Experimental Renal Failure and Iron Overload: A Histomorphometric Study in Rat Tibia

Abstract

Renal failure (RF) is a serious disease of relatively high incidence, known to cause bone alterations. RF patients frequently suffer anemia, which is usually treated with iron. Given that iron overload inhibits bone formation, the aim of the present study was to evaluate the effect of iron on the subchondral bone of rat tibiae, using a model of renal failure. Male Wistar rats were subjected to experimental nephrectomy in order to induce renal failure and to iron overload by daily intraperitoneal injections of 88 mg/kg body weight of iron-dextran for 16 days. Tetracyclines were injected intraperitoneally to evaluate dynamic parameters of bone. Undecalcified histological sections of the tibiae were obtained. Serum urea, creatinine, and paratohormone (PTH) levels were evaluated 30 days after the onset of the experiment. Static and dynamic histomorphometric measurements were performed. Iron overload modified the response of the animals with renal failure: a reduction in bone forming activity compatible with adynamic bone disease and a decrease in peritrabecular fibrosis were observed. Our results suggest that iron is yet one more factor involved in the imbalance in bone metabolism typically found in renal failure patients treated with iron, rendering diagnosis and treatment of bone disease in these patients more complex.

Introduction

Bone disease induced by renal failure is characterized by complex alterations of the bone remodeling process. Bone disorders include fibrous osteitis, osteomalacia, osteosclerosis, delayed growth, and osteoporosis. The association among renal disease, skeletal disease, hyperplasia of the parathyroid gland and altered Vitamin D metabolism is well documented (Slatopolsky and Coburn, 1990). Most treatments conventionally administered to patients with chronic renal failure expose them to intoxication with different metals. Dialysis patients in developing countries have been found to exhibit high levels of aluminum, strontium, and iron (Afifi, 2002). Aluminum accumulation on bone mineralization fronts and its consequences has been widely addressed. Two characteristic disorders arise, i.e., osteomalacia in 95% of these patients (Ellis et al., 1979) and adynamic bone disease. Adynamic bone disease (ABD) was first described in the 1980s and is characterized by reduced or no cell activity and low turnover (Malluche and Faugere, 1990). Although its etiology is unknown, the disease has been associated to low levels of parathormone (PTH). Anemia is also associated to renal failure. Thus, many patients are treated with iron compounds and/or blood transfusions. The finding of iron deposits on bone surfaces of these patients has led some authors to suggest that iron is an etiologic factor of bone disease (Pierce-Mykli and Pierides, 1984; Ackrill et al.,1988). However, available data on the effects of iron toxicity on renal failure patients is scanty and do not refer to endochondral ossification.

Renal failure is a chronic disease that often entails many years of treatment. During this lengthy period, patients may require different treatments. From this perspective, further knowledge about bone dynamics gains particular significance.

The combination of renal failure and iron overload induces adynamic bone disease with reduced bone volume in the alveolar bone of rats (Mandalunis et al., 2002). The effect of renal failure and iron overload on long bones has not been studied to date. Based on the above, the aim of the present study was to evaluate the effect of iron overload on endochondral ossification in animals with experimental renal failure, by means of histomorphometric studies.

Materials and Methods

Fifty-six Wistar rats, 200 g body weight, were assigned to one of 4 groups. The animals were fed a normal protein diet and water ad libitum. Housing conditions included galvanized wire cages, 6 animals per cage, temperature 21–24°C, humidity 52–56%. Light dark cycle was 12 hours to 12 hours. National Institute of Health Guidelines for the Care and Use of Laboratory Animals (NIH publication 85-23 Rev. 1985) were observed.

Renal Failure

Renal failure (RF) was induced under ether anesthesia following standard procedures (Kaiser et al., 1984). Partial nephrectomy (PN) by removing two-thirds of the right kidney at the onset of the experiment (day 1) and total nephrectomy (TN) of the left kidney on day 7.

Intoxication with Iron

Iron overload (Io) was induced following the protocol previously described by de Vernejoul et al. (1984). Beginning on day 14, a daily dose of 88 mg/kg of iron dextran (a compound of iron hydroxide dextran) (Sigma D-8517) was intraperitoneally injected during 16 days.

Tetracycline Labeling

All animals were injected intraperitoneally with tetracyclines diluted in saline solution on days 14 and 28 of the experimental period to evaluate dynamic bone parameters. A 50 mg/kg body weight dose of oxytetracycline (Terramicyn, Pfizer, Buenos Aires, Argentina) was injected on day 14. A 20 mg/kg body weight dose of demeclocycline (Sigma, St. Louis, Mo., USA, D-6140) was injected on day 28.

Two days after the last injection, all the animals were euthanised. Both tetracyclines are fluorescent under ultraviolet light; oxitetracycline labeling fluoresces in yellow bands whereas demeclocycline fluoresces in orange/yellow bands. A Zeiss Axioplan microscope equipped with UV light, an excitation filter ranging from 450 to 490 nm, and a barrier filter at 515 nm was used. Table 1 shows the experimental protocol.

Determinations of Urea and Creatinine Concentration

On day 30, all the animals were anesthetized in order to collect blood samples by heart puncture. Serum urea and creatinine levels were assayed using Wiener lab. Kits (Rosario, Argentina) and evaluated quantitatively by spectrophotometry.

Determination of Parathyroid Hormone Levels

Serum PTH was assayed using an immunoradiometric assay (IRMA) for the quantitative determination of parathyroid hormone levels in rat serum (using Nichols Institute Diagnostics kit).

Histology

The animals were euthanised on day 30 and the tibiae were fixed in 15% formalin. The right tibiae were processed for embedding in methylmethacrylate. Undecalcified sections of the tibiae were obtained using a Leica polycut microtome (Wetzlar, Germany) to perform static and dynamic histomorphometry and Perls’ reaction (Prussian Blue) was used to evidence iron deposits.

Histomorphometry

Static histomorphometry involves the identification of cellular and tissue components for the measurement of lengths (mm), areas (mm²), or cell counts. Dynamic histomorphometry uses fluorochromes, such as tetracycline, that are incorporated into bone at the front of calcification. These labeled sites fluoresce under UV microscopy. When fluorochromes are administered over a time interval, the rates of formation and mineralization can be calculated from measurements of tissue growth between the labels. The static and dynamic histomorphometry in bone sections of animal models can give information about the bone response, help determine the effectiveness of pharmaceutical agents, and characterize toxicological effects.

Static histomorphometry

Histomorphometric determinations (Parfitt et al., 1987) were performed on undecalcified sections stained with Masson’s trichrome technique, employing semi-automatic, specific software (Osteoplan, Carl Zeiss, Oberkochen, Germany). The following static parameters were evaluated in 15 fields corresponding to the proximal tibia metaphysis (Gerharz et al., 2003) (Figure 1).

In order to ensure that the cancellous or trabecular bone measured was of the same kind, tibia length was measured in five tibiae from each group using a Vernier caliper (Mitutoyo, Japan) showing no differences among groups (data not shown). All measurements were performed by the same observer. The total bone-tissue area measured was 2.99 mm² in all the tibia sections.

• BV/TV (%): Bone volume: fraction of bone tissue in total volume. Total volume was taken as bone tissue plus bone marrow.

• OV/BV (%): Osteoid volume: fraction of osteoid tissue in the total bone volume.

• OS/BS (%): Osteoid surface: fraction of total bone surface presenting osteoid tissue.

• O.Th (μm): Osteoid thickness.

• Ob.S/OS (%) :Osteoblast surface/osteoid surface.

• N.Ob/B.Pm (×100 mm): Number of osteoblasts: number of osteoblasts per 100 mm of bone perimeter.

• Ob.S/BS (%): Osteoblast surface: fraction of total bone surface with active osteoblasts.

• ES(Oc+)/BS (%): Active erosive surface: Active erosive surface (with osteoclasts) as a fraction of total bone surface.

• ES(Oc−)/BS (%): Inactive erosive surface: Inactive erosive surface (without osteoclasts) as a fraction of total bone surface.

• ES/BS (%): Total erosive surface: Total erosive surface (inactive + active) as a fraction of total bone surface.

• Fb.S/BS: Surface presenting peritrabecular fibrosis.

Dynamic Histomorphometry

Undecalcified and unstained sections were used to perform dynamic histomorphometry employing specific software (Osteoplan, Carl Zeiss, Oberkochen, Germany). Dynamic parameters were measured in the same area (15 fields) indicated in Figure 1.

The following parameters were evaluated:

• DiL (μm): Distance between labels: distance between both labeling lines.

• DoL/BS (%): Double labeling: Double labeled surface as a fraction of total bone surface.

• SiL/BS (%): Single labeling: Single labeled surface as a fraction of total bone surface.

Statistical Analysis

The data were statistically evaluated using Duncan’s test for multiple comparisons.

Results

Body Weight

Body weight was recorded every 7 days throughout the experimental period. Initial and final weights of all the animals are plotted in Table 2. The final weight of all the animals in the experimental RF groups was significantly lower than that of sham.

Determinations of Urea and Creatinine Concentration

Serum urea was significantly (p < 0.05) higher in the RF and the RF+Io groups as compared to sham and Io groups. Creatinine levels were higher in the RF, RF+Io, and Io groups. This finding would appear to indicate that iron toxicity contributes to elevating creatinine levels in these groups (Table 3).

Determination of Parathyroid Hormone Levels

Although no significant differences were found among groups when comparing PTH levels (measured at the end of the experiment), an increase in PTH was observed in RF (38 ± 10 pg/ml) and RF+Io (43 ± 14 pg/ml) compared to sham (34 ± 0.7 pg/ml), and Io (34 ± 8 pg/ml) animals.

Histology and Histomorphometry

Iron intoxicated rats with and without renal failure presented diarrhea and changes in the color of oral mucosa. The dose used in our study definitely affected target organs, as are the liver and bone. Macroscopic observation of the livers showed them to be enlarged and microscopic evaluation evidenced the presence of a large number of macrophages with iron uptake, as evidenced by Perl’s reaction.

Histologic sections of the tibiae in the RF and RF+Io groups evidenced greater bone volume and growth cartilage plate sealed to trabeculae immediately below the metaphyseal cartilage than sham. RF sections evidenced more clearly defined osteoid surfaces mainly presenting active osteoblasts and peritrabecular fibrosis, whereas Io and RF+Io sections showed a predominance of inactive cells on bone surfaces, and iron deposits in bone marrow cells, osteoblasts, and osteocytes (Figures 2 and 3).

All the groups exhibited active and inactive erosion surfaces. Fluorescence microscopy failed to reveal double labeling in tibiae sections of any of the groups. However, tibiae in the RF group exhibited increased single tetracycline labeling.

Results of the histomorphometric study are shown in Table 4.

Discussion

Our results show that final body weight of animals in all three experimental groups (Io, RF, and RF+Io) was lower than that of sham animals. The difference was more marked in both groups with renal failure (RF and RF+Io). The dehydration typically associated with renal failure would explain the loss of weight in RF animals. Rats in the RF+Io group also suffered diarrhea, which would account for the more severe weight loss observed in this group.

Approximately 20% of animals were lost during the experiment. In all cases, death occurred postnephrectomy. The results obtained with the static histomorphometric study (number of osteoblasts, bone surface with active osteoblasts, osteoid volume and surface) revealed no significant differences between the RF+Io and the absolute sham group. This could lead to the conclusion that RF+Io does not affect bone formation. However, the RF group exhibited a larger amount of osteoblasts, higher percentage of bone surfaces with active osteoblasts, greater osteoid volume and surface, and peritrabecular fibrosis, as compared to the sham group. Furthermore, all the static parameters of bone formation were found to be lower in Io animal than in sham group. This finding is in agreement with previous experimental studies on iron intoxication in pigs (de Vernejoul et al., 1984) and Wistar rats (Mandalunis et al., 1997).

It is noteworthy that in the present study, the RF+Io group did not show the characteristic features of renal osteodystrophy. It is possible that the histological features observed in the RF+Io group are a result of the combination of RF and Io. These two factors may have a combined effect, or may act separately.

It is well documented that parathyroid hormone, whether in hyperthyroidism induced by renal disease or in primary hyperthyroidism, stimulates osteoclastic activity resulting in bone calcium removal and causing osteopenia in long-term patients. Osteoblasts are known to have receptors for PTH, so that the effect of PTH on osteoclasts is indirect and mediated by osteoblasts (Habener and Potts, 1990). Given that RF animals exhibited increased osteoblastic activity, the experimental times used in our study and the results obtained herein allow us to infer that the condition induced in this experiment corresponds to the first stages of renal disease.

Iron may affect osteoblasts and lining cells, inhibiting or decreasing collagen matrix synthesis, or it may alter osteoblast PTH receptors, inhibiting the effect of this hormone. Although it was not the aim of this study to evaluate changes in bone marrow, animals treated with iron (Io and RF+Io) exhibited an increase in the number of cells presenting iron in their cytoplasm. The osteoblastic inactivity and reduced bone formation rate observed in our study may be associated to a decrease in IGF-1 resulting from changes caused to hematopoietic cells by the presence of iron deposits, as described by Mahachokletwattan et al. (2003). However, in contrast with our results, the aforementioned authors also reported the presence of osteoclastic activity and osteomalacia. It is possible that the different pathologies observed in bone and bone marrow cells in beta-thalassemia and renal failure are associated to different mechanisms of action of iron.

Although our results showed increased PTH levels in RF+Io animals, the typical histologic features of bone exposed to increased PTH were not observed. Therefore, iron may alter the effect of PTH, possibly altering osteoblast PTH receptors or affecting its biological pathway.

In the present study, iron was detected in bone cells and bone marrow cells, but was not observed on mineralized bone surfaces. This finding is not in agreement with studies by de Vernejoul et al. (1984) and Huser et al. (1988) reporting iron deposits on the osteoid-mineralized bone interface. In our study, iron deposits identified by Perl’s stain were observed in bone marrow cells, osteoblasts and osteoclasts, but not on the mineralized bone surface. Ebina et al. (1991) found iron deposits in osteoblasts and osteoclasts of normal Wistar rats injected with aluminum and iron complexes. Our results are also in agreement with in vitro studies by Diamond et al. (1991) demonstrating that iron causes a diminution in cell proliferation and in osteoblast activity. Jacob et al. (1997), reported accelerated apoptosis in endothelial cells in the presence of iron. These data lend support to the results obtained herein which show that osteoblast surfaces decreased in the Io group compared to sham group, and in the RF+Io group compared to the RF group.

Matsushima et al. (2003) reported that iron lactate induced osteomalacia in rats. In the present study using dextran iron, osteomalacia was not observed in Io or RF+Io animals. Though it holds true that the static parameters were similar in sham and RF+Io animals, revealing no significant differences between these two groups, results obtained with the dynamic bone study show that bone response of these groups was different throughout the experimental period. Because RF+Io animals exhibited less single labeling in the subchondral bone than sham and IR rats, it could be thought that RF+Io does not affect bone formation. However, analysis of bone dynamics does not lend support to this conclusion. Comparison between static parameters of the RF+Io and sham group showed no significant differences, whereas comparison between dynamic results of these same groups did show significant differences. The RF+Io group exhibited fewer tetracycline labeled surfaces, indicating less bone mineralization.

Both static and dynamic parameters of bone formation were found to be diminished in the RF+Io group, and were significantly different compared to the RF group. Comparison among results obtained with the different groups, showed that both the Io and the RF+Io groups presented adynamic bone disease with normal osteoid volume, and that static bone formation parameters for the RF+Io group were similar to the sham group, probably due to the effect of renal failure.

Animals in the RF and RF+Io groups presented significantly greater subchondral bone volume with the presence of growth cartilage plate sealed to trabeculae compared to the Io and sham groups. This finding may be associated with bone resorption, since inactive and total resorption was found to be lower in these groups. Cobo et al. (1999) reported an increase in growth cartilage height and in the number of chondrocytes in the hypertrophic zone in nephrectomized rats. These findings show that renal failure inhibits endochondral ossification. Our observation of growth cartilage plate sealed to trabeculae, which is associated with impairment in endochondral ossification and/or bone remodeling, is in keeping with Cobo’s report.

Whereas substantial peritrabecular fibrosis was found in the IR group, it was only observed on few surfaces in the RF+Io group. Peritrabecular fibrosis is a histopathological entity that manifests not only in renal osteodystrophy but also in other bone pathologies. Peritrabecular fibrosis is commonly found in bone disorders presenting alterations in bone turnover, and is associated with erosion and formation surfaces. As mentioned above, the RF group exhibited significant peritrabecular fibrosis, whereas it was only found on few surfaces in the RF+Io group. Differences between the IR+Io group and the sham and Io groups were not significant. The imbalance in bone metabolism caused by renal failure and treatment for renal failure is yet to be clarified, given the complexity of the condition and the many factors involved.

Aluminum intoxication is often the cause of anemia in renal failure patients. Because of its affinity for transferrin, aluminum competes with iron thus causing anemia (Van Landeghem et al., 1997). Anemia is usually treated with erythropoietin or iron compounds. The results obtained in this work demonstrate that iron overload produces adynamic bone disease (ABD) that features reduced or no cell activity (Malluche and Faugere, 1990) and is defined as low bone turnover and decreased bone formation (Velazquez-Forero et al., 1996). Adynamic bone disease has only been recognized as a pathologic entity in the last few years, and is not fully understood. The bones may have normal strength and overall appearance, yet they are under active. Although ABD has been associated to low levels of PTH, the results obtained in this study did not show this association. This would seem to indicate that iron overload causes ABD, regardless of PTH levels.

It is known that acidosis is a possible contributing factor to the bone loss seen in renal disease. Further studies should be conducted to clarify the role of acidosis in this experimental model. Our results studying the combined effects of RF+Io suggest that iron overload causes changes in the typical histopathological features of renal osteodystrophy, suggesting the need for careful evaluation and treatment of renal failure patients receiving iron therapy.