| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Molecular Oncology (H.T., D.S.B.H., N.U.) and the Breast and Endocrine Program (N.C.G., N.M.H., A.E.G., F.R.S.), John Wayne Cancer Institute at Saint John’s Health Center, Santa Monica, California 90404
Address all correspondence and requests for reprints to: Frederick Singer, M.D., John Wayne Cancer Institute, 2200 Santa Monica Boulevard, Santa Monica, California 90404. E-mail: singerf{at}jwci.org.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Objective: The objective was to determine whether hypermethylation of either gene plays a role in their expression and 99mTc-sestamibi retention.
Design: This was a retrospective study on a convenience sample of paraffin-embedded parathyroid glands.
Setting: The study was performed at the John Wayne Cancer Institute at Saint John’s Health Center (Santa Monica, CA).
Patients: Forty-eight patients with primary hyperparathyroidism and five patients without parathyroid disease undergoing thyroid surgery provided 27 adenomatous, 10 hyperplastic, and 16 normal parathyroid glands.
Intervention: We performed immunohistochemistry, real-time quantitative RT-PCR, and methylation-specific PCR for MDR1 and MRP1 on archival parathyroid tissue and correlated these results with the patient’s 99mTc-sestamibi scan.
Main Outcome Measure: The main outcome measure was to determine whether hypermethylation of the genes for either transporter is associated with loss of their expression and with a positive 99mTc-sestamibi scan.
Results: The MDR1 gene was methylated in none of 12 normal glands, 19 of 27 adenomas, and three of 10 hyperplastic glands. Methylation of the MRP1 gene was uncommon (five of 48 tested glands). Methylation of the gene affected the transcript level only for MDR1. Among all glands, hypermethylation for MDR1 was more likely in 99mTc-sestamibi-positive scans (P < 0.001).
Conclusion: In parathyroid tissue, hypermethylation of the MDR1 gene decreases its expression and is associated with increased detection of parathyroid adenomas by 99mTc-sestamibi parathyroid scans.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The mechanism of the parathyroid scan appears to involve passive diffusion of the isotope across the cell membrane and then sequestration in mitochondria by a large negative transmembrane potential (5, 6, 7). Prolonged retention of the isotope is due to the functional loss of at least multidrug resistance 1 (MDR1), originally termed p-glycoprotein (8, 9, 10, 11, 12), and possibly to another protein, multidrug resistance-associated protein 1 (MRP1) (10, 12), both belonging to the large ATP-binding cassette (ABC) superfamily of membrane transporters (16). These transporters are better known for their ability to cause resistance to chemotherapeutic drugs. If present and active, as they are in normal parathyroid glands, they appear to extrude the isotope and produce a negative scan (8, 11).
However, not all studies agree on the role of MDR1 (13, 14) and MRP1 (9) in 99mTc-sestamibi retention. In addition, it appears that gland size and location may be important determinants of gland localization by 99mTc-sestamibi, with smaller and superiorly located glands more likely to escape detection (3). Furthermore, limitations of prior studies on the determinants of parathyroid 99mTc-sestamibi scan results are that few cases of primary hyperplasia (8) or only large (>1.5 g) adenomas (11, 12) were studied, thus excluding from analysis the glands most likely to have a negative scan.
The purpose of the present pilot study was to elucidate further the basis for 99mTc-sestamibi retention by abnormal parathyroid glands. First, we aimed to determine whether the mechanism of down-regulation of the MDR1 and MRP1 genes in patients with primary hyperparathyroidism is due to hypermethylation, one of the most common molecular mechanisms responsible for gene silencing in neoplasia (15). Second, if hypermethylation is responsible for the silencing of the gene for either MDR1 or MRP1, then does it help to predict the results of a sestamibi scan?
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
After approval by the Saint John’s Health Center/John Wayne Cancer Institute Institutional Review Board, paraffin blocks were obtained from a convenience sample of 48 patients who had surgery for primary hyperparathyroidism between 1999 and 2003 and had undergone a preoperative sestamibi scan (the latter became routine in August 1998). In 27 patients, a single parathyroid adenoma was removed. A balanced number of cases with positive (13) and negative (14) 99mTc-sestamibi scans were selected before analysis so that a comparison could be made. A range of adenoma weights from low (<200 mg) to high (>1000 mg) was intentionally included in the selection of both sestamibi-positive and sestamibi-negative patients. One parathyroid gland from each of 10 patients with primary parathyroid hyperplasia was studied. Only two of the 10 glands happened to be sestamibi-positive. Of the 16 normal glands, 11 were from a patient with primary hyperparathyroidism due to a single adenoma who had had a normal gland removed during neck exploration. In five cases, a normal gland (often intrathyroidal) was inadvertently removed during thyroid surgery. Care was taken to avoid specimens that may have had a false-positive scan due to coexisting thyroid pathology or a false-negative scan to an ectopic location. Of the abnormal glands that were sestamibi-positive, 12 glands were inferior and one was superior, whereas of those which were sestamibi-negative, 10 were inferior and four were superior. All patients had normal renal function, and all adenoma patients became normocalcemic after parathyroidectomy, confirming that the excised tissue was responsible for the patient’s hypercalcemia. Hyperplasia patients had three or more pathologically abnormal glands.
The normal glands in the patients with hyperparathyroidism were sestamibi-negative preoperatively, and it was assumed that normal glands taken from patients undergoing thyroid surgery would have been scan negative too.
Immunohistochemistry (IHC)
Expression of MDR1 and MRP1 in parathyroid specimens was assessed by IHC. Four normal glands were not available for IHC study because of insufficient tissue. Cut sections (5-µm) were deparaffinized in xylene, and the sections were incubated with polyclonal goat antihuman MDR1 antibody (Santa Cruz Biotechnology, Santa Cruz, CA) at a dilution of 1:1000 or polyclonal goat antihuman MRP1 antibody (1:1000; Santa Cruz Biotechnology) and kept at 4 C overnight. Pathologically verified melanoma specimens were used as positive controls. Negative control slides were treated with nonimmunized goat IgG under equivalent conditions and with no primary antibody. Staining was performed as previously described (16). Multiple fields of each specimen were evaluated in a blinded fashion. The IHC results for MDR1 and MRP1 were classified as positive if the staining pattern was either diffusely or heterogeneously positive, or negative if there was no staining.
RNA isolation
All parathyroid tissues were examined by standard histopathology with hematoxylin and eosin staining. For quantitative real-time RT-PCR (qRT) assay, 10 x 10-µm-thick sections were cut from each block of paraffin-embedded archival tissue (PEAT) parathyroid as previously described (17). The PEAT used was less than 5 yr old. A new sterile microtome blade was used for each tissue block cut. Sections were then deparaffinized and microdissected by laser capture microdissection (Arcturus Engineering, Mountain View, CA). Parathyroid tissue adjacent to thyroid tissue was microdissected on the slides after deparaffinization to prevent contamination by thyroid tissues. Parathyroid tissues were digested with proteinase K before RNA extraction using a modified protocol as described by Takeuchi et al. (17). Total cellular RNA from specimens was extracted, isolated, and purified as previously described (18). The RNA was quantified by UV spectrophotometry and the RiboGreen detection assay (Molecular Probes, Eugene, OR). Only tissue specimens that had a sufficient amount of quality RNA were used. Established cell lines and characterized PEAT tumor specimens were used as positive and negative controls for the qRT assay.
Multiple marker qRT assay
Primer and probe sequences for MDR1 and MRP1 were designed for qRT. The probes used were: MDR1, 5'-FAM-AGC ATT GAC TAC CAG GCT CGC CAA-BHQ-1–3'; MRP1, 5'-FAM-ATG GTC CTC ATG GTG CCC GTA AT-BHQ-1–3'; and glyceraldehyde-3-phosphate dehydrogenase (GAPDH), 5'-FAM-CAG CAA TGC CTC CTG CAC CAC CAA-BHQ-1–3' (Proligo, Denver, CO).
Moloney murine leukemia virus reverse-transcriptase (Promega, Madison, WI) was used for first-strand cDNA synthesis with both oligo-dT (GeneLink, Hawthorne, NY) and random hexamers (Roche, Chicago, IL). qRT assays were performed as previously described (17). Briefly, each PCR mixture consisted of cDNA from 250 ng of total RNA. Samples were amplified with a precycling hold at 95 C for 10 min, followed by 45 cycles of denaturation at 95 C for 1 min, annealing at 55 C for 1 min (annealing at 60 C for MDR1 and MRP1), and extension at 72 C for 1 min using the iCycler iQ RealTime PCR detection system (Bio-Rad Laboratories, Hercules, CA).
Normal peripheral blood lymphocytes from noncancer patients and negative controls (reagents without RNA or cDNA) for qRT were included in each assay. Expression of the housekeeping gene GAPDH served as an internal reference for PEAT mRNA integrity (17). Positive and negative controls for RNA extraction, RT, and PCR procedures were performed in all assays. Each assay was performed in duplicate, and the mean copy number was used for analysis.
Standard curves for each gene were generated by using the threshold cycle of dilutions of known number cDNA templates. The mRNA copy number was calculated using the RealTime Detection System Software (Bio-Rad Laboratories). MDR1 or MRP1 mRNA copy number was generated by the ratio with GAPDH mRNA copy number; MDR1/GAPDH x 10–4 and MRP1/GAPDH x 10–3 were used as MDR1 and MRP1 detection values as previously described (16).
MDR1 and MRP1 promoter region methylation analysis
Methylation-specific PCR (MSP) after sodium bisulfite modification was performed for assessing the methylation status of the MDR1 and MRP1 gene region (19, 20, 21, 22). Five samples from normal glands were not available for MSP because of insufficient tissue. MSP was performed using fluorescent-labeled methylation and nonmethylation-specific primers. Primers used for amplification were as follows: methylated MDR1 forward primer, 5'-TTT GGA ACG GTT ATT AAG ACG T-3', and reverse primer, 5'-AAA CTC AAA AAA ACA AAA ACC G-3'; unmethylated MDR1 forward primer, 5'-TTT TGG AAT GGT TAT TAA GAT GT-3', and reverse primer, 5'-AAC TCA AAA AAA CAA AAA CCA CT-3'; methylated MRP1 forward primer, 5'-TTT TCG GAA GGC GAG TTA AC-3', and reverse primer 5'-TCT CGA CGT AAA CAA CCG AA-3'; unmethylated MRP1 forward primer, 5'-TTTT GGA AG G TGA GTT AAT GT-3', and reverse primer 5'-TCT CAA CAT AAA CAA CCA AA-3'. SssI methylase-treated and untreated normal DNA from peripheral blood lymphocytes was used as the positive and negative control, respectively. Detection of PCR products was analyzed by CEQ 8000XL capillary array electrophoresis system and CEQ 8000XL software version 6.0 (Beckman Coulter Inc., Fullerton, CA) as previously described (22).
Statistical analysis
To assess the association between mRNA marker(s) and clinical and pathological parameters, Student’s t test was used for continuous variables. Fisher’s exact test and 
2 test were used for categorical variables. Spearman correlation coefficient analysis was used to assess the relation between baseline patient characteristics and between copy levels of MDR1 and MRP1 mRNA. A univariate and multivariate analysis was performed on variables relevant to sestamibi results. All P values were assessed as two-sided and significant at 
0.05.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The serum calcium and PTH of patients with primary hyperparathyroidism were 10.8 ± 0.7 mg/dl (normal, 8.5–10 mg/dl) and 132 pg/ml (normal, 10–65 pg/ml), respectively. Among the abnormal glands chosen for study, the adenomatous glands weighed more (median, 400 mg) than the hyperplastic glands (median, 125 mg). 99mTc-sestamibi-positive cases had a higher serum PTH (191 ± 21 vs. 113 ± 20; P = 0.01), and their adenoma weighed more (median, 520 mg) compared with 99mTc-sestamibi-negative cases (median, 260 mg). Detection by 99mTc-sestamibi occurred in none of the normal glands removed from scanned patients with hyperparathyroidism, in 48% (by design) of the adenomatous glands, and in 20% (by chance) of the hyperplastic glands.
MDR1 and MRP1 expression (IHC) in different gland types
IHC evidence of both MDR1 and MRP1 transporters was found in 12 of 12 normal parathyroid specimens with sufficient tissue for study. In the 10 hyperplastic glands, MDR1 was detected in eight and MRP1 in nine. In the 27 adenomas, MDR1 was detected in eight and MRP1 in eight, with five glands being positive for both. There was a modest agreement in detection of MDR1 and MRP1 transporter proteins by IHC (for all glands, 
= 0.62, P < 0.001; for adenomas alone, = 0.47, P = 0.01).
MDR1 and MRP1 expression (mRNA) in different gland types
The assay for mRNA copy number for each gene was first validated by comparing the mRNA copy number for each gene in tissues that stained positive for transporter protein with those that stained negative. For all glands, the copy number for MDR1 was higher in IHC-positive than in IHC-negative tissues: mean ± SE MDR1 mRNA/GAPDH mRNA, 93 ± 13 x 10–4 vs. 35 ± 15 x 10–4 (P = 0.005). The same was true for MRP1: mean ± SE MRP1 mRNA/GAPDH mRNA, 898 ± 85 x 10–3 vs. 313 ± 102 x 10–3 (P < 0.001).
As shown in Table 1
, ABC transporter mRNA levels were highest in normal parathyroid glands, intermediate in hyperplasia, and lowest in adenomas. Expression of MDR1 and MRP1 mRNA copies was correlated (r = 0.6; P < 0.001).
|
There was no evidence of hypermethylation of MDR1 in 11 of 11 normal glands, whereas hypermethylation was found in three of 10 (30%) hyperplastic and 19 of 27 (70%) adenomatous glands (Table 1
). In contrast, hypermethylation of MRP1 was uncommon and was found in only one of 11 (1%) normal glands, none of 10 hyperplastic glands, and four of 23 (17%) adenomas.
To determine whether gene hypermethylation was associated with gene silencing, mRNA copy number and IHC staining for each transporter were compared in tissues in which the transporter gene was and was not hypermethylated (Table 2). Among all gland types, the mRNA for MDR1 was 3.8 times higher in those glands in which the gene was unmethylated compared with those in which the gene was hypermethylated. Similarly, positive IHC for MDR1 was more likely in glands in which MDR1 was unmethylated compared with those in which it was hypermethylated (22 of 26 vs. 5 of 22; P < 0.001). In contrast, there was no obvious relation between hypermethylation of MRP1 and either the level of MRP1 mRNA or positive staining on IHC for MRP1 protein (26 of 43 vs. 2 of 5; P = 0.64).
|
To determine whether the loss of expression of transporter genes paralleled gland enlargement, the expression of ABC transporter genes and their hypermethylation was assessed in small (200 mg), intermediate (201–599 mg), and large (>600 mg) glands (Table 3). Among all glands that could be evaluated, positive IHC staining for both MDR1 and MRP1 tended to be higher in smaller compared with larger glands. Hypermethylation of MDR1 also was less likely in smaller compared with larger glands, but for MRP1, hypermethylation of the gene was uncommon and appeared to be unrelated to gland size.
|
Among all abnormal parathyroid glands, the 99mTc-sestamibi scan was more likely to be positive in larger glands. For glands weighing 200 mg or less, 201–599 mg, and 600 mg or more, the proportion of positive scans was one of 28 (4%), seven of 13 (54%), and seven of 12 (58%), respectively. However, there were exceptions to this pattern: one hyperplastic gland weighing 161 mg was scan-positive, and five adenomas that weighed 775, 825, 951, 1245, and 2352 mg were scan negative.
Relative importance of MDR1 vs. MRP1 IHC for results of sestamibi scans
Abnormal glands that were IHC-positive for only MDR1 or MRP1, for both, or for neither were analyzed for sestamibi retention to gain insight into the relative importance of the two transporters for scan results. This analysis was limited to glands more than 200 mg since very small glands may not be visualized simply because of the limitation of the imaging system. At our center, the smallest gland a gamma-detector can detect has been estimated to be approximately 200 mg (23). In the three cases in which only MDR1 was detected, all were scan-negative, whereas in the two cases in which only MRP1 was detected, both were scan-positive. In an analysis of IHC results in the 13 largest glands (520–3605 mg), the status of MDR1 IHC alone explained 11 (85%), whereas MRP1 alone or combined MDR1 and MRP1 results were less predictive.
MDR1 vs. MRP1 expression, hypermethylation, and 99mTc-sestamibi accumulation
To determine the role of MDR1 and MRP1 expression and hypermethylation in sestamibi scan results, we compared IHC for each transporter and hypermethylation of each gene in patients with positive and negative sestamibi scans (Table 4). This comparison was made for all abnormal glands and for adenomas alone.
|
To examine the importance of hypermethylation of MDR1 in explaining the results of sestamibi scanning further, an additional analysis was performed on the 13 largest glands (12 adenomas and one hyperplastic gland weighing >500 mg). The hypermethylation status of the MDR1 gene predicted the sestamibi scan result in 11 (85%; eight hypermethylated glands that were IHC-negative and scan-positive, and three unmethylated glands that were IHC-positive and scan-negative). Two glands weighing 951 and 1245 mg were hypermethylated but scan negative. These were the superior glands noted above; the former gland was actually IHC positive.
In contrast, although MRP1 mRNA tended to be higher and IHC for MRP1 was more likely to be positive in glands that were 99mTc-sestamibi-negative, there was no relation between scan results and hypermethylation of the MRP1 gene.
Relative importance of gland activity, gland size, and transporter expression
Because the sestamibi scan was more likely to be positive in glands from cases that had a higher serum PTH, a larger abnormal gland, and a gland in which ABC transporter was not expressed, an analysis was performed to determine the relative importance of each factor. This analysis was restricted to adenomas that were larger than 200 mg. The analysis did not include hyperplastic glands because multiple glands contribute to the total weight of abnormal tissue and serum PTH in patients with hyperplasia.
First, a univariate nominal logistic regression analysis including eight studied variables possibly related to sestamibi retention in 23 adenomas was performed. Variables included serum PTH, gland weight, and the IHC status, mRNA level, and methylation status for each transporter. The most significant variable was MDR1 hypermethylation (P = 0.004). Neither weight nor any of the MRP1 variables were significant; the significance of serum PTH was borderline (P = 0.09). In a multivariate regression model including MDR1 hypermethylation and serum PTH, the ß-coefficient ± SE (0.35 ± 0.09) for hypermethylation was highly significant (P = 0.007) for predicting sestamibi scan results whereas serum PTH was not (0.002 ± 0.001; P = 0.09).
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Although we also found evidence for decreased transcription of the MRP1 gene in abnormal parathyroid glands, hypermethylation of this gene was found in only five glands (four of 27 adenomas and one of 11 normal glands). Future studies will examine the mechanism(s) accounting for the down-regulation of MRP1 expression.
Our results agree with findings of previous studies on the determinants of detection of parathyroid adenomas by sestamibi scans. We found that adenomas that are associated with a higher serum PTH are larger and have undetectable MDR1 transporter protein (with or without the loss of MRP1 transporter protein) by IHC are more likely to have a positive scan.
This study also extends our current understanding of the role of transporter proteins in parathyroid scanning by extending it to primary hyperparathyroidism due to hyperplasia. We found that both MDR1 and MRP1 are likely to be present and their respective genes unmethylated in primary hyperplasia. In addition to the small size of hyperplastic glands, the activity of ABC transporters may help to explain why hyperplastic glands are more likely to be sestamibi negative compared with adenomas.
Although this study was not designed to test the relative importance of the loss of one transporter vs. the other, there were several suggestions that the loss of MDR1 may be the more important determinant of a positive scan: several glands that were positive for MRP1 but negative for MDR1 were sestamibi-positive, whereas the reverse was not true; using IHC, results for MDR1 alone were the best predictor of scan results; and MRP1 variables (IHC, mRNA, and methylation) were not predictive of scan results in a univariate analysis.
Although lack of detection of MDR1 by IHC with or without the lack of detection of MRP1 seemed to provide a sufficient explanation of why glands retain sestamibi and become scan-positive, the reason why scans may be negative is not as straightforward. There were a few glands that had no detectable MDR1 and/or MRP1 and yet had negative scans. In some cases, failure to detect such a gland could be explained by its size being below the resolution of the imaging system. However, in other cases a negative scan could not be ascribed to small gland size because there were a few large glands that had undetectable transporter proteins and yet were scan-negative. Other explanations for why such glands fail to be visualized could possibly be that IHC is not sensitive enough to pick up low but physiologically important levels of transporter protein, methylation status may not be the only mechanism of genomic DNA silencing, sampling error (if the transporter is not evenly distributed within the tumor), the involvement of other transporters, or superior location of the glands.
If the level of ABC transporter proteins could be reversibly decreased in pathological parathyroid glands, then they could possibly be made to retain 99mTc-sestamibi reliably. This would be especially helpful for patients who have not responded to prior parathyroid surgery and have a negative 99mTc-sestamibi scan because in these cases the culprit gland is notoriously difficult to locate at surgery even by an experienced surgeon. There are drugs that can modulate these transporter proteins (26), but clinical trials would be needed to determine whether their inhibition could improve the sensitivity of 99mTc-sestamibi scans without causing an unacceptable decrease in their specificity.
|
Footnotes |
|---|
Disclosure: The authors have no conflict of interest to report.
First Published Online February 13, 2007
Abbreviations: ABC, ATP-binding cassette; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; IHC, immunohistochemistry; MDR1, multidrug resistance 1; MRP1, multidrug resistance-associated protein 1; MSP, methylation-specific PCR; PEAT, paraffin-embedded archival tissue; qRT, quantitative realtime RT-PCR; 99mTc-sestamibi, technetium-99m-sestamibi.
Received September 13, 2006.
Accepted February 5, 2007.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
probe-guided parathyroidectomy. Arch Surg 139:1175–1179
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
