R E M O V A L OF M E T A L I O N S F R O M A Q U E O U S S O L U T I O N S BY PENICILLIUM

BIOMASS :

KINETIC AND UPTAKE PARAMETERS

M. G A L U N

Department of Botany, The George S. Wise Faculty of Sciences, Tel-Aviv, Israel E. G A L U N

Department of Plant Genetics, Weizmann Institute of Science, Rehovot, Israel B. Z. S I E G E L

Pacific Biomedical Research Center, University of Hawaii at Manoa, Honolulu, HI 96822, U.S.A. P. K E L L E R

Department of Botany, The George S. Wise Faculty of Life Sciences, Tel-Aviv, Israel H. L E H R

Department of Botany, The George S. Wise Faculty of Life Sciences, Tel-Aviv, Israel and S. M. S I E G E L *

Department of Botany, University of Hawaii at Manoa, Honolulu, H1 96822, U.S.A.

(Received March 28, 1986; revised July 14, 1986) Abstract. The uptake and binding ofNi, Zn, Cd, and Pb by the mycelium ofPenicillium digitatum are highly pH-sensitive, being severely inhibited below pH 3. In the case ofNi, Zn, and Cd, H + inhibits competitively. The Cu-ion, like UO~ + studied previously, is nearly pH-insensitive. All of these cations except Pb are taken up to a greater extent by mycelial preparations preheated at 100 ° C for 5 min. Other activators include alkali and dimethyl sulfoxide (DMSO) pretreatment, but formaldehyde inhibits, Combining current and previous data, the ion-selective character of uptake is reflected, on a molar basis by the rank order Fe 3 +, Ni 2 +, Zn 2+ > Cu2+ > Pb2+UO~ + ,> MOO]-. P. digitatum appears to act like a mixture of neutral and acidic glycans with no real evidence for cationic amino-functional sites. In addition to the technological applications in water treatment, we suggest that fungal biosorption may be of natural geochemical importance in the concentration of metals and formation of minerals.

1. Introduction

The uptake of metals from aqueous solution by bacteria, yeast, filamentous fungi, and algae has received a great deal of attention over the past decade. Heavy radionuclides, especially Ra, Th, and U, have been singled out (Galun etaL, 1983a, b,c, 1984; Sakaguchi et al., 1978) although a wide range of non-radioactive heavy metal ions from Fe to Pb has also commanded attention (Beveridge and Murray, 1976, 1978, 1980; Corpe, 1975; Crist et al., 1981; Heldwein et al., 1977; Norris and Kelly, 1979; Olafson et al., 1979). Recent work in our laboratories has shown that heat-activated (killed) * Author for all correspondence.

Water, Air, and Soil Pollution 33 (1987) 359-371. © 1987 by D. Reidel Publishing Company,

360

M. GALUN ET AL.

Penicillium digitatum mycelium can accumulate Fe 3 +, Ni 2 +, Cu 2 +, Zn 2 +, Cd 2 +, and Pb 2 +, as well as UO 2 +, but not CrO 2- or MoO42- (Galun, E. et al., 1983). In the course of a study of uptake determinants such as pH, and ion-parameters, a number of differences as well as similarities to other complex exchange systems have been noted comparatively (Tsezos and Keller, 1983; Tsezos and Volesky, 1981, 1982a, b). This report seeks to characterize the effects of heat and chemical pretreatment, and pH on uptake and mycelial loading as related to specific ion properties, and to facilitate comparisons with other systems. 2. Materials and Methods

Most of the methods and procedures used in conjunction with this study have been described in detail (Galun et al., 1983a, b, 1984; Galun, E. et al., 1983; Siegel et al. 1983, and will only by summarized here. All experiments were carried out using an isolate ofPenicilliurn digitatum from orange peel (Galun et al., 1983b). This isolate was originally selected from among a number of filamentous species for its activity in uptake, and is here applied to a general study of heavy metal biosorption. Precultures were prepared in 500 mL Erlemeyer flasks containing 200 mL potato dextrose solution (PD) on a rotary shaker (150 rpm) for 3 days at 25 °C. Inoculation was carried out by spore suspension at a final density of 108 spores m L - 1. After we found that with 0.5~o Tween 80 in the medium the fungus grows in the form of homogeneously dispersed beads (Galun etal., 1983c), thus optimizing the surface contact with the metal ions it was added (in some experiments, as indicated below) to the medium of the precultures. The fungal mass that developed was then harvested over a nylon filter (20 mesh), washed several times with 100 mL distilled water, slightly squeezed to remove excess of liquid and transferred as desired. Uptake experiments were generally carried out at 25 °C, pH 5.5 (or as specified). The standard charge offungal biomass was 3.0 g (fresh w) or 0.13 g (d.w). Reaction volumes were 20 mL containing an initial concentration of metal-ion (as halide) at 10, 25, or 50 ppm, or as specified. In most experiments, duplicated runs were made, however, in some cases, triplicates were used. Standard deviations did not exceed + 15~o. Special conditions or procedures are described in connection with experiments as appropriate. 3. Results

3.1.

SITE COMPETITION

BETWEEN

H÷

AND METAL IONS

Increasing H + concentration generally suppresses metal-ion uptake, although exceptions exist. The nature of H + inhibition was analyzed graphically using the Lneweaver-Burke test (Boyer et al., 1958; McElroy and Glass, 1954) originally applied to the kinetics of substrate binding to enzyme proteins. Here (Figure l a - d ) reciprocal

REMOVAL OF METAL IONS FROM AQUEOUS

SOLUTIONS

361

(a) 0,07

.t:

2

0.05

0.O4

6 7'

0.03 Z

[

4

6

8

IO x Io -5

,.,T,A..,(~) o.2c

(b)

o.18

2

o.16

o,14

.E _~ 0.12

J o.lo

0.08 =

N

3 0.06

4 5

0.04 002

I

I

i

2

4

6

INITIAL

I

8

I

Io x Io -5

Zn(~) I

Fig. I. Lineweaver-Burke double reciprocal plots of metal-ion uptake rates vs initial metal-ion concentrations over the pH range 2-7. A family of curves varying in slope with H-ion concentrations (i.e., inversely with pH) but of constant intercept value is typical of site competition, in this case between H-ion and metal-ion. Graphs a, b, c, and d are Ni, Zn, Cd, and Pb, respectively.

M.GALUNETAL.

362 0.40

(c)

I

~

I

I

I

2

t o.32 0.36

0.28 -0.24 v

0.20 0.16 0.12 0.08 4 6,5

0.04 0

2

4

6

8

I0 x 10-5

,.,T,A.

0.52

(d)

2

0.281

I

0,24

i

0.20

0.16

4 I

6 I

,~,~,~ ~ ( ~ I )

8 '

iloXi0 5

363

REMOVAL OF METAL IONS FROM AQUEOUS SOLUTIONS

metal ion uptake rates are plotted against reciprocal initial metal ion concentrations at each

experimental pH. If the family of curves generated thereby shares a fixed intercept but increase in slope with increasing H + concentration, then H + and metal ion are presumed to be competing for the same site. The competitive pattern is almost classically displayed by Zn 2+ and Cd 2+, and nearly so by Ni 2 +. The extreme displacement of the pH 2 curve for Pb 2 + relative to those at pH 3 to 5 clearly indicates a non-competitive interaction without appreciable pH dependency above pH 3. The uptake of Cu 2 + (not shown graphically) was virtually pH-insensitive. At 10 ppm initial Cu-ion concentration, uptake increased only 5 ~o from pH 2 to pH 3, and only 2 2 ~ over the entire range, pH 2 to pH 5. Even at the higest concentration, 50 ppm Cu-ion, the rate only increased by 11~o at pH 3 and by 67 ~o at pH 5. 3.2. ACTIVATION OF Penicillium BIOMASS BY CHEMICAL OR THERMAL PRETREATMENT

Mycelium pre-treated with 0.1 N NaOH and subsequently washed free of base to pH 5.5 to 6.0, nevertheless behaved as if it had been incubated at a relatively high pH with respect to uptake of Ni-, Cu-, Zn-, and Cd-ions (Figure 2). The already high too

904

/ Ni

Zn

Cu

Cd

~ t [ ~ ] UNTREATED

8O

Pb

[ ~ ] NaOH PRETREATED

7O

[ ~ ] HCI PRETREATED

60"

[~

NaOH THEN HCI PRETREATED

r~HCl THENNoOHPRETREATED

5O

3O 20 I0 0

Fig. 2.

I

2

$

5

2

4

5

I

2

4

5

I

2

5

4

Effect of alkali and acid pretreatment on subsequent metal-ion uptake at pH 5.5. Vertical bars represent standard deviations of triplicates.

capacity of the fungal mycelium for Pb-ion had little potential for enhancement. In contrast, HC1 pretreatment had no effect on subsequent uptake at pH 5.5 of Ni 2 ÷ but may have improved marginally the uptake of Cu 2 ÷ and Zn 2 ÷. The improvement in Cd 2 + uptake conferred by acid treatment was unexpected. A small loss was noted in Pb 2 + biosorption capacity. Combinations involving acid and base in sequence (treatments 4 and 5) as expected exhibit the effect characteristic of the second reagent. Thus treatments 4 and 3 are similar as are treatments 2 and 5. The only exception was in the capacity for Pb 2 + uptake, where all pretreatment effects were minor.

364

M. G A L U N ET AL.

Reportedly formaldehyde pretreatment can improve biosorption (see 26, for example). I n our experiments Penicillium biomass was pretreated with 5 to 10 ~o formaldehyde but the uptake of the cation was blocked to a substantial degree (Table I). Untreated, the mycelium eventually lost most of its Ni-ion, but formaldehyde-treated ftmgal preparations retain whatever they had taken up. I n contrast, D M S O pretreatment activated uptake and enabled the mycelium to hold the metal even after 24 hr. TABLE I Effects of formaldehyde and dimethylsulfoxidepretreatment on uptake of Ni-ion from 10 ppm solutions at pH 6 Ni-ion remaining in solution (~o) after

Pretreatment

None Formaldehyde ~o 5 10 Dimeth~sulfoxide

5 min

4 hr

24 hr

50

47

75

85 80 38

83 73 20

78 73 17

Mycelial pretreatment at 100 ° C (5 min) increased the ion uptake rate in all cases (Table II). The effect ranged from a 23 to 46 ~ elevation of Ni uptake to a 135 to 227 ~o elevation of Cu uptake. A consistent trend relative to initial metal-ion concentration is not evident. TABLE II Effect of fungal heat treatment on initial rates of metal uptake, eaction volume 20 mL containing 0.14 g dry mycelium,pH 5.5, 25 °C. Initial rates are based on incubation 5 min. Average of duplicates. Cation (all + 2)

100 °C for 5 min

Initial concentration (g mL- 1) 10

25

50

Uptake (g.g- 1. min- 1)a Ni

+

7.2 10.5 (146)b

17.2 22.4 (130)

25.9 31.9 (123)

Cu

+

4.3 10.1 (235)

7.2 20.1 (279)

8.6 28.1 (327)

Zn

+

3.7 7.3 (193)

9,7 14.1 (145)

14.1 20.5 (145)

Cd

+

9.0 14.4 (160)

13.3 18.9 (142)

18.0 34.2 (190)

Pb

+

8.8 12.1 (138)

23.1 28.1 (122)

23.8 45.5 (191)

a Average of duplicates. Mean error 15~o. b Uptake after heat treatment as % unheated.

365

REMOVAL OF METAL IONS FROM AQUEOUS SOLUTIONS

rO,O00

(a)

5,000

Cu

J

f

2,000

//'/A

~

:&

g #

//oA /

I,OOO

,.A//

ONHEATED

i/

215

5OO

510

IO0

TOTAL ION CONCENTRATION (jug.ml -I) Fo,ooo

(b)

5,000

Cd

/

/

/7o

HEATED

0 . ' 7 /

2,000

T

<.

1,ooo

P~ m

5oo

5i

I IO

[ 25

510

I00

TOTAL ION CONCENTRATION (2ug, ml -I)

Fig. 3. Effect of thermal pretreatment (100 °C 5 min) on metal-ion binding. (a) and (b) Cu- and Cd-ions, respectively, are also representative of Ni- and Zn-ions. The binding of Pb-ion is not affected by heat treatment (c). All tests were run at pH 5.5. Average of duplicate runs.

366

M . G A L U N E T AL.

(c)

fO,O00

I

5,000

Pb

I

I

I

/

2,000 UNHEATE/~~ Iu, I,OOO

g m

5o0

200

I

I

I0

5

I

25

I

50

i00

TOTAL ION CONCENTRATION (~g.ml -I )

Pretreatment of the mycelium also increased the total amount of metal ion bound except in the case of Pb (Figure 3a-c). Curves for Cu and Cd are also representative of Ni and Zn; heat treatment displaces the curve relating amount bound to total metal ion without slope alteration. Heat activation is also reflected, of course, in residual metal-ion in solution. The effect of heat activation is also well illustrated by the % metal-ion remaining in solution at equilibrium, especially at lower ion concentrations (Table III). Again, the exception is Pb-ion, which is almost completely removed without pretreatments. 3.3. I0N COMPARISONS On a molar basis the initial rate of uptake of metal ions for 0.35 mM solution in (mol kg- ~ m i n - 2) based on the first 5 min was: Ni 2 +

290,

Cd 2÷

118,

UO2a÷ 175,

Cu 2+

113,

Zn 2+

Pb 2+

112,

150,

yielding the series: N i > UO 2 > Zn > Cd > Cu, P b . The place of Fe is not established.

REMOVAL OF METAL IONS FROM AQUEOUS SOLUTIONS

367

TABLE III Effect of fungal heat treatment on residual metal ion in solution at equilibrium (2.5hr). Reaction volume 2 0 m L containing 0.12g dry mycelium, pH 5.5, 25 °C. Cation (all + 2)

100 ° C for 5 rain

Initial concentration (g m L - 1) 10

25

50

~o Remaining in solution a Ni

+

48 22

50 38

59 50

Cu

+

24 5

38 21

53 45

Zn

+

37 13

38 22

58 50

Cd

+

21 12

58 36

58 42

Pb

+

7 15

10 7

18 18

Average of duplicates. Mean error 15~.

TABLE IV Metal loading of Penicillium biomass in relation to ion parameters Ion charge

Element

Parameters ~ Coordination No.

Radius A

Equilibrium metal content of Penicillium biomass after 2.5 hr in solutions containing initially 0.35 mM

3.5 mM

moles k g - 1 Fe Ni Cu Zn Cd Pb U (as UO 2 + ) b

3+ 2+ 2+ 2+ 2+ 2+ 2+

6 4 4 4 4 4 6

0.63 0.66 0.70 0.68 0.88 1.03 3.40

84 35 39 38 30 32 13

270 250 230 260 100 90 42

a After Fairbridge, 1972; Weast, 1967; Whittaker and Murtos, 1970. b See Galun et al., 1983b.

In contrast, the metal content at equilibrium (loading) when dilute solutions are used f o l l o w s t h e s e r i e s ( T a b l e IV). F e }> C u , Z n , N i > C d , P b > U O 2 .

368

M. GALUNET AL.

When a 10-fold concentrated solution was used, the sequence observed was similarly arrayed, but more distinctly ranked: Fe > Ni, Z n > Cu > Cd > Pb > U O 2 .

4. Discussion 4.1. GENERAL BIOSORBENT FEATURES OF Penicillium MYCELIUM The biosorbent properties of Penicillium reside in the hyphal cell wall. These walls can be characterized as a system o f biopolymer fibers interwoven to from an ion-exchange and coordination surface. This surface is patterned with fixed ionogenic groups and fixed neutral ligand sites derived from basic neutral and acidic glycans, phosphoglycans, and proteins. Given the mix of ionogenic groups, from phosphate and carboxylate to amino and guanidino (pK range 2 to 10), strongly charged centers for both cationic and anionic exchange should be present. As we have shown previously, however, (Galun, E, et al., 1983; Siegel et al., 1983), Penicillium biomass does not bind M o O ] - or C r 2 0 ~- even in an acid medium unless preloaded with a strongly bound cation such as Pb 2 +. Thus the cationic centers presumed to be present in Penicillium m a y be blocked chemically (e.g., acetylated) or architecturally inaccessible. Further evidence that positively charged centers do not contribute significantly to the surface characteristics o f these fungal preparations is provided by comparisons of TABLE V Representative chemical composition of cell walls in Penicillium and related fungi Component Neutral Carbohydrates a Glucose Galactose Mannose Glucuronic acid a chitin Galactosamine a Protein Lipid Phosphate Ash

Content (% dw) 12.0-56.0 trace- 8.5 2.0-11.5 1.5- 5.0 8.0-39.0 8.3-10.8 7.3-"18.0 3.4-11.4 0.1- 1.8 1.8- 5.2

a Includes a- and b-glycans, heteroglycans, proteoglyeans. Wall protein may contain 0.2-3.9 mole % L-arginine, 1, 2-7, 7 mole % L-lysine; traces to 4.8 mole histidine; and traces to 10.3 mole % cysteine. After Aronson (1981).

REMOVAL O F METAL IONS FROM AQUEOUS SOLUTIONS

369

uptake with and without formaldehyde pretreatment. Contrary to previous reports (Beveridge et al., 1978; Strandberg et al., 1981) the uptake of UO~ + ion by Pencillium is completely unaffected by formaldehyde treatment and, as we show here, Ni 2 + uptake is actually suppressed. The uptake of UO 2 + by yeast is highly pH-sensitive, with a pH optimum at about pH 3.5 (Strandberg et al., 1981). In our Penicillium preparation, UO 2 + uptake varied little in rate between pH 2.5 and 9.5. The Cu-ion, like uranyl-ion, varied little in uptake with pH. This has also been reported to be the case for another species of Penicillium (Stokes and Lindsay, 1979). Uptake of the remaining 4 cations, Ni 2 +, Zn 2 +, Cd a +, and Pb 2 + drops markedly between pH 3 and pH 2 and in 3 cases inhibition of uptake by H + is the result of the removal of anionic binding sites by their association with H +. The competitive nature of the equilibria involved is shown in Figure 1 by convergence of uptake curves as the ion concentration rises (toward the origin). In essence, when the metal ion is present at sufficiently high concentration, biosorption proceeds even in the presence of excess H +. This obviously does not apply to Pb-ion, which is about 70~o inhibited at pH 2 at all Pb concentrations. It seems obvious that although a general coulombic mechanism governs metal ion uptake and binding, the specific electronegative centers are not all equivalent. It is also clear that pH-insensititve uptake of UO2-ions and Cu-ions must involve neutral ligands (Tobin et al., 1984). Penicillium preparations possess latent binding sites which can be rendered active by suitable pretreatments. This was observed in the uptake of UO2-ion (Galun et al., 1983a). Chemical activation was achieved by a 5 rain treatment with molar KOH; lesser degrees of activation were achieved by alcohols and by dimethyl sutfoxide. In the present study, with the exception of the Pb-ion, all metal cations tested were taken up at higher rates following alkali treatment. Even largely pH-insensitive Cu-ion was more rapidly biosorbed by preparations by mycelium that had been pretreated with alkali. Acid pretreatment was selective, having no effect on Ni uptake, marginal effects on Cu and Zn, but a definite positive effect on Cd. Latent ion-binding groups could have been HC1 activated by displacement of a blocking cation, Ca 2 + for example, or by hydrolysis of an ester, releasing the dissociable carboxyl group. Alternatively, denaturation of fungal wall protein could have exposed anionic functions. These options are not mutually exclusive. Unlike site-competitive pH effects, which tend to disappear at higher metal-ion concentrations, thermal activation varies comparatively little for each specific ion as a function of concentration. It is obvious that heat, like alkali treatment, exposes latent binding sites, but we assume that it does so via a denaturative change in protein, as in helix (closed) to random coil (open) transition. The evidence is: (a) there is a time lapse between heat treatment and use, hence the thermal effect, like protein denaturation, is essentially irreversible; (b) although the thermal pretreatment described here was 100 °C for 5 min, as little as 2,5 min at 70 °C can activate the mycelium, at heat sensitivity best accounted for by proteins. Thermal activation has the same effects on uptake and loading as would the simple addition of more biosorbent.

370

M. GALUN ET AL.

It would not be surprising, however, to find some degree of bio-selectivity in ion uptake, although thermal pretreatment eliminates all capacity for growth. In contrast, the actual capacity at equilibrium for metal ion (load) uptake by heat killed preparation is Fe 3 + > Ni 2 +, Zn z+ > Cu 2 + > Cd 2 + > Pb 2 + > UO 2 + a decrease in capacity with increasing ionic radius. This behavior contrasts with that reported for other fungi where binding and ion size increase together (Shumate et aL, 1978; Strandberg etal., 1981). 4.2.

BIOGEOCHEMICAL IMPLICATIONS

In concept, bioaccumulation of metals includes representatives of all major groups of organisms (Vinogradov, 1953; Watabe and Wilbur, 1976). Nevertheless, the fungi, although widely recognized for their essential role in bio-recycling or organic residues, have not been given the attention accorded to S and Fe bacteria, reef algae and marine invertebrates in metal accumulation. Evidence both for the natural solubilization by fungi of Au and U and for their experimental accumulation of Cu, Co, Zn, Cd, Sn, and Pb (Korobushkina et al., 1973; Berthelin and Munier-Larny, 1983; Siegel, 1973; Siegel etal., 1983) points to a significant place for fungal organisms in the genesis of localized minerals and metal deposits. The present study, although addressed primarily to biosorption as the basis for water quality control and metals recovery technologies, also has relevance for the study of metal biogeochemistry. Thus, we suggest that the following features of our current work are of especial interest: (a) Uptake of Ni, Zn, Cd, and Pb is pH-sensitive, being highly suppressed at pH 2 whereas uptake of Cu- and UOz-ions is virtually pH-insensitive. (b) Heat-killed mycelium is comparable or superior in metal uptake and binding to living preparations. (c) Metal binding is selective. If the equilibrium load for Fe 3 + = 100, then Ni, Cu, and Zn are 85 to 96, Cd 37, Pb 33 and U (for UO~z+) only 16. In addition, oxyanions of Cr v1 and U VI a r e not taken up unless the mycelium is first doped with a di- or polyvalent cation such as Pb z + These features suggest that soils rich in fungi and fungal remains may have long served as sites for the accumulation and persistence of a wide variety of metals and their compounds.

Acknowledgments This work was carried out with the support of a grant from the U.S.-Israel Binational Science Fotmdation.

REMOVAL OF METAL IONS FROM AQUEOUS SOLUTIONS

371

References Berthelin, J. and Munier-Lamy, C.: 1983, Ecol. Bull. (Stockholm) 35, 395. Beveridge, T. and Murray, R.: 1976, J. Bacteriol. 127, 1502. Beveridge, T. and Murray, R.: 1980, J. Bacteriol. 141, 876. Beveridge, T., Williams, F., and Koval, J.: 1978, Can. J. Microbiol. 24, 1439. Boyer, P., Lardy, H., and Myrback, K. (eds.): 1958, The Enzymes, Vol. I, Kinetics-Thermodynamics, Mechanism, Basic Properties, Academic Press, New York. Corpe, W.: 1975, Dev. Ind. Microbiol. 16, 249. Crist, R., Oberhauser, K., Shank, N., and Ngkuyen, M.: 1981, Environ Sci. Technol. 15, 1212. Fairbridge, R. (ed.): 1972, Encyclopedia of Geochemistry and Environmental Sciences, Vol. IVA. Van Nostrand Reinhold Co., New York, p. 189. Galun, M., Keller, P., Malki, D., Feldstein, H., Galun, E., Siegel, S., and Siegel, B: 1983a, Science 219, 285. Galun, M., Keller, P., Malki, D., Feldstein, H., Galun, E., Siegel, S., and Siegel, B: 1983b, Water, Air, and Soil Pollut. 20, 221. Galun, M., Keller, P., Feldstein, H., Galun, E., Siegel, S., and Siegel, B: 1983c, Water, Air, and Soil Pollut. 20, 277. Galun, M., Keller, P., Malki, D., Feldstein, H., Galun, E., Siegel, S., and Siegel, B: 1984, Water, Air, and Soil Pollut. 21, 411. Galun, E., Keller, P., Lehr, H., Siegel, S., Siegel, B, and Galun, M.: 1983, Bioaccumulation of Heavy Elements by Penicillium digitatum : Parameters Affecting Entrapment in Batch Cultures and Column Systems, Abstract, III Intl. Mycological Cong., Tokyo, 28 August - 30 September, 1983. Heldwein, R., Tromballa, W., and Broda, E.: 1977, Z. Algg. Mikrobiol. 17, 299. Korobushkina, E., Karaviko, G., and Korobushkin, I.: 1973, Ecol. Bull. (Stockholm) 35, 325. McElroy, W. and Glass, W. (eds.): 1954, The Mechanism of Enzyme Action, Johns Hopkins Press, Baltimore. Norris, P. and Kelly, D.: 1979, Dev. Ind. Microbiol. 20, 299. Olafson, R., Abel, K., and Sim, R.: 1979, Biochem. Biphys. Res. Commun. 89, 16. Sakaguchi, T., Horikoshi, T., and Nakajima, A.: 1978, 3". Ferment. TechnoL 58, 561. Shumate, S., II, Strandberg, G., and Parrott, J., Jr.: 1978, Biotechnol. Bioeng. Syrup. 8, 13. Siegel, S.: 1973, Env. BioL Ned. 2, 19. Siegel, S., Siegel, B., and Clark, K.: 1983, Water, Air, and Soil Pollut. 19, 229. Siegel, S., Keller, P., Galun, M., Lehr, H., Siegel, B., and Galun, E.: 1985, Watet, Air, and Soil Pollut. (in press). Stokes, P. and Lindsay, J.: 1979, Mycologia 71, 798. Strandberg, G., Shumate, S., and Parrott, J.: 1981, Appl. and Environ. Microbiol. 41, 237. Tobin, J., Cooper, D., and Neufeld, R.: 1984, Appl. and Environ. MicrobioL 47, 821. Tsezos, M. and Ketler, D.: 1983, Biotechnol. Bioeng. 25, 201. Tsezos, M. and Volesky, B.: 1981, Biotechnol. Bioeng. 23, 583. Tsezos, M. and Volesky, B.: 1982a, Biotechnol. Bioeng. 24, 385. Tsezos, M. and Volesky, B.: 1982b, Biotechnol. Bioeng. 4, 955. Vinogradov, A. 1953, The Elementary Chemical Compositions of Marine Organisms, Sears Foundation for Marine Reseraeh, Yale University Press, New Haven. Watabe, N. and Wilbur, K. (eds.): 1976, The Mechanism of Mineralization in Invertebrates and Plants, University of South Carolina Press. Weast, R. (ed.): 1968, Handbook of Chemistry and Physics, 48th ed., 1967-1968, Chemical Rubber Co. of Cleveland, pp. 143. Whittaker, E. and Murtos, R.: 1970, Geochem. Cosmoschim Acta 34, 945.