How to measure and predict the molar absorption coefficient of a protein

8/22/2019 How to measure and predict the molar absorption coefficient of a protein

1/13

Prorein Science (1995), 4:2411-2423. Cambridge University Press. Printed in the USA.Copyright 0 995 The Protein Society

How to measure and predict the m olar ab sorptioncoefficient of a protein

C . NICK FELIXAJDOS,'ANETTEEE,'ERALDRIMSLEY,'AND THERONICA GRAY'' Department of Medical Biochemistry and Genetics, 'Dep artm ent of Biochemistry and Biophysics, andCenter fo r Macromolecular Design, Texas A& M University, College Station, Texas 77843-1 I14(RECEIVED biy 12, 1995; ACCEPTED eptember 8, 1995)

AbstractThe m ola r absorp t ion coe f f i c ien t , E ,f a protein is usually based on conce ntration s measu red by dry weight, ni-t rogen, or amino acid analys is . The s tudies reported here suggest that the Edelhoch m ethod is the best methodfor measuring E for a protein. (This method is described by Gill and von Hippel [1989, Anal Biochem 182:319-3261 and is based on data from Edelh och [1967, Biochemistry 6:1948-19541.) The absorb ance of a pro tein at280 nm depend s on the c ontent of T rp, Ty r, and ys tine (disulf ide bonds). The average E values for these chro -mop hores in a samp le of 18 well-characterized proteins have been estim ated , and the values in wat er, pro pan ol,6 M guanidine hydrochloride (G dnHCI), and M urea have been measured. For Trp, the average values for theprotein s a re less than the E values measured in any of the solvents. For Tyr, the average values for the pr oteinsare intermediate between those measured in 6 M Gdn HCl an d those m easured in propanol . Based on a sampleof 116 measured t values for 80 proteins, the t at 280 nm of a folded protein in water, t(280), can best be pre-dicted with this equation:

~ ( 2 8 0 )M" cm") =(#Trp)(5,500) +(#Tyr)(1,490) +(#cystine)(l25).These ~( 28 0) alues are qui te re liable for prote ins containing Trp res idues, and less re liable for prote ins that donot . However, the Edelhoch methods convenient and accurate , and the est appro ach is to measure ra ther thanpredict E .Keywords: molar absorpt ion coeffic ient ; molar ext inct ion coeffic ient ; near U V absorbance; t ryptophan absor-bance; tyrosine absorbance

Biochemists must fre quen tly determ ine the co ncentra tion f aprotein solut ion. This s most often don e y measuring the ab-sorbance, A , near 280 nm and using the Beer-Lamb ert law:

A =IC (1)where E is the molar absorptioncoefficient (M " cm-I) , I is thepathlength (cm), and C is the p rote in concentra t ion (M). Thisis an excellent meth od for measuring protein concentrations pro-vided that an accurate value of 6 is avai lable . The goal of thestudies described here was to answ er the following questions.First, wha t is the best exp erimental method for determiningE for a prote in? Second, can be predicted accurate ly from theamino acid composi t ion of a prote in?

Reprint requests to : C. Nick Pace, Departmentof Medical Biochem-istry and Genetics, TexasA&M University, College Station,exas 77843-1 1 14; e-mail: [email protected].

To determine E requires an accurate measurementof A a n dC (Equ at ion 1). The measurem ent of A is straightforward(Schmid, 1989; Mach et al. , 1995), but the measurement of Cis not . The four techniques most often used to measure C are:am ino acid nalysis (Benson et al., 1975), Kjeldahl nitrogen de-terminat ion (Jaenicke, 1974), the dry weight method (Hun ter ,1966; Kupke &Dorr ier, 1978; Nozaki, 1986), and the Edelh ochmethod (Edelhoch, 1967; Gill & von Hipp el, 1989). Table 7shows t values determined for r ibon uclease T1over the years.It is clear that these techniques do no t always lead t o t valuesin good ag reement. The same is true f or rib onuc leaseA (Wet -laufer, 1962) and other prote ins (Gill & von Hippel, 1989). Inthis paper, we report t values determined for RNaseT1 and s ixmutants , and for even other prote ins by the dryweight and/orEdelhoch method s . We conclude that the Edelhoch method isthe simplest and most reliable experimental method for deter-mining t .The absorbanceof a prote in solut ion above 75 nm dependson jus t th ree chrom ophores : the s ide cha ins o f T rp , Tyr , and

24 1 1


2/13

2412 C.N. Pacetl.cystine (Wetla ufer, 1962). (Thro ugho ut this paper, cystine con-tent is identical to disulfide bond content. In con trast to cystineresidues, cysteine residues do not contributeignificantly to theabsorb ance abov e 275 nm [Bailey, 19681.) Th e E values for theprotein chromophores depend n their environment (Yanari &Bovey, 1960; Bran dts &Kaplan, 1973). O u r main interest is inpredicting the E values for globu lar prote ins . In globular pro-teins, the three contributing chromophores aremost often bur-ied in the interior of the prote in: 87% buried for Trp, 76%buried for Tyr, and 92% buried for cyst ine (Lesser & Rose,1990). Consequently, i f we knew the solvent that best repre-sented the interiorof a folded protein, then we could measuret values for the Trp, Tyr, andystine chrom opho res in th at sol-vent an d might be ab le to use these values to predict E values forproteins with acceptable accuracy. W etlaufer (1962) an d Perkins(1986) used E values for model chromophoresdetermined in wa-ter and showed that th e E values for protein s could e predictedreasonably well. More recently,Gill and von Hipp el (1989)showed that t values for the model chromop hores determinedin 6 M GdnHC l can also e used to predict E values for p roteinsreasonably w ell. Mo st recently, Mach et al. (1992) used a sta-tistical method to estimate thebest average E values at 280 nmfor the three chromophoresn a sample of 32 proteins an d thesefurther improved the fit between mea sured and predicted t val-ues for their sample of p roteins. In this p aper, we report E val-ues for Trp, Tyr, and cys t ine models measured in water , 8 Murea, 6 M GdnH C1, and propanol . In addi t ion, we have ana-lyzed the absor ption spectra of a sample of 18 globular pr otein sto es t imate the average t values for Trp, Tyr, and cystine be-tween 272 and 286 nm. Fina lly, we have analyzed a sample of116 measured E values for 80 different proteins to estimate thebest E values for Trp , Tyr , and cystine to predict t values at280 nm for proteins.

ResultsProtein molar absorption co efficients determinedby the dry weight and Edelhoch m ethodsMolar absorpt ion coeffic ients for wild-type RNase TI and s ixmutants determined in buffer and in 6 M Gdn HC l by the dryweight meth od and in buffer by the Edelhoch method are istedin Table 1 . For the RNase T1 values in the second row and foral l of the mutants, the same absorbance easurements were usedfor both methods . The t values in Gd nH Cl a ver age 4.1070 lessthan he values n buffer . In general , X shif ts o shorterwavelengths and t decreases when protein s aredissolved in 6 MGd nH Cl or 8 M urea (Sela et al. , 1957; Lee & Timasheff, 1974;Prak ash et al. , 1981; Noz aki, 1986). Fo r RNase TI in buffer,th e t values from the Edelhoch method arelways higher tha nt values from the dryweight method. For wild-type RNase T I ,the difference s abou t 5 % , and on ly fo r Asp 49rp is the dif-ference greater tha n 10%. Some of the factors thatmay contrib-ute to the ifference are considered in the Discussion. As a checkon our dry e ight method, E values for RN ase A an d lysozymewere determine d an d the agreem ent with literature values wasreasonably good (Table 2 ) .The results obtained over a period of years and summarizedin Table 1 suggested that the Edelhoch method might be morereliable than the dryweight meth od for determining t for a pro-tein. To test this further, we measured E values with the Ede l-

Table 1. Molar absorption coefficients at 278 n m fo r RNaseTI and six muta nts determined b y the dry weight andEdelhoch methodsa" . ~

~ ~_ _ _ _ _ _ _ _ _ ~ ~ ~ ~ ~ ~Dry weight Edelhoc h

Protein~

6 M GdnHClufferuffer~ ~

"~_ _ _ ~~~ ~ ~ ~ .~ ~RNase T1RNase TIA sp 49 A laA sp 49 P heAsp 49 TyrA sp 49 T r pT r p 59 TyrT r p 59 P h eRNase TI averageh

n.d.17,420 f 59018,03017,470 f 59018,470 ? 49 021,700 f 37013,62012,000 ? 17017,640 f 260

18,520 f 33018,220 f 39018,84018,310 f 6019,300 rf- 18022,050 f 21014,08012,590 f 19018,470 f 210

19,160 f 27019,300 f 1 0 019,200 f 1,13020,300 i 5024,500 t 16013,800 i 18019,215 f 40

19,200

14,700

~ ~~ ~ ~

~""" ~

~ ~ ....... ~ ~ ~Buffer =30 m M M O P S , p H 7 . 6 M G d n H C l =6 M guanidine hy-drochloride, 30 m M M O P S , p H 7. t value =mean ? mean deviation.For the f irst entry, the dry weight results based on three indepen dentmeasurements and was reported in Hu et al . (1992), and the Edelhochresult is based on three independent measurements. Fo r the other en-tries, the number of independent measurementswas three for wild-typeRNase T1 , one for D49A, two for D49F, two for D49Y, three for D49W.one fo r W59Y, and two fo r W59F.value =mean i mean deviation of the first four entries. D49Aan d D49F were included with wild type because they have the same con-tent of Trp, Tyr, and cystine andhey must be folded because hey havegreater enzyme activity rhan the wild-type enzyme.

hoch metho d for even well-characterized protein s. The resultsare summarized n Table 2. The E values based on the Edelhochme thod are in excellent agreement with results from the litera-ture determined by other methods . The average deviat ion isk2.0070. The largest deviations are for RNase A and 0-lactoglob-

Table 2. Molar absorption coefficients f o r several proteinsdetermined by the dry weight and Ed elhoth m ethods"~" ~~ ~

. ~ ~- ~ .. .ProteinRNase TILysozymeRNase ABS A0-lactCTgenBPTIInsulin

.. ~h

27828 127827827 828 2276276

D ry weightdelhoch18,470 f 210 19,215 ? 4037,070 rf- 450 38,010 2 809,490 k 160 9,460 f 18044,600 t_ 2016,550 f 8050,380 f 1,080

5,740 f 106,020 f 2

. .."" ~ ~ ~ Literature19,330 i 1,04037,860 i 4509,880 f 150

43,820 i 53 017,460 ? 44 051,600 f 880

5,700 rf- 1006,010 k 23 0~ ~. . ~ ~ ~ ~-~

t value =mean f mean deviation. For RNase TI, the dry weightand Edelhoch values are from Table 1 and the l i terature value is fromTable 7. For lysozyme, the t values are based on three dry weight de-terminations, two Edelhoch determinations, and five l i terature values(2.65 & 0.03 m L mg-l cm"). For RNase A, the e values are based onthree dry weights, f ive Edelhoch determinations, andix literature val-ues (0.722 f 0.011 mL mg" cm-I). Fo r the other proteins, the valuesare based on two Ed elhoch determin ations, and the following values(mL mg-' cm-I) from the l i terature: BSA, 0.660 f 0.008 fo r 10 t val-ues; &lactoglobulin, 0.951 f 0.024 fo r 5 t values; chymotrypsinogen,2.01 f 0.04 fo r 3 t values; BPTI, 0.867 k 0.012 for 7 c values; insulin,1.04 f 0.04 for 7 t values.


3/13

Protein absorption coefficientsulin. Fo r RNase A, E =9,800 M" cm" at 277-278 nm is theE value used most often over the years. Both our dryweight an dEdelhoch values are abou t4% less than this. Ou r dry weight re-sults fo r RN ase A yield E =9,870 M" cm" if we do not cor-rect the absorba nces for ight scattering. One possibility is tha tmany of the literatu re values for R Nase A were not correctedfor light scattering andmight be toohigh. For /3-lactoglobulin,th e E value in comm on use is 17,550 M" cm" at 278 nm (Town-end et al . , 1960). How ever, we determine d a value of 16,650M" cm" i f the absorb ance values were corrected for lightscat tering (Pace, 1966; Cup 0 & Pace , 1983), in excellent agree -ment with our result from the Edelhoch method.Model compound molar absorption coefficientsAbsorption spectra for N-Ac-Trp-OEt andN-Ac-Tyr-OEt weremeasured in the following solvents: water, 8 M urea, 6 MdnHC1,I -p ropanol , d ioxane , and fo rm am ide . Mola r absorp t ion coe f -ficients calculated for the first four solvents at fou r wavelengthsare given in Table 3. The resul ts for dioxane and formamidewere similar to the results for propano l. For Trp and Tyr , thevalues in 8 M urea and 6 M Gd nH Cl are essentially identical,and differ only lightly from th e values in water. For Trp, X,

2413is about 2 nm higher and E, about 9% grea te r in p ropanolthan in water , and for Tyr, is abou t 3 nm higher andabout 25% greater in propanol than in water . The E values fo roxidized glu tathione in 6 M G dnH Cl are s imilar to the valuesfor cystine reported by Edelhoch (1967). Note that the valuesfor oxidized gluta thio ne are abou t 2% f the t values for Trp,a n d a b o u t 9 %of the E values for Tyr. We also determined theabsorpt ion spectrum for N-Ac-Phe-OEtn water and found =188 M" cm-I at X =257.6 nm , E =3.7 M" cm" at 275 nm ,an d E is less at longer wavelengths. Thus, the contribution of Pheto the absorpt ion of prote ins above 27 5 nm is negligible.Average t values between 272 and 286 n m f o rTrp,Tyr, and cystine based on the absorption spectraof 18 globular proteinsAbsorption spectra from 350 to 250 nm were measured for 18proteins: the first 1 1 in Table 5 plus trypsin, thermolysin, staph-ylococcalnuclease (SN), bovine pancreatic rypsin nhibitor(BPTI), insul in, RNase TI, and barnase. The p rotein concen -tra t ion for each protein was based on th e av erageof the mostreliable t values from the literature a nd s given in the Meth odssection. The E values were calculated from 272 to 286 nm at 2-nm

Table 3. Molar absorption coefficients of Trp, Tyr, and cystine model compounds in water, 8 M urea,6 M GdnHCI, and propanola -~~~~~~ ~ ~ ~

Wavelength~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Water 8 M urea 6 M G d n H C lropano l~ ~ ~

~

Tryptophan models '~ ~ ~ ~ ~

282 5,525,580 5,635 6,125280 5,630,635 5,685 6,075278 5,590 5,535 5 ,565276 5,7805,455 5,385 5,405,680

Tyrosine models '282 1,145 1,240 1,220 1,560280 1,215 1,300 1,285278 1,6801,295 1,390 1,395 1,735276 1,390 1,450 1,455 1.660

Cystine modelsd282 I O 0 105 I IS280 I10 115 I25 135278 125 130 140276 I50145 I50 I60 I65

120

- ~ -~ ~ . ~ ~~ ~"" T h e u n i ts fo r all values are M" cm".'Th e values given for water, 8 M urea, and propanol are the average of at least two measurements using N-Ac-Trp-OEt. The values given for6 M GdnHCl a re the average of ou r values and Edelhoch's values obtained with N-Ac-Trp-NH2 and given in Gill and von Hippel (1989).Th e values given for water, 8 M urea , and p ropano l a re the average f at least two measurements using N-Ac-Tyr-OEt. The values given fo r6 M GdnHCl a re the average of our values and Edelhoch's values for Gly-Tyr-Gly given in Gill and von Hippel (1989).Th e values given fo r cystine in water, 8 M urea , and 6 M GdnHCl a re the average of at least two measurements using oxidized glutathione.Th e value given for cystine in propan ol is the average of ou r values with cystine in propanol and Bailey's (1968) values with cystine in ethanol.

N-Ac-Trp-OMewater): 5,600 (279.8 nm )Bailey, 1968)N-Ac-Trp-NH2 (6 M G d n H C I ) : 5,690 (280.8 nm) (Edelhoch, 1967)N-Ac-Trp-O Et propanol): 6,200 282.5 nm) (Solli & Herskovits , 1973)N-Ac-Tyr-OEtwater): 1,390274.6 nm )Brand ts & Kaplan, 1973)N-Ac-Tyr-OMewater): 1,420 (274.6 nm )Bailey, 1968)N-Ac-Tyr-OEt (6 M G d n H C l ) : 1,450 (275.3 nm) (Brand ts & Kaplan, 1973)N-Ac-Tyr-OEt (6 M GdnHCl) : 1,500 (275.5 nrn) (Edelhoch, 1967).N-Ac-Ty r-OEt propanol): 1,710 (277.8 nm) Brandts & Kaplan, 1973)N-Ac-Ty r-OEt propanol): 1,710278.3 nm ) (Solli & Herskovits , 1973)

~ .. ~ ~~ ~

e Some li terature values for tmux nd X for comparison (see Nozaki, 1990, fo r other values):


4/13

2414Table 4. Molar absorption coefficients fo r Trp, Tyr,and cystine based on an analysis of the absorptionspectra of 18proteinsaWavelength TrPyr

Cystine not fixed272 5,084 f 170 1,405 f 84 20 f 108274 5,248 k 130 1,490 f 63 65 f 82276 5,286 f 107 1,563 k 5204 f 67278 5,315 & 100 1,557 f 48 170 f 61280 5,520 f 106 1,457 f 50282284 5,473 f 201 1,234 f 98 232 k 126286 5,026 f 217 986 & 135 354 k 174

Cystine fixed

~ ..".

~

173 k 655,603 f 130 1,362 k 63 182 f 82

272 5,095 f 238 1,345 k 7795274 5,253 f 176 1,452 k 57 I75276 5,289 f 134 1,545 f 44 I60278 5,313 f 120 1,568 f 39 I40280 5,517 f 128 1,474 i 125282 5,599 k 164 1,387 f 54 I10284 5,466 f 262 1,281 i 86 95286 5,013 f 346 1,078 f 126 85

~ ~ ~ _ _ _ _ ~ ~ ~ ~ ~ _ _ _ _ _ _ ~ ~_ _ _ _ _ _ _ _ _ ~ ~

a These values in M" cm" are based on an analysisof the measuredabsorption spectraof the first 11proteins in Table 5 plus thosefor tryp-sin, thermolysin, staphylococcal nuclease, BPTI, insulin, RNase T1, andbarnase, as described in the text.

C.N. Pace et al.intervals, and the results were used to estima te the verage E val-ues for Trp , T yr, and ys t ine given in Table 4. The (X) valuefor a protein can be calculated with

( A ) = [#Trpl t , (Trp) +[#TyrIt,(Tyr)+[#cyst ine] ,(cystine) (2)

where #Trp, #Tyr, and cys t ine are the numbersof residues ofeach type n the protein (see Table 5 ) , and Eh(Trp, Tyr, or cystine)are the aver age values for thes e residues in the same pro tein.We have used the nonlinear leas t-squares method of John sonand Fras ier (1985) to f ind thevalues of t , (Trp, Tyr, or cystine)that minimize the sum of the squares of the deviations of the ( A ) values calculated with Equ ation 2 from the measured (A )values. The results of this analys is aregiven in Table 4. In th etop panel , the t , (Trp, Tyr, andcystine) v alues were all allowedto vary . No te tha t the values for cystine increase with increas-ing wavelength. This is not reasona ble, and proba blyesults be-cause cystine makes such a small con tribu tion to t ( h ) tha t i tcann ot be estimated reliably. Consequently, we also did the anal-ysis with th( cystin e) fixed at t he values determ ined for oxidizedglutathione in 6 M GdnHCl (Tab le 3) , and only the t,(Trp) andt,(Tyr) values were allowed to vary. These results are given inthe lower panel of Table 4. We also showed that changing thet value for a given pro tein by k 10% has only a small effect o nthe t values for Trp and yr derived from theanalysis. Note tha t

Table 5 . Observed and predicted molar absorption coefficients at 280 n m f o r 80 proteins"~ ~

~ ~

N o.ro te in1 C AH - C )2 C P A3 CTgen4 a - L a c t6 Lysozyme7 Papa in8 RNase9 BS A

10 H SA1 1 3KI12 Y14F13 Y55F14 Y88F15 Y55.88F16 Y14.88F17 Y14,55F18 Aldolase19 A lc Hase(Y)20 C P A21 C P B22 CTgen23 G A P H a s e24 G lu Hase25 Insulin26 La cep27 a-Lac t28 P-Lgb29 Lysozyme

~ ~

~ ~ ~

5 P-Lgb

~~

M W29,11834,40925,66614,18318,30014,31423,42613,69066,40066,45013,40013,38313,38313,38313,36713,36713,36739,00036,70034,41434,61725,67035,60655,7555,734

34,61214,18618,28514.314

~

~ ~ ~ ~ ~~ ~

TrP TYr Cystine~ ~ ~

A (280,1070)~ ~ ~ ~ ~ ~

7 0 17.707 19 1 18.908 5 19.704 40.902 2.636 46.305 19 34.600 6 4.8920 17 6.611 18 17 5.300 3 0 3.360 2 0 2.140 2 0 2.310 2 0 2.480 1 0 I .300 1 0 I .230 1 0 0.983 12 0 9. I O5 14 0 13.107 19 1 17.908 22 3 21 .oo8 50.003 1 1 0 8.904 18 0 9.200 4 3.612 0.704 40.302 2.606 46.40

~ ~

t (obs)5 1,53965,03350,56229,64217,62337,64657,6289,432

43,89035,2194,5022,8643,0913,3191,7381,6441,310

35,49048,07761,60172,6965 1,3403 1,6895 1,2955,51023,19028,79817,55437,789

~ ~

~ ~ ~ ~

~ ~~ ~ ~

4 p r e d )50,42066,93550,58528,46017,21031,97056,1859,440

42,92534,4454,4702,9802,9802,9801,4901,4901,490

34,38048,36066,93577,15550,58532,89048,820

6,33522,92028,46017,21037,970

070De v2.17

~ ~

-2.92-0.05

3.992.34

-0.862.50

-0.082.202.200.72

-4.053.61

10.2114.259.38-13.743.13-0.59

-8.66-6. I31.47-3.79

4.82- 4.961.161.171.96-0.48

(continued)


5/13

Protein abso rption coefficients--~

Table 5. ContinuedN o.rotein30 T4 lysozyme31 Ovalbumin32 Papain33 RNase34SA35 HSA36 Tgen37 D-a.a.xidase38 AAT39 CAH - B )40 CA (B-B)41 A-crystallin42 B-crystallin43 11-crystallin44yokinase45 Thioredoxin46ransferrin47 Trypsin48 Lysozyme49 RNase50 a - C T51 Concan52 SN53 Insulin54 Papain55 Alc DH ase(H )56 Elastase57 PG Kinase58 RNase TI59 Barnase60 ApoMb61 Lysozyme62 R-lysozyme63Nase A64-RNase65 SN66 poCyt c67 Barnase68 Barnase6935 F7071 F71 W94F72 W94L73 RNase174Nase175 D49A76 D49F7749Y78 D49W79 W59Y8059 F81 T4 lysozyme82 W138Y83126,138,158Y84ysozyme85 RNase86 BSA87 &Lgb88 CTgen89 BPTl90 lnsulin91 Ovalbumin

~ - ~ _ _M W

18,70045,90023,42613,69066,29666,47023,99839,33644,93329,11528,98019,79020,06321,00321,63811,67575,18123,99314,30613,69025,18525,57216,8005,73423,42639,79425,89644,570I 1,08912,38317,20114,31414,31413,69013,69016,78611,70212,38312,38312,34412,36012,34412,3101 1,0891 1,0891 1,04511,12111,13711,16011,0661 1,05018,63618,61318,56714,31413,69066,40018,30025,6666,5 185,78245,900

~

T rP3350214

10767I24028460841052721326600113322221111120032060228003

Tyrystine6

10196

20181014158862

1572

26I O364774

19411

79733366747778779999

I O9

10967936204444

I O

0I34

1717600000000056445003304020040400000000022222222000441725331

A (280,1070)12.807.01

24.806.926.535.3415.10

18.6014.5016.8018.008.008.3021 .oo5.2011.7011.1015.7028.007.0219.7013.709.309.7224.604.55

19.904.9518.60

22.109.0026.9025.25

7.046.989.399.20

20.9022.1017.5018.9018.2018.1017.5017.3017.4017.4018.2022.0013.3012.5013.0011.306.8426.20

7.036.479.4519.808.259.896.91

"

2415

t (obs)23,93632,17658,0969,47343.29 1

35,49536,23773,16565,15348,91352,16415,83216,65244,106I 1,25213.66083,45137,66940,0579,61049,61435,03415,6245,57357,62818,1065 1,53322,06220,62627,36615,48138,50536,143

9,6389,55615,76210,76625,88027,36621,60223,36022,46622,28119,40619,18419,21819,35120,26924,55214,71813,81324,22721,03312,70037,5039,62442,96117,29450,8195,377

5,71831,717

t (pred)25,4403 1,52556,1859,44042,925

34,44537,65075,86060,85044,92050,42014,44013,98044,35010,43013,98083,36537,65037,9709,44050,58532,43015,9306,33556,18516,960

55,39021,43019,16026,93015,47037,97037,470

9,4408,94015,93011,46026,93026,93021,43022,92021,43021,43019,16019,16019,16019,16020,65024,66015,15013,66025,44021,43013,41037,9709,44042,92517,21050,585

6,3356,3353 1,525

% D ev______-6.282.023.29

0.350.852.96

-3.90-3.686.608.163.348.7916.05

-0.557.30-2.340.100.055.211.77- .967.43- .96-13.662.506.33-7.482.877. 1 11.590.071.39

-3.672.056.44- .0 7

-6.45-4.061.590.801.894.613.82I .2 70.120.300.98- .8 8

-0.44-2.941.10

-5.01-1.89-5.59- I .251.910.080.48

0.46-17.81-10.780.61(cont inued)


6/13

2416Table 5 . ContinuedNo. Pro te in

. ~~.~~ . ~.

M W92 BS A 66,29693 C A ( B -B ) 28,98094 RNase 13,69095 R-RNase 13,69096 Helicase 11 81,98997 Reprotein 76,40098 Luciferase 40,10899 Luciferase 40,108100 Luciferase 36,349101 Luciferase 36,349102 LuciferaseB 76,457103 LuciferaseB 76,457104 D H F RE.C . ) 17,680105 D H F RM ) 2 I ,446106 Thermolysin 34,334107 Peptidase 25,165108 Neurotoxin 6,867109 Glucagon 3,483I10 Acidrotease 34,239I 1 Rhodanese 32,800112 Staphylokinase 15,660113 Glycerolinase 56,106114 T P IR ) 26,750115 0-Lgb 18,300116 PG K 44,570

~

~ ~ ~ ~~ ~ ~

~ ~ ~

Ty r Cystine A (280,1070)2 20 17 6.277 0 16.9000 6 0 6.25

133 6 12.909 22 0 10.056 16 0 13.406 16 0 14.102 10 0.3 02 10 0.1 08 26 0 1 1 508 26 0 11.405 0 17.6033 28 0 16.605 14 3 18.302 I 40.80I 2 0 23.004 14 2 12.608 I 1 0 17.501 9 0 11.70

13 18 0 17.505 0 12.902 2.052 0 4.95

~.~

~ ~ ~ ~ ~

~ ~~ ~ ~

~ ~ ~ ~~ ~

~ ~~

t (obs)41,56848,976

9,1458,556105,76676,78253,74556,55226,53525,80887,92687,16131,11725,09256,99446,05214,2838.01 143,141

57,40018,32298,18634,50816,56222,062

~

~ ~

C.N. Pace et al .

~ ~ ~ ~~ ~ ~ ~

(pred) Vu Dev~

42,925 -3.2750,4202.95

9,4403.23 8,9404.49106,520 -0.7182,2807.1656,8405.7656,8400.5125,900 2.3925,9000.3682,740.9082,740 5.0733,4607.5325,440 - .3958,2202.1548,735 -5.8312,990.0 68,4805.86

43,110 0.0760,390 -5.2118,910 -3.2198,3200.1433,460 3.0417,210 -3.922 1,430 2.87

a T h e n a m e of the protein or an ab breviation is given in colum n 2. Abbreviations: CA(H -C) , i sozyme C of human ca rbon ic anhydrase ; CPA,bovine carboxypeptidase A; CTgen, bovine chymotrypsinogen; a-Lact, bovine a-lactalbumin; 0-lgb, bovine 0-lactoglobulin; BSA, bovine serumalbumin; H SA, hum an serum albumin; 3-K1, A5-3-ketosteroid isomerase; Alc DH ase(Y ), yeast alcohol dehydrogenase; CPB, bovine carboxypep-tidase B; GA P D Ha se, yeast glyceraldehyde-3-phosphate DHase; Glu DHase, bovine glutamate dehydrogenase; Lac rep, Escherichia coli lac re-pressor; Tgen, bovine trypsinogen; AAT, chicken mitochondrial aspartate aminotransferase; CA(H-B), isozyme B of human carbonic anhydrase;A-crystall in , bovine a-A2-crystall in; B-crystall in , bovine a-B2-crystall in; 11-crystall in , bov ine yll-crystall in; a-C T, b ovine a-chymotrypsin ; Con-can, jack bean concanavalin A; SN , staphylococcal nuclease; Alc DHase(H), horse iver alcohol dehydrogenase; PC kinase, yeast phosphoglycer-ate kinase; Mb, myoglobin; R-lysozyme, reduced lysozyme; R-RN aseA , reduced RNase A; Cyt c , bovine cytochrome c; BPTI, bovine pancreatictrypsin inhibitor; CA(B-B) isozyme B of bovine carbonic anhydrase; DHFR (E.c.) , E . coli dihydrofolate reductase; DHFR ( M ) , muscle dihydro-folate reductase; TPI(R), rabbit tr iose phosphate isomerase; PGK , yeast phosphoglycerate kinase.

The m olecular weight and the Trp, Tyr, and cystine contents are given in columns 3-6. We confirmed most of this information in at least oneof the available databases, but there aretill likely to be a few erro rs. In gathering t values, we would generally start w ith the comprehensive com -pilation of Kirschenbaum (1978). We tried to use only values obtainedby the dry weight method, am ino acid analysis , K jeldahl nitrogen determi-nations, or the Edelhoch method. In column7 , A(280, IVu) is the ab sorbance of a 1 % solution of the protein at 280 nm. Many of th e A(280) valueswere determined atwavelengths other th an 280 nm. To correct values at other wavelengths, A ( X ) , o 280 n m , A(280), the following equation was used:

A(280) = [ A ( X ) ] [ t (280, 6 M C d n H C l ) / t ( h , 6 M G d n H C I ) ] ,where ~(280 , M G d n H C I ) a n d (X, 6 M G d n H C I ) a r e t values for the protein at 280 nm and X calculated with Equation 2 using the Trp, Tyr,and cys t ine con ten t o r each p ro te in g iven in th is ab leand hemode lcompounddata orGdnHCl nTable 3 . I n co lumn 8 , t ( o b s ) =[O.l][MW ][A(280,1%)] is the m olar abs orption c oefficient at80 nm. In column9, t (p red) =5,500(#Trp)+1,49O(#Tyr)+125(#cystine) is the t(pred )value for he protein based on he recomm ended t values for Trp, Tyr, and cystine given in Table 9. In column 10 , %Dev = 100[ t (obs)-t (p red) ] / t (obs) .

Some of the entries were taken from tables in other papers:-1 1 are f rom Tab leV in Wetlaufer (1962); 18-36 are from Table 3 n Gill and vonHippel (1989); and 37-47 are from Table 1 in M ach et al. (1992). I n the latter two tables, we averaged the most reliable values when more thanon e t value was given. Also, any errorswe found in these tables w ere corrected. In Wetlaufer 1962), for example, the Trp, Tyr, and cystine con-tent was correct for only 1 of the 11 entries, and in Gill and von Hippel (1989), all of the cystine contents were wrong.Entries 11-17 are from Li et al . (1993), Kuliopulos et al . (1989), or from data supplied by Paul Talalay. Fo r entry 12 , Y14F denotes the mutantof 3-ketosteroid isomerase in w hich Tyr 14 is replaced by Phe. This same nom enclature was used throughout the table for other mutants. The Evalues in entries 48-57 are from Kalnin et al . (1990). Entries 58-66 are from the Privalov lab (Privalov et al . , 1989; Gri ko et al ., 1994; Yu et al.,1994). For ba rnase and mutan ts , en t ry67 is from Lees and Hartley (1966), an d 68-72 are from Loewenthal et al . (1991). For RNase T1 a n d m u -tants, entries 73-80, th e t values are from Tables 1 and 7 . Th e references for th e rest of the entries are as follows: 81-83, Elwell and Schellman(1977); 84-90, Table 3 footnote; 1-93, Nozaki (1986); 94,95, White (1961); 96, Runyon and Lohman (1989); 97, Amara tunga and Lohman 1993);98-103, Sinclair (1995); 104, 105, C. Clark (pe rs . comm.) ; 106, Voordouw and Roche (1974); 107, Robinson (1975); 108, Hauert et al . (1974); 109,Kay and Marsh (1959); I IO , Fukum oto et al . (1967); 1 1 1, P. Horowitz (pers. comm.); 112, Damaschun et al. (1993); 113, D. Pettigrew (pers. comm.),Pettigrew et al . (1988); 114, B. Gracy (pers. comm .); 115, C u p 0 a n d P a c e (1983); 116, A d a m e t a l. (1985).


7/13

Protein absorption coefficients 2417the uncerta inty in the E value estimates is lowest near 278 nm ,as expected, where i t is k2-37 0 for tk(Trp ) and th(Ty r) .Average t values at 280 n m fo r Trp, Tyr, and cystinebased o n 116 measured E values fo r 80 proteinsIn order to es t imate average values for Trp, Tyr, and ystineresidues in globu lar protein s, we compiled the 116 measuredE values for 80 different proteins in Table 5 . Similar bu t less ex-tensive tables were compiled by Wetlaufer (1962), G ill and vonHippel (1989), an d Mach et al. (1992). These tables all containederrors , and there are surely som e errors in our table , but wedou bt that hey will significantly change the average values wederive from our analys is .

The analysis used above toestimate the t values in Table 4 wasapplied to the data n Table 5 to ob tain the results in Table 6.The top l ine shows the es t imates obtained for t (Trp), t (Tyr) ,and t (cys t ine) when all three are varied. No te that (cys t ine)=12 8 M" cm-l is in the range expected based on th e results inTable 3. In rows2 and 3 , cystine is fixed at th e e xtre mevaluessuggested by Table 3 to sho w th at th e value assigned to t(cys-tine) does not significantly effect the estimates of t(T rp ) andE(Tyr) derived from the an alysis. Th e four th row gives the re-sults of an nalysis run using only the pro teins that contain Trpresidues. Th e E value depend s strongly on wavelength near 280 nmfor p ro te ins wi thou t T rp o tha t E canno t be determined as ac-curately. Also, we will see tha t thre eof the four prominent out-liers in our predicted values, 3-KI(Y14,55F), B PT I, an d insulin,are prote ins with no Tr p residues. For this smaller sample of93 E values, the estimatesof t (T rp) andE(Tyr) do not diff er ig-nificantly fro m those for the omplete data set. Thus, excludingpro te ins wi th T rp=0 does not significantly c hang e the resultsof the analys is . The f i f th row ives the results obtained for the

Table 6 . Estimated molar absorption coefficients fo r Trp,Tyr, and cystine at 280 n m based o n analyses ofthe E values in Table 5~ ~~ . .. -~ ." ~ ~ - - ~.

TrP TY r CystineI1 6 t values"Cystine fixedbCystine fixedbTr p p resent 'T r p =OdCystine present'Cystine =0'First 1 1 entriesg

5,443 f 1415,444 f 1575,437 f 1575,434 f 1755,430 f 1395,417 k 2575,485 f 129

-

1,503 k 7228 f 1041,501 f 62 1351,508 k 62 I101,508 k 71251,472 ? 38251,484 f 52 1251,540 f 1061,507 f 51

-I25"" "

a These t values at 280 nm in M" cm-l were estimated by minimizingthe sumof the square of deviationsf th e ~ ( 2 8 0m) values calculated withEquation 2 from the measured2 8 0 nm) values given in Table 5 using thenonlinear least-squares analysis of Johnson and Frasier 1985).T h e E value for cystine was fixed at th e extreme ossible values (Ta-ble 3) to show that the values estimated for Trp and Tyrvalues are notsignificantly changed by the value c hosen for for cystine.T h e 93 t values for proteins containing Trpwere included.T h e 23 E values for proteins w ith Trp=0 were included.e T h e 56 t values for proteins containing cystinewere included.'The 60 values for proteins with cystine =0 were included.This analysiswas run on the irst 11 entries in Table 5. These 11 pro-teins were the proteins included in Table V n Wetlaufer's review (1962).

23 proteins that contain no Trp. The es t imate for Ty r differsby only about 2% from th e es t imate based on the ent i re dataset . In rows 6 and 7 , we subdivide the data set into approxi-mately equal halves on th e basis of cystine content. Here, theTr p es t imates d o not differ s ignificant ly, but the Tyr es t imatesdiffer by a lmost 4%.Finally, in the last row , we give an analysisof 11 proteins used by Wetlaufer (1962) to first test the possi-bility of estimating E values for proteins using model com poun dda ta. W e will see in Table 9 tha t these t es t imates for Trp, Tyr,and cystine are considerably better at predicting t values for aprotein than the E values used by Wetlaufer (1962).

DiscussionMeasuring the absorption coefficient of a proteinT h e E value for RNaseTI has been measured by all of the stan-dard metho ds over the years and the results are not in goodagreement (Table 7). In the mid-l9 80s, we sent a sample ofRNase T1 to Dr. Yas Noza ki , who had eveloped an im provedproc edur e for determining dry weights on small protein sam ples(Nozaki , 1986). He fo und t278=17,190 M" cm- ' for RN aseTI and this was reported in a paper describing the p urificationof recombinant R Nase T1 Shirley & Lauren ts, 1990). Becausethis differed by =20% from he value of 21,180 M" cm"(1.91 mL mg" cm") tha t was in general use, E was measuredseveral times in our la bora tory by the dry weight meth od andthis resulted in the value of 18,520 M" cm-l tha t was repo rtedin the calorimetric studies of RNase T1 by Hu et al. (1992).In that paper , we found that AH,,,, , , H o f f >AH,,, and were notsure why. In t rying to resolve this ques t ion, the Privalov labo-ratory measured E =20,630 M" cm-I using amin o acid anal-ysis and a Kjeldahl nitrogen method (Yu et al. , 1994). Theirpreferred method is the Kjeldahl procedu re of Jaenicke (1974);we tried th is me thod in ou r laboratory and found 19,630 M"cm". For other reasons , we had begun a s tudy of mutan ts ofRNase T1 differing in their aromatic amino acid content . Wereplaced T rp 59, which is 100% buried, with Tyr and Phe, a ndwe rep laced hemosthyperexposed esidue in theprotein,

Table 7. Measured m olar absorption coefficientsat 278 n m fo r RNase TIReference e ( M - l cm")ethod

~

Takahashi (1961)Takahash i (1962)Minato et al . (1966)Shirley and Laure nts (1990)O k a j i m a e t al . (1990)Grunert et al . (1991)Hu e t a l . (1992)Yu et al. (1994)This paper

18,52021,18020,29017,19020,51017,30018,520 k 33 019,160 k 27020,630 t 78020 , 630 f 1 , 11019,630 f 67 018,470 k 21 019,215 f 40

??Dry weightAmino acid analysisVacuum driedDry weightEdelhoch methodNitrogen analysisAmino acid analysisNitrogen analysisDry weightEdelhoch method

Mean f meaneviation 19,330 f 1,040


8/13

2418 C.N. Pace et al.A sp 49 , with Trp, Ty r, Phe, an d A la . Thevalues for w ild-typeRNase TI an d hese mutan ts were determined on the same tocksolut ions of prote in us ing the same absorbance measurementsby the dryweight and th e Edelh och m ethods . T heseesults aresummarized in Table 1 . Based o n this tortu ous 10-year experi-ence, we thought that the delhoch method might be the methodof choice for determining the value for a protein.We will sum-marize some of our reasoning.We have had on ly limited experience with am ino acid analy-sis or Kjeldahl nitrogen determin ation s, but the accuracy of thesemethods does not eem adequate for determining an value fora prote in. The average percent error for 16 amino acids deter-mined by amino acid analyses a t 62 different sites was 10.9%(1994 report from theAssociation of Biomolecular Resource Fa-cilities on the quality of amino acid analyses) . S imilarly, withthe Kjeldahl me thod , most of ten used with proteins, Jaenicke(1974) states: The reproducibility in control analyses with add eds tanda rds is k 570. his is the differen cebetween the t valuesdetermined by nitrogen analyses by our lab and the rivalov lab(Table 7).

We have had more experience with the dry weight method .Und er optim al cond itions, the precision in a given dry weightdeterminat ion is a b o u t 5 0 . 2 % (Kupke & Dorrier , 1978), a ndwe estimate the reprodu cibility of independen t determinationsto be ?2-3%. With RNase T I , we consis tent ly observe lowerE values from th e dry weight method than from the Edelhochmethod (Table 1). The differen ce is not large; the values fromthe dry weight method av erage about 5% lower than the valuesfrom the Edelhoch method.We are not sure why, but there aremore uncertainties with th e dry weight method. Hunters (1966)studies with the dry weight m ethod led to these conclusions: (1 )The basic problem in dry weight determinations is a lack ofadequate cri teria for determining theompleteness of water re-moval when this process may be attended y oxidative destruc-tion of the prote in; and (2 ) . . . the actual value obtaineddepended on the method of dryingwhich had been employed.Different drying conditions gave dry weights which varied upto a lmost one partn a hund red. The best conditions for obtain-ing the true dryweight of a protein are somew hat ncertain andprobably vary from protein to p rote in.We agree. Pace (1966)used dry weights to determ ine the conce ntrationf P-lactoglob-ulin stock solutio ns over a period of years and concluded: . . .the difficulty lies in the determination of protein concentrationsby dry weights, w here, althoug h the precision was uniformlygood, the accuracy was not . Here are some of the qu es t ions :Firs t , does drying a t107 C to a const ant e ight remove al l ofthe water? Second, doesextensive dialysis against distilled wa-te r or passing a protein through a mixed-bed ion exchangecol-umn remove al l of the counterions? Third, why does the dryprotein slowly begin to gain weight when drying is continu edover a period of days? (I t should be ossible to answer some fthese questions using mass spectroscopic techniques.) Becauseof these uncertaintie s and because th e dry weight method is sot ime-andprotein -consu ming , it is clear hat heEdelhochmethod, i f sufficiently accurate, would be the preferred methodfor measuring t for a prote in.Edelhoch methodEdelhochs 1967 p.aper describes a method for determining theT r p a n dTyr content of proteins. The information in this paper

was first used to determ ine values for p roteinsby groups at theUniversity of Oregon (Butler et al. , 977; Elwell & Schellman,1977), and has s ince become know n as the Edelhoch method.A clear description of he meth odwas given by Gill an d von Hi p-pe l (1989). The basic assumption of the delhoch method is thatt values determined for model com poun ds for Trp, Tyr, andys-tine in 6 M GdnHC l can be sed to approximate the values forthe same chromophores n a prote in unfolded n 6 M GdnHC1.Figures 2-5 in Edelhochs pap er (1967) show that this is a rea-sonable assumption. They show that the spectrumf a proteinin 6 M GdnHCl canbe reproduced quiteaccurately using a mix-tu re of model comp ounds for Trp, Tyr, and cyst ine n 6 MGdnHC1. T he absorpt ion spectrum of the model compoundmixtures is shifted slightly to sho rterwavelengths compared tothe protein spectru m, but the magnitude o f the ab sorbances nearA,,,,, are very similar.Further support for the basic assumption of the Edelhochmethod comes from an analys is f 39 measured E values on 27different prote ins n 6 M G d n H C l a n d8 M urea. In addi t ion toth e 7 measured t values in 6 M G dnH Cl given in Table 1, we used9 t values measured in 6 M GdnH Cl by Nozak i (1986), 9 t val-ues in 6 M GdnH Cl f ro m Lee and T im ashe f f (1974), 1 1 E val-ues in 8 M urea from Prakash e t a l . (1981), 2 t values in 6 MGdnHC l from Span et a l. (1974), an d 1 value in 8 M urea fromSela et al. (1957). We fixed E = 125 M c m a t 28 0 nm forcystine (Table 3) for reasons explained above. We found thatE ~ ~ , ~5,450 M cm for T rp , and tzg0=1,265 M cm forTyr minimized the percent deviation between the t values cal-culated with Equat ion 2 and the 39 t values measured in 6 MG d n H C l a n d 8 M urea. This Trp value is about 4% lower andthe Tyr value abou t 2% lower than the values given in Table 3fo r 6 M GdnHC1. T his is reasonable agreement considering theuncertainty in the experimental E values for protein s in 6 MG d n H C l a n d8 M urea, andshows that the t values for Trp , Ty r,and cystine in proteins un folded in these solvent ar e similar toth e t values for model chromophores in the same solvents .

O u r results on RNase TI with the Edelhoch method (Table 1)encouraged us to t ry the method n some o ther well-character-ized proteins. These results are sum marize d in Table 2 . Thesevalues were obtained by preparing a solution of the protein,il-tering the solut ion, di lut ing the solut ion into buffer and 6 MGdnH Cl solut ions , and measuring the absorpt ion spectra ofthese two solutions. All of this can be done in a few hours. Forthese eight protein s, the a verage percen t differ enceetween re-sul ts from the Edelhoch method andesults from the literatureis 2 .0 %. The only deviat ion greater than% is for P-lactoglob-ulin , and , as explained in the Results, we think this is becausethe l i tera ture t value is too high. Thus, we conclude that theEdelhoch method is both the simplest an d most accurate methodfor determining the molar absorpt ionoefficient for a protein.Predicting the absorption coefficient of a proteinWe first measured the t values for the model compounds forTrp, Tyr, and ys tine (Table 3) in order to es t imate the78/25 1absorbance ra t io expected for pure RNaseTI . We used this ra-t io to monitor o u r success in remov ing a yellow pigment du r-ing the purification of RNase TI (Pace et al., 1987). At the time,we though t it might be possible to use the t values in w ater tomodel the exposed chromophoresn a protein , a nd the valuesin propanol to model the buried chrom ophoresn a protein and


9/13

Protein absorption coefficients 2419that this m ight allow us to calc ulate E values fo r p rote ins withreasonable accuracy. As illustrated in Table 8, the results werenot e ncourag ing. Pro panol was a surpris ingly poor model forburied Trp and Tyr chrom ophores . Con sequently, we neededa better appro ach for estimating the average E values for the Trp,Tyr, andcystine chro mop hore s in globular proteins, an d this ledus to the experiments summ arized in Table 4.T h e E values in Table 4 a re the ave ragevalues for Trp, Tyr,and cys t ine tha t can best reproduce the measured absorptionspectra for 18 well-characterized p roteins using Eq uat ion 2. InFigure I , we comp are the a bsorpt ion spectraased on the pec-tra l propert ies of Trp andyr in these18proteins to the absorp-t ion spectra for Trp and Tyr in propanol andM GdnHC1. ForTyr, theaverage spectral p roperties n the proteins are intermedi-atebe tween hespectra lp rope r t i e s in p ropa nol a n d 6 MG d n H C I . Fo r Trp, the spectra l propert iesn prote ins arecloserto hose observed in 6 M G d n H C l h an h o s e o b s e r v e d i npropanol . In numerical terms, the t values ob tained from ananalys is of the p rote ins (Tables , 6 ) a re m ore than10% lowerthan the E values measured in propanol (Table 3). T he E valuefo r Tr p n proteins is particularly surprising because it is less thanthe value measured in any of the olvent systems, including wa-ter. It is not clear to us why these largely buried T rp residues be-have mo re like they are in water than in a nonpolar environment.Richards (1977) has pointed out that "The interior ofproteinis not anoil dro p but esembles rather a molecular crystal," andthis might explain why propanol is such a poor model for theinterior of a protein, a t leas t with regard to the sp ectra l p rop-erties of Trp and Tyr . Theseesults explain why E values basedo n 6 M GdnH Cl are reasonably successful in predicting E val-ues for folde d prote ins (Gill & von Hippel , 1989). Note in Ta-ble 8 that the t values for Trp an d Tyr based on proteins are

Table 8. Estimating the molar ab sorption coefficientof RNase TI at 278 nm a "uriedpropanol)xposedwater)T r p (1) 1.0 x 5,780 =5,780 0.0 X 5,590 =0Tyr (9) 7.7 x 1,735 = 13,360 1. 3 X 1,295 = 1,684S-S (2) 1.6 X 15 0 =240 0. 4 X 125 =50Tota l9,380 + 1,734 =21,114M o d e l t (calculated)Water (Table 3) 17,495 (9.1% low)6 M GdnHCl (Tab le 3) 18,380 (4.5% low)Propano l (Tab le 3) 21,695 (1 2.7% high)Propano l (bur ied ) +water (exposed) 21,114 (9.7% high)Pro te in (Tab le 4 ) 19 ,200 (=equa l ) (measured) (Tab le 1) 19,215 M" cm "

a All of the t values ar e in M" cm". The upper portion of the tableshows the calculation o f t a t 278 nm fo r RN ase TI us ing the t valuesfor water for the exposed chromophores and thevalues for propanolfor the buried chromophores. The percent buried was calculated usingthe Lee and Richards program (Richards, 1977) and the 9RST crystals t ruc tu re of R N a s e T1 f r o m h e B r o o k h a v e n P r o t e i n D a t a B a n k(Martinez-Oyanedel et al . , 1991). The lower portion shows the calcu-lated absorption coefficients using values from the tables noted. Thelast l ine shows the measured t value.

1 Tryptophan62006m l80 048 W 44600 4 6 M GdnHCl

I270 2727476 278 28082 284 286 288Wavelength (nm)

I800 7

1600

14001300 --E I200 -

8 1 1 0 0 --1000-

9W -80 0 -700 - ' 06 M GdnHCl600 10 272 214 216 218 280 282 284 286 288

Wavelength (m )Fig. 1. Absorption spectra for ryptophan and yrosine. The curveslabeled 6 M GdnH Cl and p ropano l a re ased on the absorp tion spec-tra used to determine the results given in Table 3. The curveabeled pro-tein is based on the results in Table 4.

much bet ter a t predict ing the value for RNase TI than thosebased on any of the solvent systems.These results encouraged us to examine a arger sample of pro-teins. We compiled a list of 116 measured E values for 80 dif-ferent proteins (Table ) . We focused our a t tent ion o n80 n m ,where most E values are measured. When an E value was mea-

sured a t a wavelength other than 28 0 nm , i t was corrected to280 nm using the contentof T r p , T y r , a n dystine and the val-ues in 6 M Gd nHC l given in Table 3, as explained in Table 5 .The error ntroduced by this correction will generally be less ha n2% . In Table 6 , we show the values of E f o r T r p , T y r , a n dys-tine that minimize the sum of the square s of th e dev iationsofE values calculated with Equation 2 from the measured valuesgiven in Table 5. The first row s based on an nalysis of all 116E values in Table 5 . The t values are determ ined to G!.6% fo rT r p , t o +4.8% fo r Tyr , and to + 8 l % for cystine. The seco ndand third rows show that the E value assigned to Cys does notsignificantly change th e E values estimated for Trp and Tyr . C on-sequently, we fixed E =125 for cys t ine for theest of th e an aly-ses. The next five rows analyze various subsets f the 116 E valuesranging in size from 1 1 t o 93 entries . The es t imates off o r T r prange f rom 5,417 to 5,485 (1.3% difference), and the es t imatesfor Tyr range rom 1,472 to 1,54 0 (4.6oio difference) for thevar-


10/13

2420 C.N. Pace et al.ious subsets and the comp lete data et. Thus, we doub t that thet values es t imated for Trp, Tyr, and cys t ine would change ap-preciably i f a larger set of t values were an alyzed.The average percent d eviation, aveyodev, is a different mea-sure of how well a given set of t values for Trp , Tyr, a nd cys-tine will give agreement between t(obs) and t(pred) (Table 9) .It should be a better measure because each E value will beweighted mor e equally than in the least-squares method of Jo hn-son and Frasier (1985) that was used to obta in the esults in Ta-bles 4 a nd 6. In Table 9, we show the aveyodev obtained usingmeasured and estimated E values for T rp, Tyr , a nd ystine fromvarious sources. Th e first fo ur lines show the avevodev obtainedusing the E values for Trp , Ty r, an d cyst ine measured in foursolvents (Tab le 3). The fifth line shows the result fo r the val-ues used by Wetlaufer to predict E values in his 1962 review. Itis interes t ing that 8 M urea an d 6 M GdnH Cl are bet ter m odelsolvents for the buried chromophores of a prote in than wateror propa nol. However, as the rest of Table 9 shows, none of thesolvent systems allow us to predict t values for folded proteinsnearly as well as the E estimates for Trp, Tyr, and cystine de-rived from an alyzing prote ins directlyT h e t values from Table 6 that ere based on our analys is ofall of the proteins in Table 5 gave an ave%dev=3.856%. Note ,how ever, that the avevodev is slightly smaller using t values de-rived f rom an analy sis of just the first l l proteins in Table 5 .This shows that the combination o f t values that gives the bestfit using the east-squaresprogram ofJohnsonandFras ier(1985) does not give the minimum value for the avevodev. N otealso that the values that Mac h et al. (1992) derived from a sim-ilar analysis of a list of 81 measured E values for 32 different p ro-teins gave ave%de v =3.860%. This shows that qui te differentsets o f t values for Trp , T yr, and ys t ine can give comparablevalues for aveyodev.We next used trial a nd error to search for rounded values oft for T rp an d T yr tha t w ould ive an avevodev less than thoseobtained using the E values from Tables4 and 6. The est com-bination hat we fou nd was eTrp = 5,500 M" cm", t T y r=1,490 M -l cm", and t ,y5 t ine 125 M" cm" (avevodev =

Table 9. Average percent deviation of measured E values inTable 5 fro m predicted t values using various Trp, Tyr,and cystine E values"Source T rp Tyrystineve%devh6 M G d n H C lTab le 3) 5,685 1,285 125 6.8818 M ureaTable 3) 5,635,300 115 6.700WaterTab le 3) 5,630 1,215 110 9.188PropanolTable 3) 6,075,680 135 11.858Wetlaufer (1962) 5,550,340 150 5.894Table 6,443 128 3.856Table 6 (first 11 entries) 5,485 1,507 125 3.852Table 4,517 1,474 125 3.900Machtl. (1992) 5,540,480 134 3.860Recommended 5,500,490 125 3.836

~~~....~. - .~. ~~ -~ ~

~ " . .."~." T h e t values are in M-I cm-l. Predicted values were calculatedwith Equation 2 using the t values given here and the num ber of T r p ,Tyr, and cystine residues fo r each protein in Table 5.AveVodev = (1/116)C I % D e v . % D e v = 100[c(obs)- ~ ( p r e d ) l /t (obs) .

3.836), an d this is denoted as recom mended at the botto m f Ta-ble 9. In the last column in Table 5 , we list the percent devia-tion, VoDev, of the measured t values, E(obs), from the t valuescalculated with Equation 2, t(pr ed), when these values are used.(Another possible combination was 5,450 for Trp,1,500 for Tyr,an d 125 for cystine [aveVodev =3.8501.) This is encouraging.It shows thatwe can predict E reasonably well knowing just theTrp , Tyr , and ystine content of the pro tein.As we will now see,we can do even better i f the protein contains at least one Trp.The only Trp-containing protein that gave a percent deviationgreater than 10% was a-B,-crystallin (entry 42 in Table 5 ) . Thist value came from theist of Mach et al. (1992), and they pointout that the ight-scattering correction for this protein was 7%,greater than for any other proteinn their list. Th us, the E valuefor this protein is surely less reliable than mos t. In contr ast, 7 f23 entries with no Tr p residues had percent deviations greaterth an lo %, with BPTI th e greatest at -17.81% (Table 5 ) . Usingour recommended values for Trp , Tyr , andystine, the ave% dev =3.167% for the 93 entr ies containing Trp, but 6.541% fo r the23 entries with Trp =0. Th us, it is clear that we can predicte values for proteins containing T rp much better than we canpredict t values for p roteins lacking Trp. This is not surprising.Based o n our recommended E values, a tryptop han esidue isequiv alent to 3.7 tyrosine residues and 44 cystine residues. Fur-thermo re, it is clear f rom the mod el comp ound data that thet values for Trp are less sensitive to the environment than theE values for Tyr (Table ). This is even clearer when the proteindata are considered. For the three 3-KI mutan ts containing asingle Tyr residue and o T r p or cystine residues (entries 15-17in Table 5 ) , the ~(2 80 ) a lues are 1,310, 1,640, and 1,740 Mcm " . For the 14 proteins in Table 5 with no Trp res idues , thea v er ag e ~ ( 2 8 0 ) alue per Tyr = 1,490 f 140 M -' cm". (A cor-rection for cystine was made using t(280) =125M" cm -' [Ta-ble 31.) Thus, the aver age Tyr in RN ase A has t(280) = 1,490M-l cm", and the t (pred) values for RNase A are in goodagreement with th e t(obs) values; but for the average Tyr inBPT I, ~(2 80) 1,270, and for he average Tyrn insulin, t(280)=1,290, and their %D ev values in Table 5 are the major out l iers .Th us, in folded proteins there is a wider range of E values forTyr residues than there is for Tr p esidues and this is the mainreason that it is more difficult to predict E values for proteinsthat contain no Tr p residues. (See Brandts and Ka plan [19731for an excellent discussion of the spectral propertiesof Tyr res-idues in RNase A , insul in, and BPTI. )

In sum ma ry, we suggest using this equatio n:t(280 nm) =5,50O(#Trp) +1,49O(#Tyr) + 125(#cystine)

to predict the E value for a folded protein in water at 280 nm.I t is quite reliable for proteins that con tain Trp esidues, an d lessreliable for proteins that do not .However, the Edelhoch methodis convenient and accurate , a nd theest app roac h is still to m ea-sure ra ther than predict t . If you are un lucky and have a pro-tein with no T r p or Tyr , th e pap er by Scopes (1974) should beconsul ted.Materials an d methodsThe water used was purified by reverse osmosis and then double-distilled n a glass still. MOPS buffer was purchased fromSigma . Urea was pur chased from United States Biochemicals


11/13

Protein absorption coefficients 24 2 1(u l t rapure ), and f rom S igm a (u l t ra ) . Guanid ine hydroch lor idewas purchased from Heico (extreme pu ri ty) , and from UnitedStates Biochemicals (ultrapure). I-Pr op ano l was purchased fromFisher. Most of the model compo und spectra ere determinedwith these compounds: N-acetyl-tyrosine ethyl ester (N-Ac-Tyr-OEt) f rom A merican Tokyo Kasei; N-acetyl- t ryptophan ethylester (N-Ac-T rp-OEt), oxidized glutathione (Grade111and ul t ra-pure) , and cyst ine (ul t rapure) from Sigma. The proteins usedwere from the fol lowing sources : r ibonuclease TI and mutantsprepared as describedby Shirley and Lau rents (1990); barnas eprepared as described y R.W. Hart ley (pers . comm.); 0-lacto-globulin A prepared as described by Aschaffenburg and Drewry(1957); 0-lacta lbumin prepared as described by Robb ins andKronman (1964); ribonuclease A (Sigma, R-5500 and TypeXllA ); hen lysozyme (Wo rthing ton, 2 x crysta ll ized); hu mancarbonic anhydrase I1 (Sigm a, C-6165); carboxypeptidase A(Sigma, 65 10); chymotrypsinogen (W orthin gton , 5x crystallized,and S igm a Type11); papain (Sigm a, P-4762); hum an se rum al-bumin (Sigma, A-3782); bovine serum albumin (Calbiochem,12659, and Sigm a, A-4503); trypsin (Sig ma, T-8253); thermol-ysin (Sigma, T-1512); staphyloc occa l nuclease (a kind gift fro mDr. David Shortle); bovine pancreatic trypsin inhibitor (Sigma,T-0256, and Boehinger Mannheim); ovalbumin (W orthington,2 x crystallized); an d bovin e insulin (Sigma, 1-5500).Model compound spectraThe m ode l com poundswere dried in a vacuum desiccatoroverphosphorus pentoxide. Stock solutions of N-Ac-T rp-OEt, N-Ac-Ty r-O Et, oxidized glutath ione, and cystine were prepared byweight. The U V spectra (250-350 nm) were measured at room tem-perature with a Cary model 15 or model 219 spectrophotom eter.

Dry weight methodThe dry weight procedure used to determ ine protein conce ntra-tions was based on a methodused routinely in the Charles Tan-ford laboratory that came earl ier from the Cohn and Edsal llaborato ry (R obinso n, 1975; Nozaki, 1986). Kupke and Dorrier(1978) called a metho d similar to this "the comm only taughtmethod of yesteryear," and stated that t has a precision of about1 part in 90 . Detailed descriptions of the dryweight method canbe foun d in Hunter [1966], Kup keand Dorrier [1978], and N ozaki[19861.) The dry weight me thod used to determ ine the E valuesgiven in Tables 1 and 2 is as follows. Protein was dissolved inwater to a concentra t ion of 2 5 mg/m L an d f i ltered through a0.2-pm Acrodisc filter. This solutionwas then dialyzed againstat least three changes of water to b ecom e o ur protein stocko-lution. Aliquots of the stock solution were weighed into care-fully cleaned, d ried, an d tared weighing bottles. T he so lutionswere first evap orated to dryness at =100"C, and then r ied fur-ther in a vacuum oven at20 torr and 107 "C. The samples usu-ally reach a constan t weight w ithin about 12 h, but heating wascontinued for 1-2 days to be sure that a constan t eight was at-tained . Under these cond itions, the increase in weight tha t is ob-served at longer times is not significant during 2 days f drying.Typically, the dry weights were run on triplicate samples, eachcontaining enough protein solut ion that a t eas t IO mg of pro-tein were present after drying. Theweighing bottles were cooledin a greaseless desiccator befo re weighing on a Mettler balanceaccurate to kO.01 mg. The s tock solut ionused for determining

the dry weights was di luted into 30 mM MO PS b uffe r , pH 7,and into the same buffer containing 6 Gdn HC l to give solu-tions with absorb ances between 0.5 and 0.9, and their spectrawere recorded from 250 to 350 nm with a Cary 15 spectropho-tometer. When the solutionsshowed significant light scattering,i.e. , significant absorption in the 320-350-nm region, a correc-tion was applied as described by Leach and Scheraga (1960).This information was then used with Equa t ion 1 to calculateth e E values determined by the dryweight proce dure given in Ta-bles l and 2.

Edelhoch methodThe resul ts for the Edelhoch procedureiven in rows 2-8 in Ta-ble l were determined on the same s tock solut ionssed for thedry weight procedure described above. For the other esults fromthe Edelhoch procedure in Tables I and 2, the tock solution wasprepare d by simply iltering theproteinsolution hrougha0.2-pm acrodisc filter. Identical dilutions of the protein stocksolution were mad e into 30mM MOP S buffer a t pH 7 (buffer) ,and into the same buffer containing 6 GdnHCl (6MG). Theabsorpt ion spectra of the prote in n buffer and 6M G was mea-sured between 250 and 350 nm. If these solutions showed ab-sorbance above =325 nm , then the absorbance measurementsnear 280 nm were corrected for a con tribution from light scat-tering as described above. (The implest method to correct forlight scattering is to multiply the absorbance at 30 nm by 1.929to get the light-scattering contrib ution at 280nm or by 1.986 toget the light-scattering con tribu tion at 278 nm. Alternativ ely,2 times the absorb ance at 33 nm gives the scattering co ntribu-tion at 280 nm, and 2 imes the abso rban ce at 31 nm gives thescattering contr ibutio n at 278 nm. These appro aches assume thatthe scattering co ntributio n varies as the nverse fourth power ofthe wavelength as in Rayleigh scattering.) At the wavelengthwhere the absorbance is maximal in 6M G, tA(6M G) s calcu-lated using:

q,(6MG) = ( # T r p ) t A ( T r p , 6 M G )+(#Tyr)eA(Tyr,6MG)+(#cyst ine)tA(cyst ine ,6MG),

where the E, values in 6 M GdnH Cl fo r T rp , T yr , and cyst inewere taken from Table . Now the prote in concentra t ion n the6 M GdnHC l solut ion can be calculated using:C(6MG) =A ( ~ M G ) / E ( ~ M G ) ,

and this gives the protein conce ntration n buffer, C(bu ffer) =C(6MG). The E value at any wavelength fo r the folded proteinin buffer, ,(buffer), can then be calculated using:, (buffer) =A,(buffer)/C(buffer) .

Absorption spectra of various proteinsThe resul ts in Table 4 are based on an analys is f the absorp-tion spectra of th e following proteins (the t value in M" cm "used t o calculate the protein concentration s given in parenthe-ses): carbon ic anhyd rase (hum an C) (51,540 at 280 nm); car-boxypeptidase A (66,750 at 278 nm); chymotrypsinogen (51,330


12/13

2422 C.N. Pace et al.at 282 nm); a-lactalbu min (29,070 at 280 nm); 0-lactoglobulin(17,630 at 278 nm ); lysozyme (37,640 at 280 nm ); papa in (58,570at 278 nm); RNaseA (9,800 at 278 nm); bovine serum albumin(44,070 at 278 nm); human serum albumin36,040 at 278 nm);3-ketosteroid isomerase 4,935 at 278 nm) (the absorp tion pec-t rum fo r this protein was kindly provided by Dr. Paul Talalay);trypsin (36,480 at 278 nm); thermolysin (58,650 at 278 nm);staphylococcal nuclease (16,220 at 278 nm); bovine pancreatictryps in inhibi tor (5,700 at276 nm); insulin (6,070 at 276 nm );RNase TI (19,290 at 278 nm); and barnase 27,370 at 280 nm).The absorpt ion spectra e tween 250 and 350 nm for each pro-tein was recorded using a Cary 15 spectrophotometer . The ab-sorban ce values at each wavelength were corrected fo r lightscattering as described abov e. Theprotein concentration was cal-culated using the E values and wavelengths given above, and henth e e values at 2 -nm intervals etween 272 nm and 286 nm werecalculated us ing Equat ion 1 .

AcknowledgmentsTh i s r e se a rc h wa s su p p o r t e d b y g ra n t s f ro m NIH (GM 3 7 0 3 9 ) , t h e R o b -e r t A . W e l c h F o u n d a t i o n ( A060), and T h e T e x a s A d v a n c e d R e s e a r c hP r o g r a m (010366-181), a n d b y f u n d s f r o m t h e T o m a n d J e a n M c M u l -l i n P ro fe s so rsh i p . T .G. wa s su p p o r t e d b y Ho wa rd Hu g h e s M e d i c a l n-s t i tu te g ra n t 71 195-539101) hrough the Undergraduate Science Tra in ingPro gram . We thank the fo l lowing co l leagues for he lp wi th th i s research :Y a s N o z a k i , G e o r g e R o s e , L e o n a r d P r e s t a , G e o r g e M a k h a t a d z e , P e -ter Priva lov , Bre t Sh i r ley , Doug Lauren ts , Rick Erickson , Tony G i le t to,an d Marty Schol tz . We a lso thank the fo l lowing co l leagues who providedi n fo rma t i o n u se d i n Ta b l e 5 : R o b e r t G r a c y , P a u l H o r o w i t z , D o n P e t -t i g re w, J i m S i n c l a i r , To m B a l d wi n , Da v i d Go l d e n b e rg , T i m Lo h ma n ,C l a y C l a r k , a n d D a v e S h o r t l e , w h o a l s o p r o v i d e ds with s taphylococ-c a l n u c l e a se . Pa u l Ta l a l a y d e se rv e s sp e c i a l t h a n k s fo r p ro v i d i n gs witht h e a b so rp t i o n sp e c t ra o r 3 -k e t o s t e ro i d so me ra se a n d so me o f t sm u t a n t s .

ReferencesA d a m B , B ur ge ss R J , P a i n R H .1985. The folding an d mutual intera ctionof t h e d o ma i n sof yeast phosphoglycerate kinase. Eur J Biochem 152:715-720.A m a r a t u n g a M , L o h m a n T M .993. Escherichia coli Rep helicase unwindsDN A by an ac t ive mechanism. Biochemistry 32:6815-6820.Aschaffenburg R, Drew ry J. 1957. Improved method for the preparation

of crystall ine P-lactoglobulin and a-lactalbumin from cows milk. Bio-chem J 65:273-277.Bailey JE. 1968. UV absorption characterist ics f N-acetyl methyl esters ofthe aromatic amino acids, cystine, and N-acetylcysteine. In : Sober HA,ed . Handbook of biochemistry. Cleveland, Ohio: Chemical Rubber Co.p B-18.B e nso n AM , Su ru d a AJ , Ta l a l a y P. 1975. Concentration-dependent asso-ciation of A-3-ketosteroid isomerase ofPseudomonas testosteroni. JBi olChem 250:276-280.Brandts JF, Kap lan LJ. 1973. Derivative spectroscopy applied o tyrosyl chro-mophores . Stud ies on r ibonuclease, l ima bean inh ib i to rs , insu l in , andpancreatic trypsin inhibitor. Biochemistry 12:2011-2024.Butler AP, Revzin A, von HippelPH . 1977 . Molecu lar parameters charac-terizing the interactionof Escherichia coli lac repressor with non-operatorCup 0 JF, P ace CN . 1983. Conformat ional s tab i l i ty o f m ixed d isu l f ide de-DNA a n d i n d u c e r . Biochemistry 16:4757-4768.

Da ma sc h u n G, Da ma sc h u n H, C a s t K ,isselwitz R, ZirwerD, Gu h rs KH,rivatives of (3-lactoglobulin B. Biochemistry 22:2654-2658.H a r t m a n n M ,Schlott B, Triebel H, Be hnke D. 1993. Physical an d con-Acta 1161:244-248.formational properties of staphylokinase in solution.iochim BiophysEdelhoch H . 1967. Spectroscopic determination of tryptophan and tyrosine

Elwell ML, Sc hellm an JA . 1977. Stabil i ty of phage T4 lysozymes. Nativein proteins. Biochemistry 6: 1948-1954.

properties an d thermal stabil ity of wild type and two m utant lysozymes.Biochim Biophys Acta 494:367-383.Fukumoto J , TsuruD, Yam am oto T. 1967. Studies on mold protease. Part I.Purification, crystallization, and some enzymatic properties of acid pro-tease of Rhizopus chinensis. Agr Biol Chem 31:710-717.Gill SC, von Hippel P H . 1989. Calculation of protein extinction coefficientsGriko YV, Makhatad ze G I, Pr ivalov P L, H artley W. 1994. Therm odynam -from amino ac id sequence data . Anal Biochem 182:319-326.ics of barnase unfolding. Protein Sci 3:669-676.Gru n e r t HP , Zo u n i A,Beineke M, Q uaas R, Georga lisY , Saenger W, Hah n

J Biochem 197:203-207.U . 1991. Studies of RNase TI mutants affecting enzyme catalysis. EurHau ert J, Maire M, Sussmann A , Bargetzd JP. 1974. The major lethal neuro-toxin of the venom of Naja naja philippine nsis. J Pept Protein ResH uC Q ,St u r t e v a n t M ,Th o mso n A,Er i c k so nR E ,Pa c eCN. 1992 .

487-4882.Thermodynamics of r ibonuclease TI denatura t io n . Biochemistry 3 1 :Hun ter MJ . 1966. A method for the determination of protein partial spe-cific volumes. J Phys Chem 70:3285-3292.Jaen icke L . 1974. A rapid micromethod for the determination of nitrogenand p hosph ate in biological material . Anal Biochem 61:623-627.Johnson ML, Frasier SG. 1985. Nonlinear least-squares analysis. MethodsEnzymol 117:301-342.K a h n NN, B a ik a lo v IA, Ve n y ami n ov SY. 1990. Quantitative IR spec tro-photomet ry of pept ide compounds in water so lu t ions. I l l . Est imat ionof protein secondary structure. Biopolymers 30 : 1273-1280.Ka y C M , Marsh MM . 1959. Some opt ica l p ropert ies o f fe tu in and g luca-gon . Biochim Biophys Acta 33:251-253.Kirschenbaum DM. 1978. Molar absorptivity and A (l cm , I %) values forproteins at selected wavelengths of the ultravio let and isible region s. AnalBiochem 90:309-330.Kuliopulos A, M ildvan AS, Shortle D, Talalay P. 1989. Kinetic and ultr a-violet spectroscopic studiesof active-site mutants of A -3-ketosteroid isom-erase. Biochemistry 28:149- 159.Kupke DW, Dorrier TE. 1978. Protein concentration m easurements: The dryweight. Methods Enzymol48:155-162 .Leach S J, Schera ga HA. 1960 . Effect of light scattering on ultraviolet dif-ference spectra. J A m C h em Soc 82:4790-4792.Lee JC, Timasheff SN. 1974. Partial specific volumes an d interactions withsolvent components of proteins in guanidine hydrochloride. Biochem-istry 13:257-265.Lees CW, Hartley RW. 1966. Studies o n Bacillussubtilis ribonuclease. I l l .Puri f ica t ion and amino ac id com posi t ion. Biochemislry 5:3951-3960.Lesser GJ, Rose GD. 1990. Hydrophobicity of amino acid subgroups in pro-teins. Proteins SIruct Funct Genet 8:6-13..Li YK, Kuliopulos A, Mildvan AS , Talalay P. 1993. Environm ents and mech-anistic roles of the tyrosine residues of A-3-ketosteroid isom erase. io-chemistry 32:1816-1824.Loewenthal R , Sancho J, Fersht AR. 1991. Fluorescence spectrum of bar-pH dependence. Biochemistry 30:6775-6779.nase: Contributions of three tryptophan residues and a histidine-relatedMach H, Middaugh CR, Lewis RV. 1992. Statist ical determination of th eaverage valuesof the extinction coefficients f t ryp tophan and ty rosine

Mach H , Volk in DB, Burke CJ, Midd augh CR . 1995 . Ul t rav io le t absorp-in native proteins. Anal Biochem 200:74-80.t ion spectroscopy. Methods Mol Biol40:91-113.Mart inez-Oyanedel J , Choe HW , Heinem annU , Saenger W. 1991. Ribonu-clease TI w!th free reco gnition and cataly tic site: Crystal struc ture anal-ysis at 1.5 A reso lu t ion . J M o l Biol222:335-352.Minato S , Tagaw a T, Nakanishi K. 1966. Crystallization of ribonuclease T I .J Biochem 59:443-448.Nozaki Y. 1986.Determination of th econcentrationof protein by dryweight-A comparison with spectrophotometric methods. Arch BiochemBiophys 249:437-446.Okajima T , Kawata Y, Ham aguchi K. 1990. Chemical modification of tryp-tophan residues and stability changes in proteins. iochemistry29:9169-9175.Pac e CN . 1966. The reversible denaturationof 0-lactoglobulin A [disserta-t ion] . Durham, North Caro l ina : Duke Univers i ty .Pace CN, Grimsley GR, Barnet t B J. 987 . Puri f ica t ion of ribonuclease TI.Anal Biochem 167:418-422.Perkins SJ. 1986. Protein volum es and hydration effects. The calculationsof partial specific volumes, neutron scattering matchpoints and 280-nmabsorption coefficients for proteins and glycoproteins from amino acidsequences. Eu r J Biochem 157:169-180.Pett igrew DW, Ma DP, Conrad CA, Johnson JR. 1988. Escherichiacoli glyc-erol kinase: Cloning and sequencing of theglpK gene and the primarystructure of the enzyme. J Biol Chem 263:135-139.

6:201-222.


13/13

Protein absorption coefficients 2423Prakash V, Loucheux C, Scheufele S, Gorbunoff MJ, Timasheff SN. 1981.

chem Eiophys 210:455-464.Interactions of proteins with solvent components in 8M urea. Arch Eio-Privalov PL , Tiktopulo El , Venyaminov SY , Griko YV , Makhatadze GI,Khechinashvili NN. 1989. Heat capacity and conformation of proteinsin the denatured state. J M ol Biol205:737-750.Richards FM. 1977. Areas, volumes, packing, and protein structure. AnnuRev E iophys Bioeng 6:151-176.Robbins FM, Kronman MJ. 1964. A simplified method for preparing a-lactal-bumin and 0-lactoglobulin from cows milk. Eiochem Eiophys Acta82:186.

Robinson GW. 1975. Isolation and characterizat ionof papaya peptidase Afrom commercial chymopapain. Biochemistry 14:3695-3700.Runyon GT, Lohman TM. 1989. Escherichia coli helicase 11 (UvrD) proteincan completely unwind fully duplex linear and nicked circular DNA. JBioi Chem 264:17502-17512.Schmid FX. 1989. Spectral methods of characterizing protein conformationand conformational changes. In : Creighton T E , ed. Protein structure:A practicalapproach. Oxford, UK: IRL Press. pp 251-285.

Scopes RK. 1974. Measurement of protein by spectrophotometry at 205 nm.Anal Biochem 59~277-282.Sela M, Anfinsen CB, Harrington W E 1957. The correlation of ribonucle-ase activity with specific aspects of tertiary structure.Eiochim E iophysActa 26:502-512.Shirley BA, Laurents DV. 1990. Purificat ion of recombinant ribonucleaseTI expressed in Escherichia coli. JEiochem Eiophys Methods 0:181-188.Sinclair JF . 1995. Equilibrium and kinetic studies of the folding of the sub-

units of bacterial luciferase [dissertation]. College Station, Texas: TexasA&M University.Solli NJ, Herskovits TT. 1973. Solvent perturbation studies and analysis ofprotein and model compound data in denaturing organic solvents. AnalEiochem 54:370-378.Span J, Lenarcic S, Lapanje S. 1974. Solvation of lysozyme and 0-lacto-globulin in aqueous guanidine hydrochloride solutions. Eiochim EiophysActa 359:311-319.Takahashi K. 1961. The structure and function of ribonuclease T1. I . Chro-matographic purification and properties of ribonuclease T1. J Eiochem49:l-8.Takahashi K. 1962. The structure and function of ribonuclease TI. 11. Fur-ther purification and amino acid compositionof ribonuclease T1. J E i o -chem 51:95-107.

Townend R, Winterbottom RJ, Timasheff SN. 1960. Molecular interactionsin 0-lactoglobulin. 11. Ultracentrifugation and electrophoretic studies ofthe association of &lactoglobulin below its isoelectric point. J Am ChemSoc82:3161-3168.

Voordouw G , Roche RS. 1974. The cooperative binding of two calcium ionsto the double site of apothermolysin. Biochemistry 13:5017-5021.Wetlaufer DB. 1962. Ultraviolet spectra of proteins and amino acids.A d vProtein Chem 17:303-391.White FH. 1961. Regeneration of native secondary and tertiary structures

by air oxidation of reduced ribonuclease. J Biol Chem 236: 1353.Yanari S, Bovey FA. 1960. lnterpretation of the ultraviolet spectral changesof proteins. J Biol Chem 235:2818-2826.Yu Y, Makhatadze GI,Pace CN, Privalov PL . 1994. Energetics of ribonu-clease TI structure. Biochemistry 33:3312-3319.

How to measure and predict the molar absorption coefficient of a protein

Documents

Transcript of How to measure and predict the molar absorption coefficient of a protein