Changeux. J.-P.. Benedetti, I... Bourgeois. J.-P., Brisson, A., Cartaud, J . . Devaux. P.. Cirunhagen, 14.. Moreau, M . , Popot. . J . . Sobel, A , ? & Weber. M . (1976) Cold Spring Harbor Sjstiip. Quaiit. BIo1. 40, .? 1 1 Clark, I>. G . %Macmurchie. D. n..Flliot, E.. Wolcott. R . G , . 1-andcl. A . M..& Rafter). M A (1972) BiochPmi.Ttr1' I I . 1595. Cohcn. J . , Weber. M.. tiuchet. L1.. & Changeux, .I. P. (1972) FEBS L.ett. 26, 43. Colquhoun. D., & Rang. H . P. ( 1976) M o l . Pharmarol. 12, 5 19- 535. Duguid. .1. R., & Raftery, V . A . (1973) Binc~hrti7istrj,I.?. 3693. Grunhagen. H.H.. & Changeiix. 1. P. ( 1 9 7 6 ) .I Mo1 Riol. 106. 5 1 7 .
Jiinin? .I.(1973) f r o g . Bioph~,.c, i f d Biol. 27, 77 120. K a t 7 , H.. & Thesleff, S. (1957) J . fhj,srol. 138, 6 3 . Lee, T., Witzemann, V., Schimerlik. M . . Rr Raftery, $1. A . ( 1 977) Arch. Biorheni. Biophj's. 183, 57 63. Lester, H A . (1972) Mol. Pharnincol. 8 , 6 3 2 644. Mooser, (3.. Schulman, H., & Sigman, D. (1972) Biochenii.ytrJ, 1 1 , 1595. Popot. J . I_., Sugiyama, H.. & Changeux Biol. 106, 469 -483. Quast. L. Schimerlik, M., & Rafter). M. A . (1978) Biocheni. Bioph,v.c.R e s Coniniitri. (in press). Raftcry. M. 4.. Vandlen, R . l...Reed. K . I .. & I.ee. T. (1975)
Cold Spring Harbor S j , i n p Qii(inr. Biol. 40, 19.1. Rang. H . P.. Pr Ritter. . I . M . ( 1 969) I f o l . f h u r n i m d . 5 . 394. Rang. H . P.. & Ritter. .i.M (1070a) .2fo1.Phnrt7iotol 6, 157. Rang, H P.. & Ritter. .I M . (1970b) M I ) / . fhcrrniwoi. 6. 3x3.
Reed. K., Vandlen, R., Bode, 1 . . Duguid. .I., & Rafter). M. 4 (197.5) Arch. Riorhrrii Biophr,r 167, l i x Schiinerlik. M . I.. Rr Rafter). I f . 4 (1976) Hiochrni. Bioph\,.s. Res. Coniniirri. 73. 607 Schmidt. I . . Rr Rirtter!. M 4 (1973) 4ntr1. RiochPn/. 52, 349. Sheridan, R.. & Ixstcr. 11. ,I (1975) Pronie were treated with trypsin. less than 10% of the label wCisrclcascd into the supernatant, although 50-55*/0 of the protein \ \ a s released, but most of the remaining membrane-bound peptide!, were not labeled. On gel electrophoresis of the digest, all the membrane-bound radioactivit) was found in a broad band at approximately 12 000 daltons, implying that selective labeling of only a portion of the largely hydrophobic segment5 occurred. On the basis of the present work. we should expect that surface residues of nucleophilic character have bcen attacked from the membrane interior by the lipophilic nitrene. Eivcn photolabile groups bound to amphipathic molecules such as lipids (Chakr;tbarti &L Khorana. 1975) niay present similar difficulties. fiere. the photolabile substituent niay loop back to thc niembrnne surface ;is has been suggested for
VOL.
spin-labeled lipids (Cadenhead et al., 1975), though this need not be a permanent configuration. The lipophilic photoactivated cross-linking reagents of Mikkelson and Wallach (1 976) and of Huang and Richards ( 1 977) may also need reevaluation with the problem of nitrene electrophilicity in mind. While this work was in progress, Abu-Salah and Findlay ( 1 977) reported t h e labeling of egg-yolk lecithin by phenylnitrene. We have been unable to confirm the report of these workers that a labeling yield of around 12% can be achieved with this system. Indeed, our results suggest why the estimate of Abu-Salah and Findlay (1977) may be erroneously high. First, if prephotolyzed phenyl azide is mixed with vesicles, we have shown that it is impossible to remove more than approximately 90% of t h e radioactivity by exhaustive dialysis. (This is in contrast to the effective removal of all unphotolyzed phenyl azide by this method.) This means that the residual radioactivity after dialysis of photolabeled vesicles (which those authors used as a measure of labeling efficiency) is largely a measure of noncocalentlj, bound photolysis products of phenyl azide. Only by thin-layer chromatography of the whole lipid can these products be separated from bona fide labeled lipid (see Figure I ) . Secondly, we have shown (Figure 3) that a large fraction of the radioactivity that cochromatographs with the fatty acid methyl esters (after lipid transesterification) derives from the photolysis products of phenyl azide that are seen even in the absence of lipid. Since Abu-Salah and Findlay (1977) chromatographed the methyl esters but not the lipids themselves, their estimate of the amount of label attached to fatty acid was too high. [In control experiments, we have shown that use of the unusual Tris-lysine buffer employed by Abu-Salah and Findlay (1977) does not increase the labeling yield.] Thirdly, these authors suggest that some 2% of label is incorporated into the head groups of the phospholipids. Yet the fact that
