| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
INNOVATIVE METHODOLOGY
1KineMed, Inc., Emeryville; 2Department of Nutritional Sciences and Toxicology, University of California at Berkeley, Berkeley; and 3Division of Endocrinology and Metabolism, Department of Medicine, San Francisco General Hospital, University of California at San Francisco, San Francisco, California
Submitted 16 June 2007 ; accepted in final form 16 August 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
50% of islet cells divided in rats between 8 and 15 wk of age. Of interest, long-term BrdU administration suppressed proliferation of islet cells significantly, but not of bone marrow cells. Physiological evidence further supported the validity of the method: older animals (24 wk old) had 60% lower islet cell proliferation rates than younger rats (5 wk old), and partial (50%) pancreatectomy increased proliferation by 20%. In addition, cholecystokinin-8 treatment significantly stimulated proliferation in pancreatectomized rats only. In summary, heavy water labeling is a quantitative approach for measuring islet cell proliferation and testing therapeutic agents.
diabetes; 
-cell; 5-bromodeoxyuridine; sulfated cholecystokinin-8
-cells is increasingly recognized as a key factor in the pathogenesis of diabetes mellitus (type 1 and type 2) (10, 35). The recent development of agents that may stimulate -cell proliferation and increase -cell mass has focused further interest on -cell regeneration as a therapeutic target in diabetes (2, 13).
Availability of sensitive and efficient techniques for measuring in vivo the proliferation rate of pancreatic -cells are therefore important for investigations into the pathogenesis and therapy of diabetes. The most widely used technique for measuring proliferation rates of -cells is through labeling with 5-bromodeoxyuridine (BrdU). This approach has been used for many years and has generated important insights into -cell dynamics during the progression to diabetes (3, 14, 31).
There are some limitations to the BrdU labeling method, however. Previous reports have demonstrated that BrdU has cytotoxic effects (26, 27) and have reported the impact of chronic BrdU exposure on proliferative capacity of various cells, such as hepatocytes, kidney and immune cells, in mice and rats (21, 32, 36). This approach is also relatively labor intensive, which limits throughput. Genetic analysis (e.g., cell proliferation as a quantitative trait) (4) and drug screening (2, 13, 33) have therefore been constrained.
We recently developed a technique for measuring proliferation rates of cells in vivo, based on heavy water (2H2O) labeling and mass spectrometric analysis (28; Fig. 1, A and B). This approach is particularly well suited for measurement of proliferation of slow-turnover cells, due to the ease of long-term administration of heavy water and the high sensitivity of mass spectrometric analysis (19). Slow-turnover cells for which dynamics have been measured by this approach include vascular smooth muscle cells, adipocytes, T lymphocytes, hippocampal neurons, hepatocytes, endothelial cells, mammary epithelial cells, spermatocytes, chronic lymphocytic leukemia cells, and others (8, 20, 34). Here, we asked whether the heavy water labeling approach can be used for pancreatic islet cells as a model to study -cell biology, in that 70–80% of islet cells are -cells, and the -cell exhibits relatively slow turnover. We also compare its use to the standard method, BrdU labeling.
|
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
Male Wistar rats (Charles River Laboratories) were maintained on an ad libitum diet with commercial chow. Other male rats underwent 50% pancreatectomy (Px) or sham operation at 8 wk old (at Charles River Laboratories) and then were administered cholecystokinin octapeptide (CCK-8; NeoMPS, Strasbourg, France) at 4 µg/kg subcutaneiously three times daily or vehicle (15% gelatin) for 14 days beginning the 5th day after surgery (22). Procedures were approved by the Institutional Animal Use Committees.
Heavy water labeling protocol.
A priming intraperitoneal bolus (0.35 ml/g body wt) of 99% 2H2O in 0.9% NaCl was given to reach a body water enrichment of roughly 5% (using an estimated 60% body weight as water) and then received 8% 2H2O in drinking water for the study duration. Animals were killed after different time points of 2H2O intake. The 2H2O enrichments in body water achieved stable steady-state values after 3 days of labeling in rodents on this protocol (28).
Pancreatic islet purification.
The islet isolation technique has been described in detail previously (9). Briefly, 30 ml of cold Hanks' buffer with collagenase solution was infused into the pancreatic duct. The inflated pancreas was digested in a water bath at 37°C. Islets were picked out by hand. DNA was extracted from all islets in each animal for measurement of 2H incorporation to avoid sampling bias.
Fresh live islets isolated by this technique, including islet cells were immunostained with anti-insulin antibodies (see Supplementary Fig. 1 online).
Bone marrow cell isolation.
Bone marrow cells were flushed out with a needle and syringe containing PBS from the femur, as described previously (28).
Enrichment and incorporation of 2H analysis in cell DNA.
The incorporation of 2H from 2H2O into the deoxyribose (dR) moiety of purine deoxyribonucleotides in genomic DNA was measured as described previously (28). In brief, DNA was extracted from islets by using Qiagen kits and was hydrolyzed to deoxyribonucleosides. DNA was heated with magnesium chloride and zinc sulfate, followed by incubation for 2 h in a 37°C water bath with DNase, nuclease P1, snake venom phosphodiesterase, and alkaline phosphatase. The dR moiety was derivatized to pentane tetraacetate, as described (28). Pentane tetraacetate was analyzed by positive chemical ionization GC-MS with a model 5973 mass spectrometer and model 6890 gas chromatograph (Agilent, Palo Alto, CA). Selected ion monitoring was performed with mass-to-charge ratios (m/z) of 245 and 246, representing the M + 0 and M + 1 ions, respectively. The excess fractional M + 1 enrichment (EM1) of dR was calculated as
 |
|
The fraction of newly divided islet cells was calculated from the ratio of 2H incorporation into DNA from islets to 2H incorporation into bone marrow cells (20, 28):
 |
Comparison of 2H2O labeling alone to the double-labeling, 2H2O-BrdU approach.
To examine the effect of BrdU on islet cell proliferation, rats were labeled continuously for periods of 7, 14, and 28 days with 2H2O alone or double-labeled with 2H2O-BrdU in parallel studies. The 2H2O labeling protocol was performed as described above. BrdU (1 mg/ml), was added to 2H2O in drinking water to create a 2H2O-BrdU solution. To ensure that BrdU was bioactive, the 2H2O-BrdU solution was protected from light exposure and was changed every 3 days with freshly prepared solution, as described previously (31). Islet cell proliferation was measured by mass spectrometry, as described above. Additionally, a portion of pancreatic segments in the double-labeled 2H2O-BrdU animals was fixed in 4% paraformalin for immunohistochemistry to explore the correlation between the 2H2O labeling and BrdU staining approaches.
BrdU/insulin immunohistochemistry.
Consecutive sections (5 µm) were double stained for BrdU/insulin, according to the manufacturer's instructions. Briefly, insulin was stained with a polyclonal guinea pig anti-porcine insulin primary antibody, followed by a peroxidase-labeled secondary antibody, developed with 3,3'-diaminobenzidine (DAB) and counterstained with hematoxylin. BrdU was detected with a mouse monoclonal anti-BrdU antibody followed by a peroxidase-conjugated anti-mouse immunoglobulin G; DAB with nickel enhancer was used as the chromogen (Cell Proliferation Kit; Amersham Life Science, Arlington Heights, IL). On these sections, the insulin-positive cells (-cells) had a red-brown cytosolic staining, and BrdU/insulin-positive cells (new growth -cells) had darker nuclei with red-brown cytosolic staining (see RESULTS). Individual animal results represented an average 41 islets counted (range 27–68), comprising on average 4,802 -cells (range 1,652–7,520), read by one observer (E. Tsang). Data were expressed as the percentage of BrdU/insulin-positive cells per total number of -cells.
Statistical analysis.
All data are expressed as means ± SE. For comparisons between two groups, the unpaired Student's t-test (two-tailed) was applied. For multiple comparisons, one-way analysis of variance was used. Pearson's correlation coefficient was use to determine the correlation between quantitative variables. A P value of <0.05 was considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
The time course of islet cell proliferation is shown (Fig. 1C). Labeling with 2H2O was begun at 8 wk of age and continued for up to 7 wk duration. Islet cell proliferation approached an apparent plateau value of 55% new cells during the 7-wk period in these growing rats. Because some smaller islets might be newer or represent recent buds from ducts (5), we determined whether smaller islets had higher growth rates than larger islets. The fraction of newly divided cells in smaller islets (<150 µm in diameter) was compared with larger islets (>150 µm in diameter) isolated from the same animal after a 14-day labeling period. There was no difference between the two groups (smaller islets 17.3 ± 2.9% vs. larger islets 17.1 ± 3.3%; Fig. 1D) (6).
The most serious practical problem in dynamic studies of slow-turnover tissues is contamination by small amounts of a high-turnover tissue (20). Islets purified by the approach that we used contain less than 5% exocrine tissue (9, 15; and Supplemental Fig. 1). Nevertheless, we also measured the proliferation rate of cells from exocrine tissue (digested acini and ducts) from the same animals for a 14-day labeling period to estimate the potential impact of contamination of islets by exocrine tissue. Proliferation rates of exocrine tissue were considerably lower than for islet cells (6 ± 0.3% in exocrine tissue vs. 16 ± 2.2% in islet cells; Fig. 1E). Contamination by small amounts of exocrine tissue would therefore be unlikely to have affected results substantially, even if present.
Correlation between fraction of newly divided islet cells by 2H2O labeling and percentage by BrdU immunohistochemistry.
To validate the reliability of the 2H2O labeling approach, we compared the fraction of newly divided cells by 2H2O labeling and the percentage of BrdU/insulin-stained cells from the same rats, which had undergone a double-labeling 2H2O-BrdU protocol. The fraction of newly divided islet cells by 2H2O labeling was highly correlated with the percentage of BrdU/insulin staining (r = 0.9581, P < 0.001; Fig. 2A) with no significant differences (9.3 ± 1.3% vs. 8.9 ± 0.8%, 7 days, NS; 18.5 ± 1.6% vs. 20.3 ± 2.0%, 14 days, NS, respectively; Fig. 2B).
|
We considered the possibility that BrdU administration might alter proliferation rates of islet cell during continuous administration (21, 32, 36). The results from 2H2O labeling alone were compared with double labeling with 2H2O-BrdU (Fig. 3A). The fraction of newly divided islet cells was significantly higher with 2H2O labeling alone than with double labeling with 2H2O-BrdU (12.7 ± 1.2% vs. 9.3 ± 1.3%, 7 days, P < 0.01, 22.2 ± 3.1% vs. 18.5 ± 1.6%, 14 days, P < 0.05, 31.2 ± 1.8% vs. 26.3 ± 4.2%, 28 days, P < 0.05; respectively,). Thus, long-term BrdU administration suppressed islet cell proliferation by 16–25%. BrdU did not affect the turnover of bone marrow cells (Fig. 3B). Long-term BrdU administration appeared to have little effect on body weight, food intake, glucose, or organ weights (Table 2), consistent with reports by others (31).
|
|
Two physiological experiments were performed that provided validation of the heavy water labeling approach. First, the effects of age on the proliferation dynamics of islet cells was determined (Fig. 4A). In younger rats (5 wk old at the beginning of the labeling period), islet cell proliferation rates were 20.8, 42.4, and 55.7% for 7, 14, and 28 days, respectively. In older rats (24 wk old at the beginning of the labeling period), islet cell proliferation rates only were 4.7, 11.3, and 21.3% for 7, 14, and 28 days, respectively. Thus, older rats had more than a 60% reduction in the fraction of newly divided islet cells compared with younger animals.
|
20% elevated compared with sham-operated, age-matched rats (Fig. 4B). This increase in proliferating fraction was observed after both 14 days and 28 days of 2H2O labeling (15.8 ± 1.8 vs. 12.8 ± 1.7% for Px vs. sham-operated, 14 days, P < 0.05, 26.1 ± 1.6% vs. 20.5 ± 4.7%, 28 days, P < 0.05, respectively). We also tested the effects of CCK-8, a peptide agent with known insulin-secretory agonist effects (22, 29). CCK-8 was given to sham-operated and Px rats (the study design, Fig. 4C). No effect was observed in the sham-operated controls. Px plus CCK-8 significantly increased islet cell proliferation further, although modestly (by 13%), compared with Px alone (Fig. 4D). All groups remained euglycemic. No differences in plasma insulin concentrations were measurable among groups (not shown). |
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
-cells in pancreatic islets. The goal was to have a method that would complement the BrdU labeling approach and perhaps has advantages, such as higher throughput (1, 2, 34). Several technical points deserve comment. The results shown herein highlight the sensitivity and quantitative reproducibility of heavy water labeling and GC-MS analysis. The results of heavy water labeling correlated well with BrdU administration (Fig. 2). Heavy water can be administered without toxicity for several weeks, allowing integrated proliferation rates to be measured over long time periods. Indeed, we (23) had shown previously that 2H2O and BrdU gave similar results for colonocyte proliferation, although with superior precision and higher throughput for the 2H2O method. We have discussed elsewhere (28) the interpretation problems inherent in use of labeled pyrimidine nucleosides like BrdU to measure cell proliferation rates, due to variability of label entry into precursor pools.
The availability of the heavy water labeling method also allowed us to directly assess the impact of BrdU administration on proliferation rates of various cell types, including pancreatic islet cells (Fig. 3A), through double-labeling, 2H2O-BrdU experiments. We observed that long-term BrdU administration significantly reduced the proliferation of islet cells. These findings are consistent with previous reports of cells other than -cells (21, 32, 36), although others have reported no adverse effects of BrdU on -cell proliferation (31). Interestedly, BrdU administration did not affect the turnover of bone marrow cells, which may reflect the fact that the bone marrow is a rapid-turnover tissue (30). In any case, the possibility of an interaction between BrdU administration and cell type or drug candidates is of some concern (16, 21, 26).
There are also methods for evaluating cell division rates that do not involve metabolic labeling [e.g., staining for Ki67 or proliferating cell nuclear antigen (17, 37)]. These methods reflect the fraction of cells that are "in cycle" but do not reveal the rate at which cell division is actually occurring in a population. However. If a G1/S-phase block is present, for example, cells will accumulate that are positive for these cell cycle antigens even though the true rate of mitosis and passage through S-phase of the cell cycle is low.
The heavy water labeling approach described herein has disadvantages as well as advantages compared with BrdU staining. Because pancreatic islet cells, not purified -cells, are isolated by our protocol, the proliferation rate of all islet cells is assumed to represent that of -cells and it is assumed that there are no highly proliferative contaminating cells present. These assumptions are reasonable, in that 70–80% of islet cells are -cells, exocrine tissue is not present (and has a low proliferation rate), and proliferating glucagon-, somatostatin-, or pancreatic polypeptide-containing cells in islets are rare in adulthood (11, 18). The correlation between BrdU and 2H2O results that we observed (Fig. 2A) also supports this conclusion. Nevertheless, BrdU labeling directly visualizes -cell proliferation and would be advantageous in any circumstance where other cells in the islet might confound proliferative results. In the future, isolation of -cells by cell sorting could resolve concerns about non -cells in islets measured by the heavy water labeling approach.
The heavy water labeling approach detects all input sources into new -cells, including neogenesis and proliferation, although it is unable to distinguish among these sources. The possibility of quantitative artifacts arising from smaller, new islets is unlikely, however, based on our finding that smaller and larger islets exhibited the same fractional proliferation rates (6). Thus, as long as the neogenic islets can be isolated physically as an islet after a 2- to 4-wk labeling period, measured proliferation results should not be biased. The proliferative capacity to include neogenic islets in the mass balance equation therefore allows true net cell growth to be calculated (2).
This approach is compatible with relatively high efficiency for an in vivo technique (5). GC-MS measurements can be reliably performed from a small number of islets, if necessary, which would allow a larger number of animals to be studied. Preparation of DNA and GC-MS analyses can be batched and performed at high efficiency (e.g., >50 samples per day by a technician).
Type 2 diabetes is characterized by a deficit in -cell mass and by an increasing incidence with age. We confirmed previous findings (12, 25) of an age-dependent decline in -cell proliferation rates (Fig. 4A). Our results with CCK in the partial pancreatectomy model represent a pilot application of a drug testing strategy. Use of heavy water labeling has also recently been used to identify strain effects on islet cell proliferation in mice (unpublished observations).
In summary, heavy water labeling represents a convenient, reliable, nontoxic, and highly efficient alternative to BrdU labeling for measurement of proliferation rates of pancreatic islet cells and thus -cells. Applications to the pathophysiology and treatment of diabetes may prove useful.
|
DISCLOSURES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
