glycine, a weak chelator. The behavior of these 

 pH buffers in our media cannot therefore be ex- 

 plained as a pure chelating effect, nor can the two 

 Synura media (Table 2) . If the miUiequivalents of 

 all the metals, including Ca and Mg, are com- 

 pared with the miUiequivalents of the chelators, 

 we find that the EDTA-glutamate medium is over- 

 chelated 1.3:1 and the histidine medium 5:1; 

 yet they give similar growth . The media for Oxyr- 

 rhis marina of Droop (1959a) present a similar puz- 

 zle: at the same pH and with the same amount of 

 trace metals and major elements similar growth can 

 be obtained by chelating with 0.6 mg .% EDTA or by 

 the joint chelation of 20 mg.% histidine, 4 mg .% 

 citric acid and 50 mg.% glycylglycine; 2-6 mg.% 

 EDTA on the contrary allows far less growth. Again 

 the level of chelation does not explain in the data: 

 hisUdine alone chelates > 60:1; 0.6 mg.% EDTA, 

 2:1 and 6 mg.% EDTA, 20:1. 



These discrepancies can be explained satis- 

 factorily if one considers the consequences of the 

 different molecular size of the chelators used. As 

 mentioned, EDTA was chosen because it was sup- 

 posed that the bulkiness of its molecule would pre- 

 vent penetration into the cells and that it would be 

 photo-stable. It was later found that Fe-EDTA 

 chelates decompose in light (Jones and Long, 195 3) 

 and that some EDTA or its breakdown products pen- 

 etrate in the algal cell (Krauss and Specht, 1958) . 

 However, the majority of the iron apparently does 

 not enter as intact iron chelate because, on a mol- 

 ecular basis, 15 to 50 times more iron was ab- 

 sorbed by the cells than EDTA. Tiffin and Brown 

 (1959) employing the iron chelate of ethylenedia- 

 mine di (o-hydroxyphenylacetic acid) (EDDHA) 

 found that roots of decapitated sunflower plants ab- 

 sorbed only about .3% of the total EDDHA and 

 large amounts of iron, leaving most of the EDDHA 

 in the nutrient solution. 



Therefore for practical purposes EDTA is a 

 non-absorbable chelator and the cells depend al- 

 most exclusively: a) on the available free metal 

 ions which are very low because of the high sta- 

 bility constants of EDTA chelates, though in the 

 case of iron more free ions may be made available 

 by the partial disintegration of EDTA in light and 

 b) on the ability of the organisms to compete for 

 the metals in the EDTA chelates . This transfer is 

 uphill since EDDHA and EDTA accumulate in the 

 medium. The data in fact show that the algae be- 

 have as if they depend mostly on free ions present 

 in the medium because any conditions, like varia- 

 tions in pH, over- and under-chelation, which up- 

 set the ratio metal chelates: free metal ions favor- 

 able for an organism and a given pH, result in tox- 

 icities or deficiencies which inhibit or suppress 

 growth . 



Histidine and other chelating small molecules 

 are readily absorbed. Since in this case the free 

 chelator, the metal chelates, and the free metals 



presumably all can be absorbed, the effect of over- 

 and under-chelation and pH changes should, and 

 do, affect far less the availability of the trace 

 metals . The transfer and the competition for trace 

 metals by the different biological internal chela- 

 tors now can proceed freely in the interior of the 

 cells. Furthermore the absorbable chelators when 

 employed in large quantities as pH buffers have 

 the power to smooth out, perhaps by mass action, 

 the inflexibilities caused by the presence in the 

 media of weak , slightly over-chelated EDTA-trace 

 metal mixtures . 



This way to supply metals by employing pen- 

 etrable chelators parallels what may happen fre- 

 quently in nature. Lichens and other plants grow- 

 ing on rocks must be able to secrete organic com- 

 pounds which dissolve and perhaps chelate the 

 mineral elements. Various fungi and bacteria pro- 

 duce and release in their media extremely strong 

 chelating substances which are specific for iron 

 such as: coprogen, produced by bacteria, actino- 

 mycetes and fungi, (Hesseltine et al . , 195 3), 

 "terregens factor", produced by Arthrobacter 

 pascens (Lochhead and Burton, 195 3) and ferri- 

 chrome, produced by Ustilago sphaerogena (Nei- 

 lands, 1952). Ferrichrome, amazingly, has a sta- 

 bility constant ten times higher than EDTA (Nei- 

 lands, 1957) for ferric iron yet is an effective way 

 to supply iron to tomato plants grown hydroponi- 

 cally . Arthrobacter terregens and other soil bac- 

 teria (Burton, 1957), Microbacterlum sp . (Demain 

 and Hendlin, 1959), and the fungus Pilobolus 

 kleinii (Hesseltine et al . , 1953) have a growth 

 factor requirement which is satisfied equally well 

 by terregens factor, coprogen and ferrichrome. 

 These substances, though apparently different 

 chemically, provide an extremely effective way of 

 supplying iron. Because of their special biological 

 activities at very low concentrations (ug./ml.) 

 they are considered by Demain and Hendlin (1959) 

 as "iron-transport factors" . Nielands (1957), in 

 a thoughtful review, postulated that they may act 

 as coenzymes for the intracellular transfer of iron. 

 Only a few of the great variety of molecules 

 tried, many of which are known chelators, can re- 

 place them. Outstanding are the compounds formed 

 upon heating sugars with amino acids , the ketose- 

 amino acids. Glucosyl-glycine is active for 

 Microbacterlum sp . , fructose-phenylalanine and 

 the products derived from autoclaving glucose and 

 glutamic acid are active for Micrococcus lysodeik - 

 ticus ■ Other bacteria , like Lacto-bacillus gayoni 

 and Proprionibacterium freundenreichii require glu- 

 cosyl-glycine , suggesting that they may also need 

 "iron transport factors". It is interesting to note 

 that the fructose- amino acids stimulate haem syn- 

 thesis and amino acid incorporation into globin. 

 (Kruh and Borsook, 1955) . Other compounds like 

 citric acid and 8-hydroxyquinoline are active for 

 M. Ivsodeikticus and aspergillic acid for Micro - -^ 



93 



/ 



* 



