The Structure of Porphyrin a, Cryptoporphyrin a and Chlorin ao 359 



two points, namely, (a) the formulation of the side chain in position 2 as 

 — CH(OH)— CO — R, and (b) the distribution of the extra alkyl groups between the 

 positions 2 and 4. The analyses of haemin a would appear to allow the presence of 

 a seventh oxygen atom, assumed by Morrison in the — CH(OH) — CO — R side chain. 

 We have observed a shift of 2 m/< of bands III and IV of porphyrin a on acetylation; 

 on hydrolysis the spectrum of porphyrin a is restored unaltered. Morrison observed 

 a similar shift of 5 m/t in the position of the a-band of the pyridine haemochrome. 

 These observations may indicate that the group formulated by us as a-hydroxyalkyl 

 group is more complicated and bears another oxygen atom. However, the presence 

 of a — CH(OH) — CO — grouping appears unlikely since it is not in harmony with 

 our dehydration of a-hydroxyalkyl to )?-alkylvinyl by /7-toluenesuIphonyl-chloride. 

 We have evidence indicating that the a-hydroxyalkyl group is not a-hydroxyethyl. 

 The ketonylporphyrins obtained by the resorcinol melt of haemins after cautious 

 oxidation of a-hydroxyalkyl to a-ketonyl by chromic acid differed from acetylpor- 

 phyrins by higher HCl-numbers and higher /?/• in chloroform-kerosene. In particular 

 the ketonylporphyrin thus obtained after conversion of formyl to methyl by the 

 Wolff-Kishner reaction was not monoacetyl-deuteroporphyrin. Again, the porphyrin 



R 

 HCOH M 



m/ \cho 



HOHjCx^ 

 P 



'M 



(obtained by a succession of reactions from porphyrin a) which differs from hydrated 

 isochlorocruoroporphyrin only by the replacement of methyl by hydroxymethyl, 

 should have a low HCl-number, if R is CH3. In fact, it has a high HCl-number and 

 still shows the a -> /? conversion typical for porphyrin a. 



There is some evidence that the vinyl group in 4 is also substituted. More important 

 than Warburg's inability to obtain cytodeuteroporphyrin from hydrogenated haemin 

 a in the resorcinol melt, is Piattelli's recent finding (see p. 357, this volume), that no 

 /3-ethyl-/J-methyl-pyrrole-2 : 5-dicarboxylic acid can be obtained by permanganate 

 oxidation of hydrogenated porphyrin a. The fact that the alkyl group substitutes the 

 vinyl group in /J, not in a is shown by the possibility of oxidizing it to formyl. Either 

 the additional carbon atoms are distributed between a-hydroxyalkyl in position 2 and 

 /S-alkylvinyl in position 4, or these two groups are interconnected by an alicyclic 

 polymembered ring. 



As we can now assume that these two groups replace vinyl, not methyl, of proto- 

 porphyrin, a mode of biosynthesis of haem a different from those discussed in the pre- 

 circulated paper must be postulated, i.e. oxidation of the two vinyls to formyl and 

 coupling with a saturated fatty acid whose a-CHj is activated, perhaps by coenzyme 

 A, followed by hydration of one of the two groups; or if an alicyclic ring is formed 

 between 2 and 4, by condensation of the two formyls with a similar compound of a 

 saturated dicarboxylic fatty acid. 



Winfield: I should like to ask Lemberg by which method the molecular weight of his 

 porphyrin was determined. If it was not by Brintzinger's diffusion method, I suggest 

 that this might be used to advantage. For some compounds it is possible to determine 

 the molecular weight with considerable accuracy, and with no necessity for pure 

 preparations. It might well be possible to determine from the molecular weight the 

 number of oxygen atoms present in the porphyrin. 



Lemberg : For the determination of the molecular weight of porphyrin a or haemin a we 

 have relied either on the iron content of the haemin, or on the ratios of the specific 

 extinctions of porphyrin a and its molar extinction as established by copper titration, 

 or on the specific extinction of pyridine haemochrome a and its molar extinction as 

 established on the basis of its iron content. These methods all depend on the purity 

 of the preparation, but give concordant results for the purest preparations which 

 agree with those of Warburg and of Morrison. 

 As Winfield has pointed out, the diffusion method has the advantage of being not 



