JOURNAL OF BONE AND MINERAL RESEARCH Volume 4, Number 5, 1989 Mary Ann Liebert, Inc., Publishers

The Evolution of Osteomalacia in the Rat with Acute Aluminum Toxicity

MARIAN0 RODRIGUEZ, ARNOLD J. FELSENFELD, and FRANCISCO LLACH

ABSTRACT

Aluminum toxicity is the presumed cause of aluminum-associated osteomalacia. In animal models, osteoma- lacia has been produced after a prolonged course of aluminum. In the present study, rats with renal failure received 20 mg intraperitoneal aluminum during a 2 day period. This model allows sequential observations in the development of osteomalacia. Rats were sacrificed and studied 5, 12, 25, and 40 days after aluminum administration. No differences were observed in serum calcium, phosphorus, or creatinine as a consequence of aluminum administration. Compared with control rats, parathyroid hormone was decreased at 12 and 25 days. A direct correlation was present between plasma and bone aluminum at 12 days ( r = 0.92, p < 0.01), 25 days ( r = 0.85, p < 0.005), and 40 days ( r = 0.88, p < 0.001) but not 5 days after aluminum administra- tion. Plasma aluminum peaked at 5 days (727 f 89 &liter, mean f SEM) and bone aluminum at 40 days (273 f 40 pg/g) . Aluminum had profound effect on bone histology. At 5 days there was a decrease in osteo- blast surface and osteoid surface; at 12 days osteoblast surface and osteoid surface returned to normal but osteoclast surface decreased. Subsequently there was a progressive increase in osteoid surface and osteoid volume. Bone formation rate measured at 12, 25, and 40 days was decreased at these intervals.

In conclusion, (1) high plasma aluminum may be directly toxic to the osteoblast; (2) progressive osteoid accumulation is secondary to matrix (osteoid) deposition, which exceeds the depressed bone formation rate; (3) the progressive decrease in plasma aluminum and increase in bone aluminum suggest that bone has a high affinity for aluminum but may have a relatively slow rate of uptake at any given time; (4) aluminum may di- rectly decrease parathyroid hormone; (5) the correlation between plasma and bone aluminum suggest an ex- change is present; and (6) aluminum toxicity may independently affect the osteoblast and bone mineraliza- tion.

INTRODUCTION

LUMINUM TOXICITY has been implicated as the cause of A aluminum-associated o s t e ~ m a l a c i a . ~ ~ - ~ ~ Evidence to support this contention includes the association with high dialysate aluminum content,(',*' the finding of aluminum deposits at the mineralization front and along the trabecu- lar bone surface," 4 , elevated bone aluminum ont tent,'^) and the production of osteomalacia in animals by alumi- num administration. ( 6 - 8 )

The mechanism by which aluminum produces osteoma- lacia is not well defined. It may have a direct effect on

bone, and this has been suggested by the finding that aluminum may be toxic to the o s t e o b l a ~ t , ~ ~ ~ ' ~ ~ may inhibit bone phosphatase activity,'"' and may directly affect cal- cification at the mineralization front.'12,131 Another factor that may modulate the effect of aluminum on bone i s parathyroid hormone (PTH). The characteristic patient with aluminum-associated osteomalacia has a relative PTH deficiency. l4. l4)

Studies in rats have demonstrated that the chronic ad- ministration of aluminum produces osteomalacia. The production of osteomalacia or inhibition of mineralization has required more than 20 days of exposure to alumi-

Department of Medicine, University of Oklahoma Health Sciences Center, and VA Medical Center, Oklahoma City, OK.

687

688 RODRIGUEZ ET AL.

n ~ m . ( ~ - ~ , ' ~ ) Data are not available in regard to a briefer but more intense exposure to aluminum. The sequential changes in the development of aluminum-induced osteo- malacia are not known. The overt disease as observed in dialysis patients or chronic rat studies represents the end result of pathophysiologic changes that can be better de- fined through the study of the initial stages of the disease.

The present study provides evidence that 2 days of high- dose aluminum produces osteomalacia. The sequential evolution of the development of osteomalacia reveals that osteoblast numbers recover before mineralization. In addi- tion, a progressive decrease in plasma aluminum and an in- crease in bone aluminum were observed.

METHODS

A total eight groups of male Wistar rats weighing from 260 to 310 g were studies. Renal failure was surgically in- duced by arterial ligation of two of the three main arteries to the left kidney, followed 1 week later by total right nephrectomy. Elemental aluminum (20 mg) dissolved in D,W was intraperitoneally injected in divided doses during a 2 day period, 7 days after the right nephrectomy. The aluminum was in the form of aqueous aluminum chloride. Rats were sacrificed 5, 12, 25, and 40 days after aluminum administration. Pair-fed groups of rats that had renal fail- ure and had received diluent but no aluminum were also sacrificed at 5, 12, 25, and 40 days. In all groups except the 5 day group, bone formation was assessed by giving 20 mg intraperitoneal tetracycline for 2 consecutive days and re- peating after 6 days. In addition to blood obtained at sac- rifice, blood samples for biochemical parameters were drawn from the tail vein pre- and postnephrectomy and at 3, 5, 12, and 25 days after either diluent or aluminum ad- ministration. All rats had free access to water and were fed standard rat chow (Rodent Laboratory Chow, Ralston Purina Company, St. Louis, MO) containing 1.2% cal- cium, 0.86% phosphorus, and 5.3 IU/g of vitamin D.

At the time of sacrifice, blood was obtained by cardiac puncture for determination of calcium, phosphate, creati- nine, aluminum, and PTH. Calcium was measured with an automated calcium analyzer (Calcette, Model 4008, Preci- sion Systems, Inc., Sudbury, MA). Phosphorus was deter- mined with a specific kit for phosphate (Fast Phosphorus Test Set, Stanbio, San Antonio, TX). Creatinine was mea- sured with a specific analyzer for creatinine (Creatinine Analyzer 2, Beckman Instruments, Fullerton, CA). Alumi- num was measured by flameless atomic absorption with a graphite furnace."6) Parathyroid hormone was measured with a N-terminal radioimmunoassay (Nichols, San Juan Capistrano, CA). The assay uses an antibody that in pub- lished results has been shown to measure rat PTH.(17) In the present study, blood was obtained from six rats before and after the induction of renal failure and 5, 12, 25, and 40 days after 20 mg intraperitoneal aluminum was given during a 2 day period. In addition, blood was obtained at the same time intervals in five control rats in whom only diluent was administered.

After sacrifice, the femur was obtained for determina- tion of bone aluminum content. The bone was cleaned of adhering tissue by scraping with a glass knife and air dried. The method for determination of aluminum was adapted from LeGendre and Alfrey.IL6) For analysis, samples of finely crushed bone were placed in a saturated solution of Na, EDTA and extracted by mixing for 24 h. Samples (20 pl) of EDTA extract were analyzed in a Perkin-Elmer graphite furnace utilizing a Perkin-Elmer HGA 2100 con- troller, AS- 1 autosampling device, deuterium background corrector, and ramp accessory and a Perkin-Elmer 2380 atomic absorption spectrophotometer. Quintuplet analyses were performed on duplicate samples.

After exsanguination, the ilium was detached from the skeleton and placed in 70% ethanol. After dehydration and defatting, the bone was embedded in methyl meth- acrylate. Sectioning of the block was performed with a sledge microtome (Jung, Model K, Heidelberg, West Ger- many). For histomorphometric analysis, 5 pm Goldner- stained sections were examined. Sections (5 pm) were stained for aluminum by the method of Maloney et al.(Is) Unstained sections (15 pm) were utilized for analysis of the tetracycline labels. All analyses were begun below the carti- lage plate. Goldner-stained sections of cancellous bone were analyzed at a magnification x500 with a Merz- Schenk reticle.(71 All histomorphometric assessments were accomplished with the aid of a Leitz Ortholux I1 micro- scope (Leitz, Wetzlar, West Germany).

Measurements of trabecular bone were made in the fol- lowing categories: (1 ) osteoblast surface, the fraction of trabecular surface covered by osteoblasts; (2) osteoid sur- face, the fraction of trabecular surface covered by osteoid; (3) osteoclast surface, the fraction of trabecular surface covered by osteoclasts; (4) osteoid volume, the fraction of trabecular bone volume occupied by osteoid; (5) osteoclast number, the number of osteoclasts per square millimeter of cancellous bone; (6) stainable bone aluminum, the frac- tion of trabecular bone surface covered with aluminum; (7) mineral apposition rate, the mean distance between tetra- cycline labels divided by number of days between doses of tetracycline; this value is uncorrected for the obliquity of the plane of section; and (8) bone formation rate, the vol- ume of mineralized bone formed within a unit volume of bone per day. The bone formation rate at the tissue level was obtained by the following formula: Bone formation rate equals fraction double-labeled trabecular surface times mineral apposition rate. The bone formation rate at the bone modeling unit (BMU), which gives the average amount of mineralized bone made per day per unit osteoid covered surface, was obtained by the following formula: bone formation rate equals fraction double-labeled trabec- ular surface times mineral apposition rate divided by the fraction osteoid surface.

Biochemical data and histomorphometric measurements were analyzed using one-way analysis of variance. The Bonferroni test was applied when multiple comparisons were performed. Statistical differences were further inves- tigated by utilization of the unpaired t-test. A P value < 0.05 was considered statistically significant. All values shown in the text represent the mean f standard error of the mean (SEM).

ACUTE ALUMINUM TOXICITY

2

6' 5 p.

689

690

20- . 0 a -

15- f a P 8 10- a

!z = 5 - d 3

t I

a a O L W

RESULTS

d+ 1 I I I

RODRIGUEZ ET AL.

In Table 1, blood chemistries are shown for the control and aluminum groups prior to induction of renal failure (pre-Nx), after renal failure (post-Nx), and 3, 5 , 12, 25, and 40 days after the administration of aluminum or dilu- ent. No significant differences are observed in serum cal- cium, phosphorus, or creatinine in the control compared with the aluminum group. However, a significant increase in serum calcium occurs after the induction of renal failure (post-Nx) in both groups. Serum creatinine increases sig- nificantly post-Nx in both groups.

As shown in Fig. 1, the serum PTH concentration post- nephrectomy increases from 6.8 f 0.36 to 10.9 f 0.64 pg/ ml, p < 0.01. At 5 days the serum PTH in the control group is not different from that in the aluminum group. At 12 and 25 days, the serum PTH concentration is greater in the control than in the aluminum group. In addition, in the aluminum group the decrease from baseline (post-Nx) is significant at 12 0, < 0.02) and 25 (p < 0.04) days. By 40 days the difference between the two groups approaches but does not attain statistical significance (p < 0.08).

Histologic data are presented in Table 2 and Fig. 2. At 5

TABLE 2. HISTOMORPHOMETRIC DATA FROM CONTROL (C) AND ALUMINUM (Al) GROUPS~

I2 Day 25 Day 40 Day 5 Day

AI C A1 C A1 C AI C (N = 9) (N = 7) (N = 8) (N = 8) (N = 10) (N = 7) (N = 8) (N 7)

Osteoblast surface ('70)

Osteoid surface ('70)

Osteoclast surface (Yo)

Osteoclast number

Osteoid volume (Yo)

Bone formation rate (tissue, pm3/pm2 per day)

Bone formation rate (BMU, Fm/daY)

Mineral apposition rate (pm/day)

Stainable aluminum

(%)

2.1 f 0.6*

7.1 f 1.3*

11.7 f 1.7

5.7 f 0.9

2.1 f 0.5

23.2 f 3.2'

22.9 f 2.4

21.2 f 2.4

9.6 f 1.0

5.5 f 0.5

3.7 f 0.7

0

10.0 14.9 f 3.5t f 2.3

25.6 22.5 f 7.0 f 3.0

2.1 10.3 f 0.5* f 1.0

1.2 4.0 f 0.2*,t f 0.3

7.5 3.5 f 1.9* f 0.7

O* 0.3 1 f 0.05

O* 1.48 f 0.27

O* 1.72 f 0.10

35.6 0 f 6.9*

12.7 f 2.7

46.6 f 6.8*

7.9 f 1.6t

2.7 f 0.5

15.7 f 3.8*

0.12 f 0.06*

0.39 f 0.19*

1.04 f 0.36*

29.4 f 6.8*

15.6 5 .O f 2.6 f 1.0

23.8 63.2 f 3.9 f 4.6*

5.6 2.7 f 0.6t f 0.2*

3.4 1.2 f 0.6 f 0.2

3.2 32.1 f 0.8 f 4.6*

0.49 0.10 f 0.09 f 0.07*

2.44 0.18 f 0.46 f 0.13*

2.05 0.78 f 0.18 f 0.37*

0 45.6 + 3.6*,t

9.1 f 2.4

13.7 f 3.3

5.6 f 1.4

1.9 f 0.4

1.9 f 0.4

0.44 0.12

3.95 f 1.63

1.92 f 0.12

0

=Mean f SEM. (*) p < 0.05 A1 versus C. (t) p < 0.05 versus same group, preceding time interval.

ACUTE ALUMINUM TOXICITY 691

6 0 -

4 0 -

20 -

days after aluminum administration, osteoblast surface and osteoid surface are significantly decreased (p < 0.001) compared with the control group. Osteoclastic activity as represented by osteoclast surface and osteoclast number is similar in the two groups. No differences are noted in 0s- teoid volume. Bone formation rate is not measured be-

cause the 5 day interval between aluminum administration and sacrifice does not provide adequate time for double tetracycline labeling.

At 12 days, osteoblast surface returns to control levels and osteoid surface is not significantly different. Both osteoclast surface and osteoclast number are significantly decreased in the aluminum group (p < 0.001). Osteoid volume is increased in the aluminum group (p < 0.05). The mineral apposition rate and the bone formation rate are completely absent in the aluminum group, and distinct separation of tetracycline labels (mineral apposition) and T

w 0 U

3 cn

a a

s w I-

$

L

1 '9' 0

Y 3 /

/

d >

0 5 12 2 5 4 0 DAYS

FIG. 2. The effect of 20 mg of intraperitoneal aluminum administered during a 2 day period on osteoblast surface, osteoid surface, and osteoid volume (Yo). The days on the abscissa represent the time after aluminum administration. Mean f SEM (G - - 0) aluminum; (e--o) control; (*) p < 0.05 versus control.

active bone formation are present in the control group (Fig. 3).

At 25 days, significant differences are present between the aluminurn and control groups in osteoid surface (p < 0.03), osteoid volume (p < 0.03), mineral apposition rate (p < 0.05), and bone formation rate; the last at both the tissue (p < 0.005) and BMU level 0, < 0.001). Although less than the control group, bone formation is present in some rats in the aluminum group.

At 40 days osteoid accumulation as represented by oste- oid surface (p < 0.001) and osteoid volume (p < 0.005) continues to increase in the aluminum group compared with the control group. The changes in osteoblast surface, osteoid surface, and osteoid volume through the course of the study are shown in Fig. 2. Osteoblast surface is de- creased at 5 days. A progressive and significant increase in osteoid surface and osteoid volume occurs between 5 and 40 days. At 40 days, osteoclast surface is lower in the alu- minum than in the control group (p < 0.001). Both min- eral apposition rate (p < 0.03) and bone formation rate at the tissue and BMU levels (p < 0.005) are greater in the control group. The bone formation rate at the tissue level in both groups is shown in Fig. 3. A representative exam- ple of the histologic findings at 40 days in the aluminum group is shown in Fig. 4.

i I 1 1 1 0 12 25 4 0

DAYS

FIG. 3. The effect of aluminum on the bone formation rate at the tissue level 12, 25, and 40 days after aluminum administration. Mean values f SEM are displayed. (0) aluminum, ( 0 ) control. (*) p < 0.05 versus control.

692 RODRIGUEZ ET AL.

FIG. 4. minum administration. ( x 200).

Large deposits of acellular osteoid (0) cover the majority of the trabecular bone (b) of the rat 40 days after alu-

As shown in Table 2, stainable bone aluminum is greater in the aluminum group at every interval. There is a pro- gressive increase in stainable bone aluminum from 5 to 40 days. As presented in Fig. 5 , a correlation is observed be- tween bone aluminum, as measured by flameless atomic absorption, and stainable bone aluminum (p < 0.001).

As presented in Table 3, the maximum level of plasma aluminum 727 * 89 pg/liter, is observed at 5 days. Subse- quent values at 12, 25, and 40 days decrease. As opposed to plasma aluminum, the maximum bone aluminum value is observed at 40 days. As displayed in Fig. 6 , a significant direct correlation exists between bone and plasma alumi- num at 12, 25, and 40 days but not 5 days.

The inverse correlation present between osteoblast sur- face and plasma aluminum is significant when data from the 5 , 12, 25, and 40 day aluminum groups are pooled ( r = -0.50; p < 0.01). The plasma aluminum concentration at 5 days is markedly increased (727 f 89 pg/liter) and de- creases progressively to 40 days. If results obtained at 40 days, when plasma aluminum levels are relatively low, are excluded, then the inverse correlation between osteoblast surface and plasma aluminum increases in significance (r = -0.72; p < 0.001). These data are shown in Fig. 7.

400 joO1

. 'i, 2 300- 3 4 ;r' 200-

z

100-

m 20 30 40 50 60 T O

STAINABLE TRABECLJLAR BONE A L W N U I (%I

FIG. 5. The correlation between stainable bone alumi- num and bone aluminum as measured by flameless atomic absorption. r = 0.81; p = <0.001.

ACUTE ALUMINUM TOXICITY 693

TABLE 3. BONE AND PLASMA ALUMINUM IN CONTROL (C) AND ALUMINUM (A]) GROUPS"

5 Day 12 Day 25 Day 40 Day

A / C A / C A / C A / c (N = 7) (N = 7) (N = 5) (N = 6) f N = 9) fN = 7) (N = 8) (N = 7)

Plasma aluminum 727 4.0 345 3.3 334 2.1 198 1.1 (FgAiter) f 89* f 4.4 f IIJI - , ~ f 3.2 f 63* f 0.45 f 31* f 0.95

Bone aluminum 79 1.41 182 1.27 I06 I .30 273 0.98 (LCg/g) f 11* f 0.50 f 44*9t f 0.18 f 19* f 0.37 f 40*7t zt 0.67

aMean f SEM. (*) p < 0.05 Al versus C. (t) p < 0.05 versus same group, preceding time interval.

DISCUSSION

The present study demonstrates that a short but intense exposure to aluminurn results in profound osteomalacia. This study, by providing sequential histologic data, sug- gests that aluminum may be directly toxic to the osteoblast and separately inhibit mineralization. In addition, alumi- num administration decreases serum PTH levels.

Pair-fed groups are compared pre- and postinduction of renal failure and 3, 5 , 12, 25, and 40 days after administra- tion of aluminum or diluent. Regarding the biochemical results, no differences are observed between the two groups at the same time interval with respect to serum cal- cium, phosphorus, and creatinine. As expected, serum cre- atinine increases after the induction of renal failure. Serum calcium also increases after the induction of renal failure. The explanation for this observation is not clear but has been reported previously."') There is not a significant in- crease in serum phosphorus after the induction of renal failure. This is not entirely unexpected since considerable data exist to suggest phosphorus excretion is enhanced in moderate renal failure. ( 2 a . 2 1 )

A relative PTH deficiency is present in the clinical syn- drome of aluminum-associated osteomalacia. ( I 4 In addi- tion, in vitro exposure to aluminum by parathyroid cells decreases PTH 2 4 ) In a recent report, daily aluminum administration for 40 days decreased serum PTH levels in In the present study in aluminum- loaded rats, PTH levels are unchanged at 5 days but signif- icantly decreased from baseline (post-Nx) at 12 and 25 days. Similarly, PTH levels are decreased at 12 and 25 days in aluminum-loaded compared with control rats. These findings support the conclusion that aluminum produced the decrease in PTH. Since the PTH levels are not sup- pressed at 5 days and plasma aluminum levels are highest at 5 days, circulating levels of aluminum may not influence PTH secretion. It is possible that aluminum deposition in the parathyroid glands decreased PTH secretion. The asso- ciation of decreased PTH secretion and increased parathy- roid gland aluminum content has been reported previously in the rat.'z5)

Definite changes in bone histology are observed in the aluminum group at all four time intervals studied. At 5 days there is a marked decrease in osteoblast surface. This decrease is likely a direct toxicity from high plasma alunii- num levels. Parathyroid hormone does not appear to be a factor since levels d o not change. In the clinical syndrome of aluminum-associated osteomalacia, osteoblast surface is decrea~ed . '~ ") Plachot et al. have detected by electron mi- croprobe aluminum deposits in the mitochondria of the os- t e ~ b l a s t . ' ~ ) Also, in vitro studies have shown aluminum to be directly toxic to the osteoblast.llO) Thus, aluminum in high doses may be directly toxic to the osteoblast. At 5 days there is also a significant decrease in osteoid surface while osteoid volume decreases. This finding would suggest that the generation of new osteoid seams decreases. The amount of osteoid present depends on the rate of osteoid synthesis by osteoblasts (matrix deposition) and the rate of removal of osteoid by mineralization of the osteoid matrix (bone formation rate). I f both matrix deposition and bone formation decrease to the same degree, then no change in total osteoid seams or osteoid volume should be observed. Since this is not found, then the rate of bone formation must exceed matrix deposition. Since osteoblast surface is markedly reduced, this would suggest that matrix deposi- tion is decreased. The rate of bone formation is sufficient to eliminate many of the thin osteoid surface seams and hence the decrease in osteoid surface.

At 12 days osteoblast surface has returned to control levels, and osteoid volume is increased while bone forma- tion is absent. These findings indicate that matrix (osteoid) deposition is greater than the rate of formation. The find- ing that aluminum toxicity may independently affect the osteoblast and mineralization has been observed previously during the implantation of bone matrix.113) At the same time, osteoclast surface and osteoclast number have sig- nificantly decreased compared with the control group and the 5 day aluminum group. Since PTH levels decrease, i t is possible that decreased PTH levels contribute to the de- cline in osteoclasts. If aluminum is directly toxic to the os- teoclast, a decreased number would be expected at 5 days. Another possibility is that the presence of osteoblasts is

694 RODRIGUEZ ET AL.

3 0 0

2 0 0

I00

0 c\

E cn

Fn \

a ” f 100 z I 3

r = .92 p<.o I . -

r = .85 p< . 005

/

z m “ I

c cn < d W c 8 10

W /

r - -0 .12 p < 0 . 0 0 1

-a----- 1

A 1 - \ O . . = \ . I -

I A 0 0 I 1 1 I 1

200 400 800 800 1000

PLASMA ALUMINUM (POlL)

FIG. 7. The inverse correlation between osteoblast sur- face and plasma aluminum. Data from 5 , 12, and 25 days after aluminum administration are plotted. The 40 day group is excluded from the analysis. 2ooY I 0 0

r = .88 p< .oo I

O< 200 400 600 800

PLASMA ALUMINUM ( p g / L )

FIG. 6 . The correlation between plasma aluminum and bone aluminum at 12, 25, and 40 days after aluminum ad- ministration. A significant correlation is present at each time interval.

important for osteoclastic r e ~ o r p t i o n . ~ ~ ~ ~ ~ ’ ~ Thus, the de- creased osteoblast surface at 5 days may induce the decline in osteoclasts at 12 days.

At 25 and 40 days there is progressive osteoid accumula- tion. Both osteoid surface and osteoid volume increase while bone formation is subnormal. These findings indi- cate that matrix deposition is more active than the mineral- ization phase. The early phase of aluminum toxicity (5 day) results in decreased matrix deposition, but by 12 days this has become more active than bone formation, which, first measured at 12 days, is decreased for the duration of

the study. Thus, it is likely that aluminum may directly af- fect bone metaboism in at least two important areas. First, aluminum affects matrix deposition presumably by de- creasing osteoblast surface. Since only 5 days after alumi- num exposure the number of osteoblasts is decreased, one must assume that both shortened survival due to high plasma levels of aluminum and decreased generation of os- teoblasts must play a role. Second, after osteoblast surface has returned to normal, absent or decreased bone forma- tion persists. Although it appears that aluminum directly impairs mineralization, the possibility that the osteoblast fails to produce the proper milieu for mineralization, that is, matrix vesicles,(z8) cannot be excluded.

During the 40 days after aluminum administration, a progressive decrease in plasma aluminum and increase in bone aluminum is observed. At the same time, as shown in Fig. 6 , a correlation is observed between plasma and bone aluminum at 12, 25, and 40 days despite a 1000-fold greater concentration in bone. This suggests the presence of an exchange of aluminum between plasma and bone. Although the distribution of aluminum between soft tis- sue, bone, and plasma and the rate of urinary excretion is very complex, the findings do indicate bone preferentially accumulates aluminum, but the rate of aluminum uptake by bone is relatively slow under these experimental condi- tions.

The percentage of stainable bone aluminum increases during the 40 day study, and as shown in Fig. 5 correlates

ACUTE ALUMINUM TOXICITY 695

with bone aluminum as measured by flameless atomic ab- sorption. This finding has been shown previously.~'')

In summary, a 2 day exposure to a large dose of alumi- num results in osteomalacia and decreased serum PTH levels in the rat with renal failure. Decreased osteoblast surface and matrix deposition are observed 5 days after aluminum administration. However, subsequently osteo- blast surface returns to normal, but matrix deposition con- sistently exceeds bone formation for the duration of the study. As a result, progressive osteoid accumulation is ob- served. These results suggest that aluminum initially re- duces the number o f osteoblasts inhibiting matrix deposi- tion and subsequently inhibits the mineralization process, producing a virtual cessation of bone formation. The find- ing of sequential changes separating matrix deposition and the mineralization process suggest that aluminum may af- fect two independent processes. An inverse correlation is observed between osteoblast surface and plasma alumi- num, suggesting that aluminum may be toxic to the osteo- blast. The progressive decrease in plasma aluminum and increase in bone aluminum suggest that bone preferentially accumulates aluminum but the rate of uptake may be rela- tively slow. The correlation between plasma and bone alu- minum at 12, 25, and 40 days suggests an exchange is pres- ent between plasma and bone aluminum. Finally, alumi- num loading decrease PTH levels from baseline and as compared with controls at 12 and 25 days after aluminum administrat ion.

ACKNOWLEDGMENTS

The authors wish to thank Mrs. Charlotte Jones-Vaultz and Mrs. Carolyn Clay for secretarial assistance and Mr. Carl Haygood and Mr. Sadegh Nematzadeh for technical assistance. This work was in part supported by a Merit Re- view grant from the Veterans Administration.

REFERENCES

Ward MK, Fees1 TG, Ellis HA, Parkinson IS, Kerr DNS 1978 Osteomalacic dialysis osteodystrophy: Evidence for a waterborne aetiological agent, probably aluminum. Lancet 1:

Parkinson IS, Ward MK, Feest TG, Fawcett RWP, Kerr DNS 1979 Fracturing dialysis osteodystrophy and dialysis en- cephalopathy. Lancet 1:406-409. Ott SM, Malone) NA, Coburn JW, Alfrey AC, Sherrard DJ 1982 The prevalence of bone aluminum deposition in renal osteodystrophy and its relation to the response of calcitriol therapy N Eng J Med 307:709-713. Cournot-Witmer G, Zingraff J , Plachot JJ , Escaig F, LeFerre R, Boumati P , Bourdreau A, Garabedian M, Galle P, Bourdon R, Drueke T , Balsan S 1981 Aluminum localiza- tion in bone from hernodialyzed patients: Relationship to matrix mineralization. Kidney Int 20:375-385.

841 -845.

5 . Hodsman AB, Sherrard DJ, Alfrey AC, Ott S, Brickman AS, Miller NL, Maloney NA, Coburn JW 1982 Bone alumi- num and histomorphometric features of renal osteody5- trophy. J . Clin Endocrinol Metab 54:539-546.

6. Ellis HA, McCarthy J H , Herrington J 1979 Bone aluminum in haemodialysed patients and in rats injected with aluminum chloride: Relationship to impaired mineralization. J Clin Pa- tho1 32832-839.

7. Robertson JA, Felsenfeld AJ, Haygood CC, Wilson P, Clarke C, Llach F 1983 An animal model of aluminum-in- duced osteomalacia: Role of chronic renal failure. Kidney Int 23:327-335.

8. Chan YL, Alfrey AC, Posen S, Lissner D, Hills E, Dunstan CR, Evans RA 1983 Effect of aluminum on normal and uremic rats: Tissue distribution, vitamin D metabolites, and quantitative bone histology. Calcif Tissue Int 35:344-351.

9. Plachot J J , Cournot-Witmer G, Halpern S, Mendes V, Bour- deau A, Drueke T, Galle P , Balsan S 1984 Bone ultrastruc- ture and x-ray microanalysis of aluminum intoxicated hemo- dialyzed patients. Kidney Int 25:796-803.

10. Lieberherr M, Grosse B, Cournot-Witmer G, Herman-Erlee MPM, Balsan S 1987 Aluminum action on mouse bone cell metabolism and response to PTH and 1 ,25(OH),D1. Kidney Int 31:736-743.

1 1 . Lieberherr M, Grosse B, Cournot-Witmer G, Thil CL, Bal- san S I982 In vitro effects of aluminum on bone phospha- tases: A possible interaction with PTH and vitamin D 3 rne- tabolites. Calcif Tissue Int 34:280-284.

12. Thomas WC Jr 1982 Trace metal-citric acid complexes a5 iri-

hibitors of calcification and crystal formation. Proc Soc Exp Biol Med 170:321-327.

13. Talwar HS, Reddi AH, Menczel J, Thomas W C Jr, Meyer JL 1986 Influence of aluminum on mineralization during nia- trix-induced bone development. Kidney Int 29:1038-1042.

14. Hodsman AB, Sherrard DJ, Wong EGC, Brickman AD, Lce DBN, Alfrey AC, Singer FR, Normal AW, Coburn JW 1981 Vitamin D-resistant osteomalacia in hemodialysis patients lacking secondary hyperparathyroidism. Ann Intern Med 94629-637.

15. Goodman WG, Gilligan J , Horst R 1983 Short-term admini,. tration in the rat: Effects on bone formation and relationship to renal osteomalacia. J Clin Invest 73:171-181.

16. LeGendre GR, Alfrey AC 1976 Measuring picoyrarn amounts of aluminum in biological tissue by flameles atomic absorption analysis of a chelate. Clin Chem 2253-56.

17. Toverud SU, Boass A, Garner SC, Endres DB 1986 Circu- lating parathyroid hormone concentrations in normal and vitamin D deprived rat pups determined with an N-terminal- specific radioinmunoassay. Bone Min 1:145-155.

18. Maloney NA, Ott SM, Alfrey AC, Miller NL, Coburn JU', Sherrard DJ 1982 Histological quantitation of aluminum in iliac bone from patients with renal failure. J Lab Clin Med 99:206-216.

19. Jastek TJ, Morrison AB, Raisz LG 1968 Effects of renal ii i-

sufficiency in the parathyroid gland and calcium homeo- stasis. Am J Physiol 215:84-89.

20. Caverzasio J , Gloor H I , Fleisch H, Bonjour J P 1982 Para- thyroid hormone-independent adaptation of the renal hand- ling of phosphate in response to renal mass reduction. Kid- ney Int 21:471-476.

21. Kraus E, Briefel G , Cheng L, Sacktor B, Spector D 1985 Phosphate excretion in uremic rats: Effects of parathyroidec- tomy and phosphate excretion. Am J Physiol 248F175- F182.

6% RODRIGUEZ ET AL.

22. Andress D, Felsenfeld AJ, Voigts A Llach F 1983 Parathy- roid hormone resonse to hypocalcemia in hemodialysis pa- tients with osteomalacia. Kidney Int 24:364-370.

23. Morrissey J, Rothstein M, Mayor G, Slatopolsky E 1983 Suppression of parathyroid hormone secretion by aluminum. Kidney Int 23:699-704.

24. Bourdeau AM, Plachot JJ, Cournot-Witmer G , Pointillawrt A, Balsan S, Sachs C 1987 Parathyroid response to aluminum in vitro: Ultrasturctural changes and PTH release. Kidney Int 31~15-24.

25. Diaz Lopez JB, D’Haese PC, Nouewen EJ, Lamberts LV, Cannata JB, DeBroe ME 1988 Estudio del contenido de alu- minio en paratiroides de ratas con insufficiencia renal e in- toxicacion aluminica cronica. Nefrologia 8:35-41.

26. Rodan GA, Martin TJ 1981 Role of osteoblasts in hormonal control of bone resorption: A hypothesis. Calcif Tissue Int 33~349-351.

27. Fallon MD 1983 Osteoblasts mediate osteoclastic bone re- sorption in cell culture (abstract). Calcif Tissue Int 35:640.

28. Anderson HC 1983 Calcific diseases. Arch Pathol Lab Med 107~341-348.

Address reprint requests to: Mariano Rodriguez, M.D.

Division of Nephrology (WIIIL) Wadsworth Veterans Administration Medical Center

Wilshire and Sawtelle Blvds. Los Angeles, CA 90073

Received for publication January 29, 1988; in revised form February 6, 1989; accepted March 6, 1989.