| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Histology Unit (A.-C.G., M.-C.M., I.M.C., J.-F.D.) and Endocrinology Unit (C.D.), Université Catholique de Louvain, Medical School, B-1200, Brussels, Belgium; Institut de Recherche Interdisciplinaire (S.C., F.M.) Université Libre de Bruxelles, B-1070 Brussels, Belgium; and Academic Medical Center (J.J.M.D.), University of Amsterdam, 1100 DE Amsterdam, The Netherlands
Address all correspondence and requests for reprints to: J.-F. Denef, Histology Unit, Université Catholique de Louvain, Medical School, Avenue Emmanuel Mounier, 52, UCL 5229, B-1200, Brussels, Belgium. E-mail: . denef{at}isto.ucl.ac.be
Abstract
In animals, as well as in humans, the thyroid gland is made of active follicles, with cuboidal cells and hypofunctioning follicles, with flattened cells. In this study, the functional status of human follicles was dissected out, based on immunohistochemical detection of TSH receptor, Na+/I- symporter, pendrin, thyroperoxidase (TPO), thyroid oxidases (ThOXs), and T4-containing iodinated Tg (Tg-I). To ascertain that angiofollicular units exist in the human, we studied the microvascular bed of each follicle, in correlation with detection of vascular endothelial growth factor (VEGF), of nitric oxide synthase III, and of endothelin in normal and goitrous thyroids. In hypofunctioning follicles, pendrin, TPO, and ThOXs were not detected, and there was no Tg-I in the colloid. At the opposite, in active follicles, pendrin, TPO, and ThOXs were detected in thyrocytes, and Tg-I was present in the colloid. In normal and goitrous thyroids, the capillary networks surrounding active follicles were larger than those surrounding hypofunctioning follicles. Immunoreactivity for nitric oxide synthase III and endothelin was solely detected in active follicles. Only a few follicles in normal thyroids were immunostained for VEGF, regardless of their functional status. In multinodular goiters, VEGF was detected in contact with the extracellular matrix at the basal pole of the cells. In conclusion, the present study endorses the likelihood of angiofollicular units in the human thyroids. Vascular changes are related to the functional status of thyrocytes.
AS FOR MANY other mammals, the human thyroid gland is characterized by a marked functional and morphological heterogeneity. This intrinsic property has been widely considered to understand the pathogeny of multinodular goiter (MNG) (1). At the anatomical level, MNG consists of clustered functioning (hot) and resting (cold) nodules, as evidenced by iodine-123 and technetium-99m scans. At the microscopical level, the interfollicular heterogeneity is clearly visible, as evidenced by autoradiography after iodine-125 injection in experimental animal models (2, 3, 4). Autoradiographic investigations in the human are obviously scarce, although one study showed that some cold follicles cannot trap iodine, whereas others do so but without organification (5).
An alternative approach to characterize the functional status of follicles is to investigate the expression of proteins involved in iodine transport and organification. Thyrocytes are able to actively transport iodine through the basolateral membrane via a Na+/I- symporter (NIS) (6, 7, 8) and to deliver it to the apical pole of the cell, likely through another iodide transporter, the product of the pendred syndrome gene (PDS), pendrin (9, 10, 11, 12, 13, 14). The Tg iodination and coupling processes take place at the apical pole, on the follicular side of the cell membrane. They require the presence of a thyroperoxidase (TPO), as well as an H2O2-generating system, including members of the NADPH oxidase family, thyroid oxidases (ThOXs) (15, 16). By using specific antibodies raised against these different components, combined with the specific detection of T4-containing iodinated Tg (Tg-I) in the follicular lumen (17), it is possible to get a functional overview of the follicles.
The likely role of the microvascular bed in the pathogeny of the glandular heterogeneity has also been considered. Changes in vascular structures have already been studied in goitrous experimental models (18, 19, 20, 21). We previously showed, in the mouse, that a correlation exists among the functional status of the follicles, the extent of the microvascular bed, and the expression of nitric oxide synthase III (NOSIII). These results suggested that the thyroid gland consists of clustered angiofollicular units regulated by local paracrine exchanges between epithelial cells and their neighbor capillary bed (22). This hypothesis remains speculative in the human thyroid gland.
In the present study, we address this specific question using human material. We first characterized the functional status of follicles through the immunohistochemical detection of the TSH receptor (TSHr), NIS, pendrin, TPO, ThOXs, and Tg-I, in correlation with the morphological aspect of the follicles. To verify the hypothesis about the presence of angiofollicular units in normal and goitrous human thyroids, we analyzed variations in the vascular bed, along with the functional aspect of the follicles, by morphometrical methods. We then correlated them with the expression of angiogenic and vasoactive factors [the vascular endothelial growth factor (VEGF), NOSIII, and endothelin (ET)].
Materials and Methods
Human thyroid tissues
Five human thyroid samples, obtained after autopsy, were used as normal glands. Five MNGs were selected from a tissue bank of the pathology department. Tissues were fixed in Bouin’s fluid and embedded in paraffin. Five other MNGs and four normal paranodular thyroids were obtained from surgical cases, after the patients gave their informed consent. Tissue samples were rapidly frozen in liquid nitrogen to perform cryostat sections.
Immunohistochemical analysis of thyroid-specific proteins (TPO, NIS, TSHr, ThOXs, pendrin, Tg-I, and Tg) and of angiogenic and vasoactive peptides (NOS, VEGF, and ET)
Cryostat sections (5-µm) were used to perform TPO [monoclonal antibody (mAb)47, gift from J. Ruf, U38 INSERM, Marseille, France], NIS [mAb (23)], TSHr (polyclonal antibody; DAKO Corp., Glostrup, Denmark), ThOXs [polyclonal antibody (15)], and pendrin (polyclonal antibody, gift from I. Royaux, NIH, Bethesda, MD) immunostainings. Paraffin sections (5-µm), laided on superfrost glass slides, were used for Tg-I [mAb (17)], Tg (polyclonal antibody, DAKO Corp.), TPO, NIS, VEGF (polyclonal affinity-purified antibody; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), NOSIII (polyclonal affinity-purified antibody; Transduction Laboratories, Inc., Lexington, KY), and ET (mAb; Biogenesis, Poole, UK) immunostainings. Paraffin sections were dewaxed and rehydrated just before the procedure. Paraffin-embedded sections used for NOSIII and NIS immunostainings were pretreated in a microwave oven in citrate buffer (0.01 M, pH 6) for one cycle of 3 min at 750 W and three cycles of 3.5 min at 350 W.
Immunostainings were performed as previously described (22). Both cryostat- and paraffin-embedded sections were washed with PBS supplemented with 1% BSA (ICN Biochemicals, Inc., Costa Mesa, CA) (PBS-BSA) and thereafter incubated in PBS-BSA containing normal serum from the animal species from which the second antibody was raised (1/50) for 30 min at room temperature. First antibodies were then applied either overnight or for 1 h (Table 1
). After two washes in PBS-BSA, the binding of antibodies was detected using a secondary antibody conjugated to a polymer labeled with peroxidase (Envision detection, DAKO Corp.) for 30 min at room temperature. The peroxidase activity was revealed with AEC substrate (3-amino-9-ethylcarbazole, DAKO Corp.). Sections were counterstained with Mayer’s hematoxylin, rinsed, and mounted in Faramount (DAKO Corp.). The binding of anti-ET antibody was detected using the ABC perox kit (Vector Laboratories, Inc., Burlingame, CA), that of anti-Tg-I by using a monoclonal secondary antibody conjugated with peroxidase, and that of polyclonal anti-Tg with a polyclonal antibody conjugated with peroxidase. For Tg-I and Tg, the peroxidase activity was revealed with 3-3'-diaminobenzidine tetrahydrochloride (Aldrich, Bornem, Belgium); sections were dehydrated and mounted in DPX (BDH, Poole, UK).
|
For all immunostaining procedures performed on paraffin sections, follicles were ranked as described below, and the proportion of positive vs. negative immunostained follicles was recorded. Follicles were considered as positively stained when all their cells were positive; otherwise, they were considered as partly positive or negative. For Tg-I immunostaining, the aspect of the follicular lumen was taken into account.
Ranking of the follicles according to their histological aspect
Ranking was performed on thick sections (5-m) stained with Hemalun-Eosine-Safran. Follicles were ranked into two categories, depending on the aspect of their epithelium. Follicles were considered as hypofunctioning when they were lined with a flat epithelium containing dense nuclei. On the contrary, follicles were considered as active when they were lined with a cuboidal or cylindrical epithelium with rounded nuclei (2, 3, 4). Follicles with an intermediary aspect were excluded from the quantitative study. The validation of this qualitative ranking by morphometrical analysis was previously reported in the mouse (22).
Morphometrical analysis of the microvascular bed
After follicles were ranked as hypofunctioning or active, according to their morphological aspect, the morphometrical evaluation of the vascular bed was performed by point counting on 150 follicles per section, according to Weibel et al. (24), as previously reported (22). The relative volume (Vv) of capillaries in contact with thyrocytes, the number of sections of capillaries in relation with each follicle, and the mean area of capillary profiles were calculated separately for hypofunctioning and active follicles.
Statistical analysis
Statistical differences between active and hypofunctioning follicles within the same group and between normal thyroid and MNG were tested by one-way ANOVA, followed by Fisher or Dunnett tests. Results are expressed as mean ± SEM for the morphometry, or as mean ± SD for immunostaining counting.
Results
Characterization of active and hypofunctioning follicles (Fig. 1)
In both normal thyroids and MNGs, a mixture of active and hypofunctioning follicles was observed. No systematic correlation between the size of the follicles and the epithelial thickness was observed. Some large follicles were lined with cuboidal or cylindrical epithelial cells with round nuclei and a well-developed cytoplasm. On the contrary, some large follicles, as well as numerous small ones, were lined with flat epithelial cells with flat and dense nuclei. In some cases, the epithelium was so flat that it was nearly indistinguishable on sections.
|
In cryostat sections, a positive staining for all markers (TSHr, pendrin, ThOXs, and TPO) was seen in active follicles. In contrast, hypofunctioning follicles were positive only for TSHr. Pendrin, ThOXs, and TPO were either not detected or only observed in very few cells (Fig. 2, A–F). The immunohistochemical pattern varied, though, from one marker to another. Staining for TSHr was almost exclusively located at the basal pole of the cells (Fig. 2A), whereas pendrin was mainly observed at the apical pole (Fig. 2D). In contrast, ThOXs staining was present in the cytoplasm under the apical pole (Fig. 2B), and TPO immunolabeling was detected diffusely throughout the whole cytoplasm (Fig. 2F). The immunohistochemical detection of NIS showed a very heterogeneous pattern. The number of follicles with positive cells varied from one gland to another, without differences between normal and goitrous thyroids. Some glands contained numerous negative follicles regardless of their functional status. Positive cells were stained at the basolateral pole of the cells (Fig. 2, G and H). There was also an intraglandular variation in the number of positive cells among follicles. Thus, NIS immunoreactivity was not indicative about the functional status of the follicles. In both normal thyroids and MNGs, Tg-I immunostaining in the colloid was positive in active follicles, whereas it was negative in hypofunctioning ones (Fig. 2, I and J). Nevertheless, Tg-I-negative follicles contained Tg, given that they were positively stained with a polyclonal antibody against Tg (Fig. 2, K and L).
|
). In normal thyroids, TPO immunoreactivity was detected in 97% of active follicles. Only 6% of hypofunctioning follicles were positive (P < 0.001), 58% of them were totally negative, and 36% were partially labeled. In MNGs, the proportion of positive hypofunctioning follicles was slightly (but not significantly) increased, compared with normal glands (Fig. 3B).
|
Morphometrical analysis of the microvascular bed
Vv of the capillaries (Fig. 4A).
In normal thyroids, as well as in MNGs, Vv of capillaries located in close contact with active follicles was 4- to 5-fold larger than Vv of capillaries located around hypofunctioning follicles (P < 0.001). In addition, Vv of the capillaries of active follicles in MNGs was larger than that of active follicles in normal thyroids (0.23 ± 0.05 in MNGs vs. 0.125 ± 0.003 in normal thyroids, P < 0.001). Similarly, Vv of capillaries around hypofunctioning follicles was significantly larger in MNGs (0.06 ± 0.004 in MNGs vs. 0.03 ± 0.003 in normal thyroids, P < 0.001).
|
).
In both normal thyroids and MNGs, the number of capillaries surrounding hypofunctioning follicles was about half of that located around active follicles (P < 0.001). The number of capillaries around active follicles in MNGs was greater than that of the corresponding follicles in normal thyroids (0.378 ± 0.01 in MNGs vs. 0.254 ± 0.007 in normal thyroids, P < 0.01). The same conclusion was drawn for hypofunctioning follicles (0.187 ± 0.008 vs. 0.128 ± 0.007, P < 0.001).
Mean capillary area (Fig. 4C).
In normal thyroids, the mean area of capillaries surrounding hypofunctioning follicles was twice as small than that of capillaries surrounding active follicles (P < 0.001). In MNGs, the difference was of 3-fold (P < 0.001). The mean area of capillaries surrounding active follicles in MNGs was larger than that surrounding active follicles in normal thyroids (121.5 ± 2.3 vs. 82 ± 2 µm2, P < 0.001). The same observation was made for capillaries surrounding hypofunctioning follicles (43.4 ± 2.5 vs. 32 ± 2.3 µm2, P < 0.001).
Immunohistochemical analysis of angiogenic and vasoactive factors
VEGF.
In normal thyroids, a weak VEGF immunoreactivity was detected in epithelial cells of some follicles, others remaining negative regardless to their functional status (Fig. 5A). In MNGs, an intense immunoreactivity was detected at the basal pole of the cells (Fig. 5B). Nevertheless, the signal was heterogeneous within the same gland. VEGF was also detected around capillaries. In both normal and goitrous thyroids, smooth-muscle cells of arteries were positive, whereas endothelial cells remained negative (data not shown). When VEGF-positive follicles were counted, there was no significant difference between the two kinds of follicles or between normal thyroids and MNGs (Fig. 6A).
|
|
). All endothelial cells were negative. A similar observation was made in MNGs, although the endothelial cells of capillaries were strongly positive (inset in Fig. 5C). Counting of follicles showed that 94% of active follicles and 8.5% of hypofunctioning follicles were positive for NOSIII in normal thyroids (P < 0.001); 49% of hypofunctioning follicles were totally negative, whereas 42% contained few positive cells. There was no significant difference when compared with MNGs (Fig. 6B).
ET.
In normal thyroids and in MNGs, an immunoreactivity for ET was detected only in follicular cells of active follicles (Fig. 5D). When ET-positive follicles were counted, 90% of active follicles and 8% of hypofunctioning follicles were positive in normal thyroids (P < 0.001); 51% of hypofunctioning follicles were totally negative, and 41% contained only few positive cells. There was no significant difference when compared with MNGs (Fig. 6C).
Discussion
Hot and cold follicles represent two separate, but coexisting, functional entities within the thyroid gland. They have been well characterized in the mouse, by correlating the aspect of the epithelium with short time autoradiographic studies after iodine-125 administration (2, 3, 4). Cold follicles are lined with a flat epithelium, whereas the lumen is filled with a dense colloid. In contrast, hot follicles are lined with a cuboidal epithelium, but the lumen is usually filled with a paler colloid. Based on these morphological features, two categories of follicles, namely active and hypofunctioning follicles can also be identified in the human thyroid. To extend the observations made in experimental models to human thyroids for which autoradiographic studies are obviously lacking, we attempted to better characterize the functional status of follicles, thanks to the immunohistochemical analysis of proteins involved in the iodine transport and organification pathways.
In follicles defined as hypofunctioning, we found that neither TPO, nor pendrin, nor ThOXs are expressed in the epithelial cells. A similar observation has been recently reported (25). Noteworthy, the colloid was also devoid of Tg-I, suggesting that the absence of iodination activity was associated with the absence of proteins involved in this pathway. When an iodination activity was present in the so-called active follicles, as shown by the positive immunostaining for Tg-I in the colloid, the epithelial cells were positively immunostained for TPO, ThOXs, and pendrin. ThOXs were localized at the apical pole of the cells, confirming recent data by Caillou at al. (25). From these data, we postulate that each follicle in the human thyroid gland can be categorized as active or hypofunctioning, based on its morphological aspect but also on the detection of proteins involved in iodination (TPO, ThOXs, pendrin), or reflecting this process, as Tg-I.
Irrespective of the functional status, TSHr and Tg were detected in both types of follicles. This suggests that cells of hypofunctioning follicles are not totally inactive, because they still are sensitive to TSH. This is in accordance with previous observations made in mice. Indeed, the proportion of cold follicles decreases progressively when a goiter is experimentally induced, indicating that when TSH increases, cold follicles become active (22, 26). These data are in accordance with those of Maenhaut et al. (27) stating that TSHr is always present in thyroid epithelial cells, even if they are dedifferentiated.
The immunostaining for TPO was seen in the whole cytoplasm, although this protein is known to act at the apical pole of the cell. This agrees with previous studies showing that only a small fraction of TPO is associated with the cell surface (28, 29). Most of the TPO is detected by confocal analysis in the perinuclear region, whereas only a small fraction is seen on the apical membrane (30). In fact, intracellular TPO is misfolded and not correctly glycosylated (29). Our data indicate that the antibody used in the present study most likely recognizes a linear epitope that is common for the different molecular TPO configurations (29).
We previously showed, in a model of old goitrous mice, that the thyroid is composed of clustered angiofollicular units having their own local vascular regulatory system. In this model, the microvascular bed was closely related to the functional aspect of the follicles. For instance, Vv of capillaries surrounding hyperfunctioning follicles was larger than those surrounding hypofunctioning follicles (22). Results of the present work in normal and goitrous human thyroids where active and hypofunctioning follicles coexist are in the same line. The capillary network surrounding active follicles was larger than that surrounding hypofunctioning follicles, because of an increased area and number of capillaries. In addition, the vascular bed was always larger in MNGs than in normal thyroids.
As in the mouse, the immunoreactivity for NOSIII in normal and goitrous thyroids was restricted to follicular cells of active follicles, suggesting a paracrine role for NO in regulating the microvascular tone. In MNGs, endothelial cells of capillaries were also intensively stained, suggesting a possible role for NO in the increased vasculature observed in nodular lesions, because NO is known to act as a vasodilator (31). The detection of NOSIII exclusively in active follicles associated with a higher area of capillaries surrounding these structures led us to hypothesize that a deficit in NO production by epithelial cells of hypofunctioning follicles could be responsible for the reduced vasodilation of capillaries surrounding them. There was also a strong correlation between NOSIII and TPO immunostainings. These data strengthen the idea that NO exerts autocrine or paracrine actions in follicular cells, as already suggested by previous reports (32, 33, 34, 35, 36).
ET was also exclusively detected in active follicles. ET may act as a vasoconstrictive or vasodilative factor, depending on the activated receptor. Vasoconstrictive effects of ET are mediated via the ETA receptor, whereas the presence of the ETB receptor on endothelial cells is associated with vasodilation (37). In normal rat thyroid, the ETB receptor is the most abundant (38). Hence, the deficit in ET production in hypofunctioning follicles could also contribute to the reduced capillary bed observed around these follicles.
A weak VEGF immunoreactivity was observed only in few follicles of the normal human thyroid. This is different from our observations in old mice, where VEGF was detected in all follicles. However, as in the mouse, no difference in this immunoreactivity was observed between active and hypofunctioning follicles in the normal human thyroid. In MNG, VEGF immunostaining was clearly extracellular, suggesting the crinopexy phenomenon, e.g. the trapping of growth factors in association with the extracellular matrix, as previously reported (39, 40).
As for ET, VEGF is known to act as an angiogenic factor by altering the mobility and the proliferation of endothelial cells (41, 42, 43). Several reports showed that VEGF and ET-1 are unable to induce, separately, angiogenesis in vivo. They must be present together to influence the expression of their respective gene (44, 45). In addition, the detection of VEGF in the extracellular matrix was associated with a strong positive signal of NOSIII in endothelial cells. Because Vv of capillaries around both active and hypofunctioning follicles increased as a result of both expended area and number of capillaries, these data indicate that both vasodilation and angiogenesis are stimulated in goiters. One could suggest that VEGF induces NO production by endothelial cells, allowing angiogenesis (46, 47, 48), and thereby increasing the number of capillaries. On the other hand, enhanced NO production could induce vasodilation and therefore expend the capillary area. NO, ET1, and VEGF are all known to induce angiogenesis. Thus, NO modulates the endothelial cell matrix adhesion, thereby activating the first steps for the migration of endothelial cells when induced by VEGF. If NO production is blocked, VEGF is unable to induce endothelial cell migration and angiogenesis (46, 47, 48). A similar role for NO in ET-induced migration of endothelial cells has been described in vitro (49). During goiter formation in the rat, both NOSIII and ET increase and are associated with the expended vasculature (50, 38). Local defects in the production of one (or more) of these factors may therefore affect the function of the others, resulting in a local decrease in the number of capillaries, as observed around hypofunctioning follicles. Moreover, other factors known to act on vascularization and detected in thyroid cells (such as the fibroblast growth factor, the angiopoietins, and thrombospondin) could also affect the local microvascularization surrounding thyroid follicles (51, 52, 53).
In conclusion, the morphological and microvascular properties of the follicles are correlated with their function, in particular with their iodination capacity, as evidenced by the regulated expression of TPO, pendrin, ThOXs, and Tg-I. In addition, angiofollicular units having their own regulation are present in normal and goitrous human thyroids. They are controlled by vasoactive and angiogenic factors. This regulation involves complex regulatory networks including NO, VEGF, and ET.
Acknowledgments
We acknowledge Dr. I. Royaux for the gift of the antipendrin antibody, and Dr. J. Ruf for the anti-TPO antibody.
Footnotes
This work was supported by the National Fund for Scientific Research (Grants 3.4506.96 and 3.4516.98).
Abbreviations: ET, Endothelin; mAb, monoclonal antibody; MNG, multinodular goiter; NIS, Na+/I- symporter; NOS, nitric oxide synthase; PDS, pendred syndrome gene; Tg-I, T4-containing iodinated Tg; ThOX, thyroid oxidase; TSHr, TSH receptor; TPO, thyroperoxidase; VEGF, vascular endothelial growth factor; Vv, relative volume.
Received September 20, 2001.
Accepted November 15, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  A.-C. Gerard, S. Poncin, J.-N. Audinot, J.-F. Denef, and I. M. Colin Iodide deficiency-induced angiogenic stimulus in the thyroid occurs via HIF- and ROS-dependent VEGF-A secretion from thyrocytes Am J Physiol Endocrinol Metab, June 1, 2009; 296(6): E1414 - E1422. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A.-C. Gerard, S. Poncin, B. Caetano, P. Sonveaux, J.-N. Audinot, O. Feron, I. M. Colin, and F. Soncin Iodine Deficiency Induces a Thyroid Stimulating Hormone-Independent Early Phase of Microvascular Reshaping in the Thyroid Am. J. Pathol., March 1, 2008; 172(3): 748 - 760. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Krause, S. Karger, O. Gimm, S.-Y. Sheu, H. Dralle, A. Tannapfel, K. W. Schmid, C. Dupuy, and D. Fuhrer Characterisation of DEHAL1 expression in thyroid pathologies Eur. J. Endocrinol., March 1, 2007; 156(3): 295 - 301. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A.-C. Gerard, M. Boucquey, M.-F. van den Hove, and I. M. Colin Expression of TPO and ThOXs in human thyrocytes is downregulated by IL-1{alpha}/IFN-{gamma}, an effect partially mediated by nitric oxide Am J Physiol Endocrinol Metab, August 1, 2006; 291(2): E242 - E253. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Verri, C. Dimitri, S. Treglia, F. Storelli, S. De Micheli, L. Ulianich, P. Vito, S. Marsigliante, C. Storelli, and B. Di Jeso Multiple pathways for cationic amino acid transport in rat thyroid epithelial cell line PC Cl3 Am J Physiol Cell Physiol, February 1, 2005; 288(2): C290 - C303. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Faggiano, J. Coulot, N. Bellon, M. Talbot, B. Caillou, M. Ricard, J.-M. Bidart, and M. Schlumberger Age-Dependent Variation of Follicular Size and Expression of Iodine Transporters in Human Thyroid Tissue J. Nucl. Med., February 1, 2004; 45(2): 232 - 237. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A.-C. Gerard, C. Daumerie, C. Mestdagh, S. Gohy, C. de Burbure, S. Costagliola, F. Miot, M.-C. Nollevaux, J.-F. Denef, J. Rahier, et al. Correlation between the Loss of Thyroglobulin Iodination and the Expression of Thyroid-Specific Proteins Involved in Iodine Metabolism in Thyroid Carcinomas J. Clin. Endocrinol. Metab., October 1, 2003; 88(10): 4977 - 4983. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
) the epithelium is flat with visible flat nuclei; whereas in the second type, the epithelium is barely visible. Scale bar, 10 µm.
