Cobalt: A review - Tahoma Clinic

Journal of Human Nutrition (1978) 32, 165-177

Cobalt : a review Andrew TAYLOR, MSC and Vincent M A M S , DM, FRCP (Lond. 8: Edin.) FRCPath Senior Bio chemist , So id t Iz bCc7st Tha m es R egio nal Heavy Metal La bo rat o ry , Department of Biochemistry, University of Surrey, Guildford, Surrev GU2 5XH; Professor of Clinical Biochemistry, Department of Biochemistry, Un>ucrsity o f Surrey, Guildford, Surrey GU2 5-1'H.

Introduction Cobalt, an essential trace element, is traditionaIIy thought of as being physio- logically active only when present as vitamin Biz, Among naturally occurring biologically active compounds this vitamin is one of the most potent with as little as 1 p g per day, containing 43 ng o f cobalt, being sufficient to avoid fatal pernicious anaemia (FAO/WHO, 19 7 0).

Inorganic cobalt is, however, by no means physiologically inert. Nevertheless, whilst it can activate the enzyme arginase (Anderson, 1945), and is responsible for a variety o f effects in several tissues, its importance for normal function and well-being has not been demonstrated.

The role of cobalt in vitamin-B12-mediated enzyme reactions involved with single carbon metabolism is well discussed in many text-books (eg, Thompson & Wootton, 1970), and will not be described here. We shall consider certain other aspects of the biochemistry and nutrition of inorganic cobalt, including its use as a pharmacological agent.

less real than apparent. Changes at a pharmacological level are undoubtedly gross manifestations of the biochemical actions of 'cobalt at a cellular level.

For convenience of discussion, this review is divided into sections that are

Physiology of cobalt Measurement of very low concentrations of trace elements presents considerable methodological problems. Much of the conflicting evidence reported in the litera- ture clearly reflects these problems and our understanding of the physiology of cobalt is an excellent example of how revision of previously accepted data inevitably follows the development of increasingly sensitive analytical techniques.

165

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Tor

onto

on

12/0

4/14

For

pers

onal

use

onl

y.

Dietary intake and absorption The comprehensive work of Tipton and her colleagues (Schroeder, Nason & Tipton, 1967; Tipton, Stewart & Martin, 1966) revealed that Western diets con- tain approxiinately 150450 g g cobalt per day. Sea-foods are the richest source of cobalt; bran and honey also contain more than 1 pg /g , but meat is probably the most important source. A considerably lower intake, 5-8 p g per day as measured by Harp & Scoular (1952). Regional variations, and choice of samples for analysis, may possible explain these results.

influenced by the concentration within the intestinal lumen, the ratio of cobalt to iron, and the status of body iron stores.

iron transport mechanism (Pollack, George & Reba, 1965). In rats, the addition of cobalt to oral test doses o f iron decreased the proportion of iron absorbed (Schade e t al., 1970). Studies with perfused duodenum demonstrated that iron and cobalt mutually compete, each inhibiting absorption of the other (Thomson, Valberg & Sinclair, 1971).

Iron deficiency, idiopathic haemochromotosis and sideroblastic anaemia promote the absorption o f cobalt in man. Iron overload, however, has no influence upon cobalt absorption, suggesting that cobalt absorption is responsive to the stimuli that promote iron absorption, but not to those which cause inhibition.

The proportion of ingested cobalt absorbed is variably reported and is probably

Absorption, through the duodenal mucosa, is effected by a common cobalt-

EN c re t io n In man the urinary concentration of cobalt was reported to be 50-190 pg/litre in eight subjects, much smaller amounts, 23-60 pg/day, being found in the faeces (Schroeder e t al., 1967; Tipton et al., 1966). It was from these data, together with analyses o f food, that cobalt balance in man was calculated (Table 1). The inference from this calculation is that at least 85 per cent of ingested cobalt is absorbed. More recently, however, much lower rates of urinary excretion have been reported, 0.5-2.2 pg/day by Cornelis, Speecke & Hoste (1975) and 0.12- 0.16 pg/litre by Lins & Pehrsson (1976), thus implying that a considerably smaller proportion of the intake is absorbed.

Table 1. Cobalt balance in man. (Schroeder eta/., 1967)

Food Water

Total

Intake (pdday)

300 6

306

Output (pcls/dav)

Urine 260 Faeces 40 Sweat and hair 6

306

The major route for the excretion of cobalt is via the kidney. Some is also eliminated into the intestine. Following subcutaneous injection of 57Co-labelled cobalt chloride to rats, 30 per cent of the administered radio-activity was recovered in the urine within 24 hours. Six days after injection, 63 per cent of the dose had been recovered in the urine and a further 12 per cent in the faeces (Shabaan & Marks, unpublished). These results are consistent with those of other workers

166

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Tor

onto

on

12/0

4/14

For

pers

onal

use

onl

y.

(Schroeder & Nason, 1971) and with observations in man. Two-thirds of intra- venously injected "Co appeared in the urine in 48 hours (Paley 8: Sussman, 1963). Lii-inary excretion following oral administration of radioactive cobalt was 9-23 per cent within the first 24 hours - further evidence that the majority of dietary cobalt is not absorbed (Sorbie et al., 1971).

Dis trib u t io T I

Table 2 shows the distribution of cobalt measured by atomic absorption spectro- photometry in tissues removed from rats one week after receiving ten daily subcutaneous injections of cobalt chloride, 40 mg/kg.

and spleen also contained considerable quantities. Using 57CoC12 .6&O a similar distribution was observed. In addition, cobalt was detected at low levels in other sol't tissues examined and in bone which contained approximately 25 per cent of the amount present in the liver (Shabaan, Marks 8: Taylor, unpublished).

Table 2. Total tissue cobalt concentration (mean k s.e.m.) in cobalt chloride treated rats

While the largest amount o f cobalt was found in the liver, the kidney, pancreas

Tissue

Liver Kidney Pancreas Spleen Blood

Cobalt

21 80 30 130k 10 32k 7 1 0 2 2 1 2 2 2

(P9)

% of Administered

dose 11.0 0.66 0.16 0.05 0.06

In the analyses of human tissue removed at autopsy the greatest concentration wdS found in adipose tissue, with liver and heart also containing substantial amounts (Tipton & Cook, 1963). However, the analytical technique employed, emission spectroscopy, is less sensitive and, since many samples contained amounts below the detection limit results are incomplete. It was suggested by Schroeder & co-workers (1967) that tissue concentrations are unaltered with age: but, since so many results were less than the limit of detection, these conclusions may not be justified and this question should be reassessed.

The total body load in man has been measured by Yamagata, Murata & Torii (1962) using neutron activation analysis. Bone contained 14 per cent of the total cobalt; 43 per cent was in muscle and the remaining 43 per cent in other soft tissues. The total body cobalt was estimated at 1.1 mg.

Problems associated with the measurement o f very low concentrations have bedevilled attempts to determine cobalt in blood and other fluids. Table 3 summarises published concentrations and illustrates the disparate findings. There is some concensus among those reporting very low concentrations, but whether these prove to be correct awaits verification.

Pharmacological effects of cobalt Either by accident or design, inorgahic cobalt has been administered in relatively large (milligram) amounts t o subjects. Whilst cobalt chloride is a recognised cytotoxic compound, and its adminsitration can be harmful, under some circum- stances a beneficial response is produced.

16

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Tor

onto

on

12/0

4/14

For

pers

onal

use

onl

y.

Table 3. Reported cobalt concentration in body fluids of healthy individuals

Sample

Whole blood Whole blood Plasma Serum Serum Serum Serum Serum Serum

Concentration range (ng/100 ml)

350-6300 7-36 18

7000 37-100

6-8 8-58 2-6 4-26

Reference

Koch et a/. (1 956) Thiers et a/. (1 955) Schroeder & Nason (1971) Kesteloot et a/. (1 966) Koch e t a/. (1 956) Thiers eta!. (1955) Parr & Taylor (1 964) Lins & Pehrsson (1976) Versieck eta / . (1976)

Ery thropoicsis For more than 40 years it has been known that cobalt produces polycythaemia in many animals and also in man (Orten, 1936; Berk, Burchenal & Castle, 1949; Duckham & Lee, 1976). It has been shown that the concentration of circulating erythropoietin, a glycoprotein hormone which stimulates bone marrow stem cells to produce mature erythrocytes, is increased in cobalt-treated animals (Brown & Meineke, 1958; Janda, Fried & Gurney, 1965). The probable stimulus to the secretion of erythropoietin is tissue hypoxia (Goldwasser et al., 1958; Fried & Kilbridge, 1969) subsequent to cobalt inhibition of tissue respiration and oxida- tive phosphorylation (see page 173).

tage has been taken of the polycythaemic action of cobalt. Robinson, James & Kark (1949) reported increased total circulating haemoglobin, red-cell count, haematocrit and haemoglobin concentration in patients with chronic suppurative infections treated with cobalt. Berk & co-workers (1949) and Gross & Spaet (1953) described similar responses in patients with anaemia of chronic renal disease and sickle cell anaemia respectively. Table 4. Increase in blood haemoglobin concentration in anaemic patients following 12 weeks oral cobalt therapy (50 mg CoC12.6H20 per day). (Duckham & Lee, 1976)

In types of anaemia that are resistant to conventional forms of therapy, advan-

Haemoglobin ( g / l O O ml)

Patient % increase Before Co After Co

5.1 8.7 4.5 6.3 5.2 7.0 6.9 4.9

7.9 11.0 9.1 7.3 6.1 8.7

10.8 6.7

55 26

102 16 17 24 57 37

The main site fo'r the synthesis of erythropoietin is the kidney although other sites exist which assume particular importance following nephrectomy. Duckham & Lee (1976) treated a group of anephric, refractory anaemic, dialysis patients with cobalt chloride by mouth in enteric-coated capsules at doses of 25 or 50 mg

168

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Tor

onto

on

12/0

4/14

For

pers

onal

use

onl

y.

per day for 1 2 to 32 weeks. In addition to responding by increasing their haemo- globin concentrations by an average of 46 per cent (Table 4): there w d S also a decrease in thc need for blood transfpsions (Fig. 1).

a * T T

:t I’ --- * *f I , , , T

G(l ill

( m d d v ) 2 0 5 a m -.... , , . , , I . , I , I , I , , . , , , , , ,

51, >#,Il,\

Fig. 1. Patient M.V.: response to three courses of cobalt chloride. Note excellent response to half dose in third course. All divisions on the abscissa are one-month intervals. Size of ’T’ arrow relates to a 2- or 4-unit blood transfusion. (Duckham & Lee, 1976. Reproduced by kind permission of the authors and Oxford University Press)

Use o f cobalt in the treatment o f anaemia has never been widespread; however, the intermittent reports of this therapy reflect the continuing requirement for a potent crythropoietic stimulant. Despite certain reservations by some authorities (Curtis et al. , 1976), the considerable improvement in the patient’s well-being demonstrates the value o f this treatment.

Liparmia In addition t o the effect upon erythropoiesis, cobalt has repeatedly been shown to induce hyperlipaemia. Thus rabbits (Boyd & hlaclean, 1959; Caren & Carbo, 1956; Reber & Studer, 1961; Caplan & Block, 1963; Telib & Schmidt, 1973), chickens (Tennent et al., 1958) and rats (Reber & Studer, 1961; Lochner, Eisen- traut & Unger, 1964; Eaton, 1972; Shabaan & Marks, 1975) all responded to treatment with cobalt with increased serum lipid concentrations. Table 5. Fasting plasma lipids (mean k s.e.m.) in cobalt-treated rats and their controls (Six animals in each group). (Taylor et a/. , 1977)

Time from final injection

2 weeks Control

Co-treated

Triglyceride mmol/l

0.58 _+ 0.04

“1.14 _+ 0.03

Control

Co-treated

Control

Co-treated

8 months

12 months

“Significant difference from control

0.57 f 0.03

“1.24 + 0.05

Cholesterol mmol/l

1.97 5 0.08

2.08 _+ 0.16

2.52 f 0.16

“3.17 f 0.16

FFA pmmol/l

654 2 20

6785 15

-

0.61 _+ 0.03

*I .80 f 0.08

3.22 2 0.10

*4.37 ? 0.1 3

754 2 24

1245k42

169

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Tor

onto

on

12/0

4/14

For

pers

onal

use

onl

y.

In the rat (Table 5 ) and in man (Table 6) cobalt produces prompt elevation of serum triglycerides with slower, less marked rises in the concentration of cholesterol and free fatty acids (Taylor et al., 1977). Table 6. Fasting plasma lipids (mean + s.e.m.1 in cobalt-treated anephric patients (Taylor e t a / . 1977)

Triglyceride Cholesterol (mmol/l) (mmol/l)

Approximate normal range

renal function

Before cobalt treatment 6.57 2 0.45 (6) At completion of treatment "3.5 +_ 0.57 (3) 7.64 0.79 (5) 1-2 months af ter Co 6.09 L 0.40 (3) 5-6 months after Co 7.1 5 2 0.49 (5)

in patients with normal 0.3-1.7 3.6-7.8

1.7 f 0.25 (1)

"2.64 & 0.23 (3) 2.1 1 +_ 0.27 (6)

*Significant difference from pretreatment value

The mechanism o f the lipaemic response to cobalt is not understood. Histo- logically, pancreatic alpha-cell degranulation and destruction arc observed tollow- ing administration of cobalt. Van Campenhout 8: Cornelis (1960) proposed, therefore, that chronic glucagon deficiency ensued with consequent hyper- glycaemia and lipaemia. In support of this hypothesis, glucagon is known to have a hypolipaemic effect (DeOya et al., 1971), to promote catabolism, and to reduce hepatic protein synthesis (Pryor & Berthet, 1960). But in the rat, radio- immunoassay has revealed elevated concentrations of plasma glucagon following cobalt chloride treatment (Lochner e t al., 1964; Eaton, 1973). Eaton (1972) also reported that the cobalt-treated rat had increased sensitivity to insulin.

The enigma posed by apparent increased insulin sensitivity in the presence of increased circulating glucagon may be reconciled by assuming hepatic resistance to glucagon (Eaton, 1973). However, data inconsistent with this attractive hypothesis have been produced; with a pancreatic-glucagon specific radioimmunoassay plasma glucagon concentrations in cobalt-treated rats are not significantly different from untreated controls. The fasting levels, as measured by this assay, are comparable to concentrations observed in man, and patients treated with cobalt for the refractory anaemia of chronic renal failure have levels of plasma glucagon similar to patients not treated with cobalt (Al-Tamer, 1978). In addition, glucagon receptors on purified hcpatocytc membranes showed no differences in membranes from cobalt-treated animals or their controls (Taylor et al., 1977).

Notwithstanding the unresolved situation concerning the mechanism of the lipaemic response, the cobalt-treated animal presents a useful experimental model for the assessment of hypolipaemic agents and dietary regimes before using them in humans. In anephric patients treated with cobalt for anaemia and receiving regular haemodialysis, the lipaemie response is not maintained beyond the period of therapy, thereby minimising a potentially unwelcome side-effect.

Cardiomy opa thy The widespread use of synthetic detergents for washing glassware presented the brewing industry with a serious problem. Despite rinsing, a film of detergent

170

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Tor

onto

on

12/0

4/14

For

pers

onal

use

onl

y.

remained within thc glass and as beer was poured in it the anticipated froth or head f i l e d to matcrialisc. Cobalt, it was discovered, added to beer at a concen- tration o f 1 to 1.5 p.p.m. (as cobalt chloride), would stabilise the froth. Breweries in Belgium, in Quebec City, and in Omaha, Nebraska and Minneapolis in the USA are all known to have used cobalt in this way. In addition, it is suspected that beer containing cobalt was sold in othcr American cities and in London (Alex- ander, 1972).

In 1966 reports began to appear o f a new syndrome of cardiomyopathy in subjects consuming large volumes o f beer containing added cobalt (Morin e t al., 1967; McDermott ct al., 1966; Kesteloot et al., 1966). Following the incrimina- tion o f cobalt in this outbreak and the discontinuation o f its addition to beer, no further cases presented.

The syndrome was characterised by fulminating heart failure in heavy beer- drinkers. Severity on presentation ranged from mild congestive failure to cardio- genic shock with metabolic acidosis. Typically symptoms had been present for about ten weeks, contrasting with ordinary alcoholic cardiomyopathy and thiamin- deficient cardiomyopathy which progress much more slowly. Pericardial effusion and elevated haemoglobin concentrations were noted in the majority of cases and were unique to this cardiomyopathic syndrome. A gallop rhythm was heard in two-thirds of the patients; electrocardiographic changes included sinus tachy- cardia and low QRS amplitude in almost 100 per cent, and about 50 per cent, of cases respectively. Very high serum levels of cardiac enzymes were measured, eg, creatinine phosphokinase [ 5,500 units (normal 6-235)] aspartate transaminase [4,400 units (normal 0-40)], and rose still further after initial treatment. Alex- ander (1972) suggested an accumulation of eiizyme within damaged ischaemic muscle and liver which was washed out into the circulation after recovery from shock, to explain this secondary rise. Normal serum enzyme levels were restored in 10-12 days. Light microscopy of myocardium revealed few abnormalities, but all samples examined by electron microscopy showed degeneration and disorien- tation of myofibres, swollen mitochondria with absent cristae and large deposits of glycogen. Mortality from the acute illness ranged from 0 to 47 per cent in the various reports.

Before this outbreak of cardiomyopathy there was little evidence linking cobalt with myocardial damage. Berk & co-workers (1 949) using cobalt as an erythro- cyaemic agent suggested that it might precipitate attacks of angina pectoris. How- ever, the amounts of cobalt consumed by beer-drinkers were never greater than 10 mg per day compared with 25 or 50 mg used in the treatment of anaemia. How cardiotoxic then is cobalt and under what circumstances does it become so?

In acute toxicity studies with rats the intraperitoneal mean lethal dose (LD5 ) was 12.9 mg/kg, the fasted oral LDso was 170 mg/kg, and the non-fasted oral LD50 was 600 mg/kg. Administered a t these doses, death due to heart failure ensued within a few hours. Oral toxicity was considerably reduced when amino acids, especially histidine, or soy-protein, milk powder, o r egg white, were fed with the cobalt. The concentration of cobalt in heart tissue was similarly reduced when protein and amino acids were added to the diet (Wiberg e t al., 1969).

Attempts to reproduce the beer-drinkers’ syndrome in rats, by administering beer containing cobalt equivalent to the amounts consumed by heavy beer-drinkers

1 7 1

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Tor

onto

on

12/0

4/14

For

pers

onal

use

onl

y.

were unsuccessful as assessed by serum cardiac enzymes, weight gain, food con- sumption and organ weights or histological examination. However, animals maintained on a poor protein diet and with beer as the source of fluid ~ve1-e more likely to develop myocardial histological changes than those treated xvith cobalt alone (Grice rt ul., 1969).

In his thorough assessment of beer-drinker's cardiopathy, .Alexander ( 19 72) drew attention to the deleterious effects of ethanol upon the heart and the requirement of adequate protein, energy and mineral intake for maintenance of a healthy myocardium. These nutritional factors, he noted, were probably deficient in the affected subjects. They were, therefore, particularly vulneruble to the cardiotoxic effects of cobalt which otherwise are manifest only at consider- ably greater levels.

Mis cellu n eo us In some anaemic patients given cobalt, side-effects develop which neccssitate interruption or termination of the treatment. Raw cobalt chloride causes gastro- intestinal symptoms, eg nausea, vomiting and anorexia, but with present regimes of low doses given in enteric-coated capsules these effects are relatively uncommon. Peripheral neuropathy, tinnitis and auditory nerve damage develop on rare occa- sions (Duckham & Lee, 1976). Cobalt also interferes with thyroid function; reduced thyroidal I'31 uptake, goitre and myxoedema have all been described (Kriss, Carnes & Gross, 1955). The incidence of thyroid dysfunction may be as high as 50-70 per cent of cases (Paley, Sobel 8s Yalow, 1958; Hopper, Anderson & Dailey, 1959), although Duckham & Lee (1976), using smaller doses of cobalt, had only yne patient out of eight with evidence of thyroid hypofunction.

Tu mo urs In common with many unrelated materials, injection of metallic cobalt powder induced the development of sarcomas a t the site of administration (Roe & Lancaster, 1964). Of 20 rats treated with cobalt chloride by subcutaneous injection, 11 survived for 12 months. Eight of these survivors dcveloped fibro- sarcomas and in four of the rats the tumours were found remote from the site of injection (Shabaan et al., 1977). Furthermore, cobalt chloride treatment of tissue cultures produces cells suggestive of mdlignant alterations (Heath, 1953).

for anaemia with oral cobalt chloride (300 mg per day for two months), who developed tumours. One had a reticulum cell sarcoma and the other a giant follicular lymphoma and, with the exception of these early cases, there are no reports of tumour formation in cobalt-treated patients.

As a consequence of these side-effects, some subjects are denied the consider- able haematological improvement afforded by cobalt therapy. Selection of patients and careful supervision during treatment has led to elimination of undesirable effects in recent studies. It now remains to be seen whether smaller doses of cobalt could be given, thereby making it possible to treat more patients with safety and with response.

Berk & co-workers (1949) described two patierits from a group of 61, treated

172

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Tor

onto

on

12/0

4/14

For

pers

onal

use

onl

y.

Biochemical effects of cobalt Erzzy me inhibition Cobalt can, by reacting with thiol groups o f amino acids, proteins and co-factors, inactivate several biochemical pathways, particularly those associated with energy production. Dihydrolipoic acid is required for the entry of pyruvate into the tricarboxylic acid c)rcle and the conversion o f 0-oxoglutaric acid to succinyl coenzyme A. This enzyme co-factor is inactive when bound by cobalt (Webb, 1962, 1964), possibly restricting thereby the activity of the tricarboxylic acid cycle and production of ATP. Yastrebov (1966) demonstrated that cobalt does indeed decrease tissue respiration and oxidative phosphorylation i y z vitro. Oxygen uptake by kidney, liver and bone marrow was decreased by 14-25 per cent when cobalt-rich plasma was added to the incubation medium. Production of organic phosphate was similarly reduced in kidney, liver and cardiac muscle tissue.

To ascribe all the observed effects of cobalt as sequelae of its inhibition o f respiration and oxidative phosphorylation makes an attractive hypothesis. As has been discussed above, tissue hypoxia is almost certainly the stimulus for the release of erythropoietin; gastrointestinal symptoms could be the consequence of local anoxia within the tract and nerve damage could be similarly induced. However, more plausible explanations exist for the interference in thyroid metabolism and production of lipaemia. It has been suggested (S. Kirkwood, personal communication cited in Kriss et al., 1955), although the experimental data were never published, that tyrosine iodinase is inhibited by cobalt ions. This communication is often referred to and if the important link between laboratory results and clinical observations is to be accepted, confirmation by independent workers must be forthcoming.

Hypertriglyceridaemia may develop as a consequence of increased triglyceride and lipoprotcin synthesis, or decreased breakdown, or a combination of both. The proposal by Eaton (1973) that cobalt causes hepatic resistance to glucagon, thereby failing to control production of apoprotein and triglyceride, was discussed above. Recently we have shown (klahmood & Marks, unpublished) that the enzyme responsible for the catabolism of circulating lipoproteins, lipoprotein lipase, is inhibited both in viuo and in vitro by cobalt.

The plasma post-heparin lipoprotein Iipase activity in cobalt-chloride-treated rats was about 75 per cent of that of untreated controls. The reduced enzymic activity was reflected in delayed triglyceride clearance following oral administr- ation of Intra-Lipid. Lipoprotein lipase activity in plasma from both untreated and cobalt-chloride-treated rats was reduced by almost 80 per cent when assayed in the presence of 10 mmol Co++ in vitro.

Cobalt inhibition of lipoprotein lipase could explain the hypertriglyceridaemia of cobalt-treated rats. Whether this mechanism is also involved in the lipaemia observed in cobalt-treated patients is now under investigation.

Cobalt and haem metabolism Cobalt has a profound effect upon hepatic haemoproteins, particularly cyto- chrome P450. This cytochrome which is involved with microsomal oxidative reactions was decreased by 40 per cent in rats following subcutaneous injection of cobalt chloride (250 pg/kg) and the total microsomal haem content was also

173

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Tor

onto

on

12/0

4/14

For

pers

onal

use

onl

y.

reduced (hlaines & Kappas, 1976). Cobalt inhibits haem synthesis (De Matteis & Gibbs, 1976; Maines et al., 1976) and induces the enzyme haem oxygenase (Maines & Kappas, 1974).

A M I N ~ L E V U L I N I C A C W

I ALA dellydrn'e

CO++ I -?

4 C" ' * \ , + 1 / Fe H A E M O G L O B l N t H A E M P PROTOPOI!FIII 1!1h - Cu - I'ROTOPORPHYRIN CATALASE CYTCCHROMES ferroChe''tdSe

+ oxygenase + \ / +

C o ' ' - ::< INTRACELLULAII BILUiUBIN IIIVUING SITE

Fig. 2. Effects of cobalt on haem metabolism

In the synthesis of haem, the rate limiting step is the formation of delta amino levulinic acid (6 ALA), catalysed by 6 ALA synthetase. A feedback system operates whereby haem inhibits the activity of this enzyme (Fig. 2). De Matteis & Gibbs (1977) showed rhat 80 per cent of 6 ALA synthetase activity was lost two hours after treatment with cobalt chloride (1 5mg/kg) subcutaneously and this low level of activity persisted for at least 5 hours.

De Matteis & Gibbs (1977) measured fhe radioactivity in crystallised haemin extracted from livers of rats injected with ["C] 6 ALA. The radioactivity in haemin from cobalt-treated rats was about 25 per cent of that from the controls. Cobalt, therefore, interferes at a second site directing the 6ALA into non-haem compound(s). These authors suggest cobalt-protoporphyrin as a possible candi- date for non-haem compound and that this may, in addition, be responsible for inhibiting 6 ALA synthetase activity by a mechanism similar to the feedback inhibition of haem.

Maines & co-workers (1976) demonstrated that the decreased activity of GALA synthetase described above was followed by an increase to about twice that of the controls. They attributed this rebound to depletion of hepatic haem (through the reactions already described and by the action of haem oxygenase) inducing dc nouo synthesis o f enzyme protein.

The importance of haem oxygenase in this mechanism is evident from the sequence of events invariably observed - increased haem oxygenase activity, followed by decreased hepatic concentration o f haem and cytochrome P450, followed by induction of 6ALA synthetase (Maines et al., 1976). Cobalt chloride, at a dose of 60 mg/kg subcutaneously, induced a seven-fold increase in haem oxygenase activity (a response which was reduced, although not entirely abolished, by puromycin and actinomycin D), indicating that synthesis of new enzyme was responsible for much of the increased enzyme activity (Maines & Kappas, 1975).

It is unclear what relevance these studies have to the treatment of anaemic patients with cobalt. Since haemoglobin concentration and red cell count

.

174

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Tor

onto

on

12/0

4/14

For

pers

onal

use

onl

y.

increase, thc decreased synthesis o f haem must be over-ruled by the erythro- poietic response to cobalt. Possibly the effects upon haem metabolism which are produced acutely are not maintained with repeated exposure to cobalt.

Miscellaneous effects of cobalt M r t a l sensitivity and total joint replacements Alloys o f cobalt, chromium, nickel and molybdenum have been successfully used in joint prostheses for many years. Several causes for the small percentage of joint loosening have been recogniscd, eg infection, trauma, faulty implantation and poorly desi<gned prostheses. Immunological sensitivity has also been suggested as responsible fo r failure.

implants (Winter, 1974), raised levels of cobalt and chromium are found in the blood and urine of patients with metallic total hip replacements (Coleman, Herrington & Scales, 1973) and the incidence of metal sensitivity (as deter- mined by skin patch tests) is about ten times that of subjects without implants (Elves e.t al., 1975).

Evans & co-workers (1974) suggested that wear causes tissue sensitisation and ultimately degeneration of bone with joint loosening. However, sensitivity could occur as a consequence rather than cause of loosening and Elves & co-workers (1975) described their intention to examine patients serially to determine the appropriate sequence. Circumstantial evidence against sensitivity leading to loosening comes from the observation that a high incidence of sensitivity (20 per cent) is found in patients with implants; there is a very low incidence of failure; there are many examples of patients with metal sensitivity having perfectly functioning joints (McKee, 1975).

Particulate matter containing alloy fragments are found around surgical

Cobalt excretion test - an index of body iron Investigation of anaemia is aided by knowledge of total body iron stores. While several techniques of assessment have been described, all have their limitations. Stainable iron in bone marrow aspirates is the most satisfactory; but, because it is traumatic, simpler tests are desirable. Since the rate of absorption of iron is influenced by body stores an indirect measurement can be made from intestinal absorption. However, measurement of iron absorption is time-consuming and requires very specialised apparatus.

As discussed above (Physiolocgy o f cobalt), studies on iron and cobalt indicate that a shared mucosal pathway exists. This led to the development of a test where the urinary excretion of orally administered cobalt gives an indirect measurement of iron absorption (Valberg et al., 1972).

To examine the reliability of this test Wahner-Roedler, Fairbanks & Linman (19 75) investigated patients with a variety of disorders of iron metabolism. Cobalt excretion and bone marrow iron were compared in normal volunteers, blood donors and patients with chronic blood loss, sideroblastic anaemia, haemo- lytic anaemia, lymphoma, idiopathic thrombocytopaenic purpura and other malignancies or chronic inflammatory processes. Blood donors and patients with chronic bleeding all had increased cobalt excretion. Patients with sideroblastic anaemia also had increased cobalt excretion. Among the patients with other

175

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Tor

onto

on

12/0

4/14

For

pers

onal

use

onl

y.

disorders o f iron metabolism there was no correlation between bone marrow iron and excretion o f cobalt.

The cobalt excretion test does not appear, therefore, to be a reliable index oi' body iron stores other than in conditions of blood loss, when it is cxtremcly sensitive. Other factors are involved in the regulation of iron and cobalt absorp- tion in addition to the body iron stores. Thus in sideroblastic anacmia ivherc increased iron absorption occurs in the presence o f more than adequate body iron, cobalt excretion was also increased. In addition, accurate collection o f urine is required and results may be unreliable when patients have impaired rend function (Vdberg e't al., 1972).

Conclusion The effects of inorganic cobalt are many and varied and, while pharmacological actions have received considerable attention in the past, it is on thc response to physiological amounts o f cobalt ions that emphasis must now bc placed.

Regulation of intestinal absorption implies that, in addition to its presence in vitamin Biz, cobalt is also an essential biological element in the inorganic form. Present understanding of the physiology and biochemistry of inorganic cobalt indicates that it may be important in the regulation or reactions of certain enzymes and that it is possibly involved with the metabolism of iron.

questions is now within reach. With recent developments of sensitive analytical techniques resolution of these

References Alexander, C.S. (1972): A m . J . Med. 53, 395. Al-Tamer, Y.Y. (1978): PhD thesis, University of Surrey. Anderson, A.B. (1945): Biochem. J . 39, 139. Berk, L., Burchenal, J.H. & Castle, W.B. (1949): New EnZg.J. M e d . 240, 754. Boyd, G.S. & MacLean, N. (1959): Q.J.Exp. Physiol. 44, 394. Brown, T.E. & Meineke, H.A. (1958): Proc. Soc. Exp. Biol. (NY) 99, 435. Caplan, R.M. & Block, W.D. (1963): J. Invcst. Dermatol. 40, 199. Caren, R. & Carbo, L. (1956):J. Clin. Endocr. Metab. 16, 507. Coleman, R.F., Herrington, J. & Scales, J.T. (1973): Br. Med.1. 1, 527. Cornelis, R., Speeke, A. & Hoste, J . (1975): Anal. Chim. Acta 78, 317. Curtis, J.R., Goode, G.C., Herrington, J . & Urdaneta, L.E. (1976): Clin. Ncphrol. 5, 61. De Matteis, F. & Gibbs, A.H. (1976): Ann. Clin. Res. 8, Suppl. 1 7 , 193. De Matteis, F. & Gibbs, A.H. (1977): Biochem. J . 162, 213. De Oya, M., Prigge, W.K., Swenson, D.E. & Grande, F. (1971): A m . J . Physiol. 221, 25. Duckham, J.M. & Lee, H.A. (1976): Q. J. Med. 45, 277. Eaton, R.P. (1972): Am. J. Physiol. 222, 1550. Eaton, R.P. (1973): Am.]. Physiol. 225, 67. Elves, M.W., Wilson, J.N. Scales, J.T. & Kemp, M.B.S. (1975): Br. Mpd:J. 4, 376. Evans, E.M., Freeman, M.A.R., Miller, A.J. & Vernon-Roberts, B. (1974): J B o n e J t Surg.

FAO/WHO (1970): Requirements of ascorbic acid, vitamin D, vitamin B I Z , folate and iron. 56B, 626.

WHO Tech. Rep. Series No. 452. World Health Organisation: Geneva and FA0 Nutrition Meetings Rep. Ser. No. 41. Food and Agriculture Organisation: Rome.

Fried, W. & Kilbridge, T. (1969): J. Lab. Clin. Med. 74, 623. Goldwasser, E., Jacobson, L.U., Fried, W. & Plzak, L.F. (1958): Blood 13, 55. Grice, H.C., Mungo, I.C., Wiberg, G.S. & Heggtveit, H.A. (1969): Clin. Toxicol. 2, 273. Gross, R.T. & Spaet, T.H. (1953): Stanford Med. Bull. 11, 196. Harp, M.J. & Scoular, F.I. (1952): J . Nutr. 47, 67.

176

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Tor

onto

on

12/0

4/14

For

pers

onal

use

onl

y.

Heath, J.C. (1954): Exp. Cell. Rcs. 6, 311. Hopper,J., Anderson, B.G. & Dailey, M.E. (1959): Am. J . hled. sci. 238, 102. Janda, W.E., Fried, W. & Gurney, C.W. (1965): Pror. Soc. Exp. Biol. (NY) 120 ,443. Kesteloot, H., Terryn, R., Bosmans, P. & Joossens, J.V. 1966): Acta Cardiol. (Brux) 21, 341. Koch, HJ., Smith, E.R., Shimp, N.F., & Conner, J. (1956): Canccr 9, 499. Kriss,J.P.,Carnes,W.H. &Gross , R.T. (1955):J. Am. Mcd , Ass. 157, 117. Lins, J,.E. & Perhsson, K. (1976): Lancet 1, 1191. Lochner,J. de V., Eisentraut, A.M. & Unger, R.H. (1964): Metabolism 13, 868. Maines, M.D., Janousek, V., Tomio, J.M. & Kappas, A. (1976): Proc. Natn. Acad. Sci. USA

Maines, M.D. & Kappas, A. (1974): Proc. Natn. Acad. Sci. USA. 71, 4293. hlaines, M.D. & Kappas, A. (1975):J. Biol. Chcm. 250, 4171. Maines, M.D. & Kappas, A. (1976): Biochem. J . 154, 125. McDermott, P.H., Delaney, R.L., Egan, J.D. & Sullivan, J.F. (1966): J . Am. Aled. Ass . 198,

McKee, G.K. (1975): Br. N e d . ] . 4, 646. Morin, Y.L., Foley, A.R., Martineau, G. & Roussel, J. (1967): Can. hled. Ass. J . 97, 881. Orten, J.M. (1936): A m . J . Physiol. 114, 414. Paley, K., Sobel, E. & Yalow, R. (1958):J. Clin. Endocr. 18, 850. Paley, K.P. & Sussman, E.S. (1963): Mc.tabolisnr 12, 975. Parr, R.M. & Taylor, D.M. (1964): L3iochcm.J. 91, 484. Pollack, S., George, J.N. & Reba, R.C. (1965): J . Clin. Inzrest. 44, 1470. Pryor, J. & Berthet, J. (1960): Biochim. Biophys. Acta 43, 556. Reber, K. & Studer, A. (1961): ExperimentaI lipaemia and attempts t o its control. In Drugs

affecting lipid metabolism, ed S. Garattini & R. Paoletti, p. 181. Amsterdam: Elsevier. Robinson, J.C., James, G.W. & Kark, R.M. (1949): New Engl. J . hled. 240, 749. Roe, F.J.C. & Lancaster, M.C. (1964): Br. hled. Bull. 20, 127. Schade, S.G., Feisher, B.F., Bernier, G.M. & Conrad, M.E. (1970): J . Lab, Clin. Med. 75, 435. Schroeder, H.A. & Nason, A.P. (1971): Clin. Chem. 17, 461. Schroeder, H.A., Nason, A.P. & Tipton, I.H. (1967): J . Chron. Dis. 20, 869. Shabaan, A.A. & Marks, V. (1975): Diabetologia 11, 376. Shabaan, A.A., Marks, V., Lancaster,M.C. & Dufeu,G.N. (1977): Lab. Animals 11, 43. Sorbie, J., Olatunboxun, D., Corbett, W.E.N. & Valberg, L.S. (1971): Can. hled. A s s . J . 104,

Taylor, A., Marks, V., Shabaan, A.A., Mahmood, M.A., Duckham, J.M. & Lee, H.A. (1977):

73, 1499.

253.

777.

Cobalt induced lipaemia and erythropoiesis. In Clinical chemistry and chemical toxicology o f metals, ed. S.S. Brown, p. 105. Amsterdam: Elsevier/North-Holland.

Telib, M. & Schmidt, F.H. (1973): Endokrinologie 61, 395. Tennent, D.M., Mushett, C.W., Kuron, G.W., Ot t , W.H. & Siegel, H. (1958): Proc. Soc. E x p .

Thiers, R.E., Williams, J.F. & Yeo, J.H. (1955): Anal. Chem. 27, 1225. Thompson. R.H.S. & Wootton, I.D.P. (1970): Biochemical disorders in human disease, 3rd edn.

Thomson, A.B.R., Valberg, L.S. & Sinclair, D.G. (1971): J . Clin. Invest. 50, 2385. Tipton, I.H. & Cook, M.J. (1963): EIlth Physics 9, 103. Tipton, I.H., Stewart, P.L.'& Martin, P.G. (1966): Hlth Physics 12, 1683. Valberg, L.S., SDrbie, J. & Corbett, W.E.N. (1972): Ann. Intern. Med. 77, 181. Van Campenhout, E. & Cornelis, G. (1960): C R S O C . Biol. 145, 933. Versieck, J., Hoste, J., Barbier, F. & Vanballenberghe, L. (1976): Lancet 1, 1403. Wahner-Roedler, D.L., Fairbanks, V.F. & Linman, J.W. (1975): J. Lab. Clin. Med. 85, 253. Webb, M. (1962): Biochim. Biophys. Acta 65, 47. Webb, M. (1964): Biochim. Biophys. Acta 89, 431. Wiberg, G.S., Munro, I.C., Meranger, J.C., Morrison, A.B., Grice, H.C. & Heggtveit, H.A.

Winter, G.D. (1974): J . Biomed. h la f . Res. 8, 11. Yamagata, N., Murata, S. & Torii, T. (1962): J. Rad. Res. 3, 4. Yastrebov, A.P. (1966): Fed. Proc. 25, Transl. 630.

Biol. hled. 98, 474.

London: Churchill.

(1969): Clin. Toxicol. 2, 257.

177

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Tor

onto

on

12/0

4/14

For

pers

onal

use

onl

y.