal populations. Compared with alterations of the hepatocytes that occurred 2 days after islet transplantation and that were obviously not monoclonal, the slow onset of the alterations after thyroid follicle transplantation might argue for a monoclonal expansion. However, the observation that, 3 and 18 months after transplantation, all transplants induced amphophilic cell alterations in all animals of the MG is an argument against monoclonality. Moreover, no single transplant without a surrounding amphophilic cell population was found either 3 months or 18 months after transplantation. There was also no indication for the induction of a basically different phenotypic hepatocellular alteration. All alterations after transplantation were strictly confined to the acinar borders, this being a further argument against early clonal expansion of the amphophilic cell populations.
The increase in mitochondria and rough endoplasmic reticulum as well as the decrease in glycogen are in line with expected alterations induced by thyroid hormones.3 The resulting amphophilic H&E staining of the cytoplasm, however, is very similar to a well-known type of hepatocytic alteration in the amphophilic cell focus. This is a preneoplastic phenotype mainly induced by chemical hepatocarcinogens belonging to the group of peroxisome proliferators, eg, hypolipidemic drugs, and the adrenal gland hormone dehydroepiandrosterone.7,9,11,25 Peroxisome proliferators and T3 are both so-called direct mitogens, which may induce hepatocyte proliferation without preceding cell loss.1,2,19 They both act via nuclear receptors on some common target genes, eg, ME and COX.26,27 Thyroid hormone receptor and peroxisome proliferator-activated receptor belong to a nuclear receptor superfamily, together with the receptors of sexual steroids and retinoic acid.2,28,29 Alterations induced by T3, such as the increase in activities of COX, mG3PDH, ME, G6Pase, G6PDH, SDH, and AP and the decrease of PHO and SYN have also been observed in preneoplastic amphophilic-cell foci induced by peroxisome proliferators.7,10,11 However, peroxisome proliferation was not present after follicle transplantation.
It is also interesting that both the amphophilic cell foci and the
amphophilic liver acini presented here exhibit neither an increase in
GGT nor an expression of tranforming growth factor- or GST-P, which
are often involved in the glycogenotic-basophilic progression in many
models of hepatocarcinogenesis.7,30
The progression from
amphophilic cell foci to hepatocellular neoplasia usually needs more
time than the neoplastic progression of glycogenotic cell populations.
Therefore, the fact that the three animals of the MG at 18 months did
not develop a hepatocellular tumor does not exclude that these altered
liver acini are preneoplastic.
The nonparenchymal cells in the altered liver acini after thyroid follicle transplantation have not yet been studied; nevertheless, they might play an important mediator role between the thyroid follicles, their hormones, and the hepatocytes.31
In conclusion, we have demonstrated a new transplantation model for the study of the effects of thyroid hormones on hepatocytes. Long-term studies have been started to investigate whether the altered liver acini downstream of the transplanted follicles not only showed morphological and enzymic alterations similar to preneoplastic amphophilic cell foci but also may proceed to neoplasia.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by the Deutsche Forschungsgemeinschaft (grant numbers DO 622/1–1 and DO 622/1–3) and the Bonner Forum Biomedizin.
Accepted for publication September 16, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  G. M. Ledda-Columbano, A. Perra, D. Concas, C. Cossu, F. Molotzu, C. Sartori, H. Shinozuka, and A. Columbano Different Effects of the Liver Mitogens Triiodo-Thyronine and Ciprofibrate on the Development of Rat Hepatocellular Carcinoma Toxicol Pathol, January 1, 2003; 31(1): 113 - 120. [Abstract] [PDF]
|
|
|
|
|
|
R. T. Miller, L. A. Scappino, S. M. Long, and J. C. Corton Role of Thyroid Hormones in Hepatic Effects of Peroxisome Proliferators Toxicol Pathol, January 1, 2001; 29(1): 149 - 155. [Abstract] [PDF]
|
|
|
|
 |
|
 L. Klotz, H. J. Hacker, D. Klingmuller, P. Bannasch, U. Pfeifer, and F. Dombrowski Hepatocellular Alterations after Intraportal Transplantation of Ovarian Tissue in Ovariectomized Rats Am. J. Pathol., May 1, 2000; 156(5): 1613 - 1626. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
