Biochemistry
`
`C Copyright
`
`1971 by the American Chemical Society
`
`Volume 10 Number 22
`
`October 26 1971
`
`Differences between Bovine and Human Serum Albumins
`
`Binding Isotherms Optical Rotatory Dispersion Viscosity
`
`Hydrogen Ion Titration and Fluorescence Effects
`
`Jacinto Steinhardt Johanna Krijn and Joan G Leidy
`
`prerequisite
`
`ABSTRACT Human serum albumin HSA offers advantages
`over bovine serum albumin BSA in studying the effects of
`ligand binding to the protein on such optical properties as
`difference spectra fluorescence and optical rotatory dispersion
`ORD because although the compositions of the two proteins
`are very similar human serum albumin HSA contains only
`than two as BSA The bind
`residue rather
`one tryptophan
`ing isotherms of HSA to a number of alkane derivatives were
`to the optical experi
`obtained as a necessary
`ments they are reported here as well as a the effects of bind
`ing on the ORD and viscosity of solutions of HSA and b
`portions of its hydrogen ion titration curve HSA differs only
`in detail from BSA in some of these properties HSA appears
`to unfold in either one stage or in an essentially continuous
`process while BSA has been shown to exist in two distinguish
`able unfolded states at pH 4856 At 686 both appear to un
`fold in a single or mixed process Comparative hydrogen ion
`the number of basic groups
`titration data demonstrate that
`in HSA
`accessible to solvent after exposure to pH 4 is larger
`than in BSA When the fluorescence
`behavior of the two pro
`teins is compared more radical differences appear The lone
`tryptophan of HSA has only about
`threequarters the fluo
`rescence emission as the average per chromophore of the two
`tryptophans of BSA Tyrosine fluorescence
`has an abnormally
`low quantum yield in both proteins but in HSA disorganiza
`
`tion by binding unfolding ligands dodecyl sulfate and
`myristyl sulfate strongly raises the tyrosine emission The
`fluo
`the tryptophan
`two proteins differ strikingly in that
`rescence in HSA is enhanced
`by com
`and in BSA is quenched
`bination with long chain n > 10 alkyl derivatives Binding
`even small amounts shifts the tryptophan
`fluorescence
`spec
`trum to shorter wavelengths in both proteins short chain
`ligands cause the same shifts but produce little or no quench
`ing or enhancement BSA quenching requires binding of only
`1 or 2 equiv but HSA enhancement
`4 and
`requires between
`6 With very large amounts bound either protein is severely
`and tryptophan
`fluorescence is
`disorganized by unfolders
`strongly quenched The relative quenching is the same in both
`and others involving polarization
`proteins This observation
`measurements suggest
`that considerable structure remains in
`the unfolded proteins The fluorescence
`effects of binding to
`in HSA are essentially the
`BSA differ at pH 56 and 686 but
`from tyrosine to tryptophan occurs to a
`same Energy transfer
`greater extent in HSA and is greatest when complexed with un
`classes of binding
`folding ligands It appears that distinct
`some
`in the native proteins the distances between
`sites exist
`of them is inferred as well as the occurrence of rapid diffusion
`of bound ligand between
`and unoccupied binding
`occupied
`sites and the existence of internal bonds which reduce fluo
`rescence in uncomplexed HSA
`
`alkyl
`
`The
`complexing of bovine serum albumin BSA1 with
`ligands including long chain detergents and fatty acids
`in the absorbance of its trypto
`is accompanied by changes
`phan and tyrosine residues Bigelow and Sonnenberg
`1962
`
`of Chemistry Georgetown University
`
`From the Department
`Washington D C 20007 Received February 3 1971 Supported by
`National Science Foundation Grant GB 13391
`used are BSA bovine serum albumin HSA human
`Abbreviations
`serum albumin MRW mean residue weight l moles of ligand bound
`to protein per mole of protein
`
`Polet and Steinhardt 1968 Gallagher et al 1970 Some of
`Si transition of
`these changes are paradoxical
`in that
`the So
`the tryptophan band differs in sign blue shift from the So
`S2 transition of the same chromophore red shift More com
`involving both tyrosine and tryptophan occur
`plex changes
`when the protein is unfolded at f > 10 Since BSA contains
`two tryptophans which may lodge in different environments
`and human serum albumin contains only one further syste
`of these proteins
`matic study of the spectral perturbations
`with the purpose of elucidating the nature of
`the binding
`sites appeared to be more promising if carried out with HSA
`
`BIOCHEMISTRY
`
`VOL 10 NO 22 1 9 7 1
`
`4005
`
`Abraxis EX2057
`Actavis LLC v Abraxis Bioscience LLC
`1PR201701101 1PR201701103 1PR201701104
`
`

`

`HSA
`
`and
`
`BSA
`
`90
`
`80
`
`70
`
`60
`
`950
`
`40
`
`30
`
`20
`
`10
`
`6
`
`5
`
`3
`4
`Log Ceq
`
`2
`
`FIGURE
`
`Effect of binding various detergent sulfates and dodecyl
`sulfonate by 01 HSA at pH 56 phosphate or 686 phosphate
`Tagged symbols indicate lot 24 HSA untagged lot 30
`where noted
`Filled symbols indicate 25° unfilled 2 0 Octyl sulfate
`HSA
`decyl sulfate D dodecylsulfonate CD dodecyl sulfate G
`dodecyl sulfate pH 686 0 myristyl sulfate L BSA lot 8167
`dodecyl sulfate pH 56
`BSA lot 8167 dodecyl sulfate pH 686
`is expressed as a function of log Ce the equilibrium
`Molal ratio t
`of ligand not bound to protein
`
`concentration
`
`rather than with BSA Except for the difference in tryptophans
`and a higher content of valine in HSA the amino acid com
`positions of the two proteins are practically identical
`reports the binding isotherms of HSA at 2
`This paper first
`and 25° at pH 56 with several alkyl
`ligands and a fatty acid
`of the ORD and viscosity on the
`as well as the dependence
`extent and type of the ligand bound Slight differences be
`functions of the two proteins
`tween
`these physicochemical
`have
`been
`including
`differences in their
`titration curves
`found In the later pages much more striking differences in
`the fluorescent properties
`including relative intensity are
`on
`reported as well as opposing effects
`these properties
`quenching
`or enhancement of each chromophore when
`they are complexed with alkyl ligands
`
`Experimental Section
`
`Materials Solutions of crystallized BSA lot 8167 Nutri
`tional Biochemicals or of HSA Pentex lots 24 and 30 were
`deionized on a Dintzis column before use Stock solutions of
`deionized protein were stored at 2° and used within 2 weeks
`Because solutions of dissolved crystallized HSA lot 24 be
`came cloudy at room temperature most stock
`solutions of
`deionized HSA were stored in phosphate buffer 14 = 0033
`at either pH 56 or 686 which kept them clear The concentra
`tions of the solutions were determined spectrophotometrically
`
`2 The only other marked differences
`in composition are in the amide
`of which BSA contains 35 residues
`nitrogen and
`the valine content
`Schultze et al 1962 and HSA 45 Phelps and Putnam 1960 p 143
`that portions of the amino acid sequences
`There is fragmentary evidence
`differ substantially Swaney and Klotz 1970
`
`4006
`
`BIOCHEMISTRY
`
`VOL 10 NO 22 1 9 7 1
`
`STEINHARDT KRIJN AND LEIDY
`
`as previously
`
`described Reynolds et al 1967 using the
`ell 7HSA = 53 and
`following extinction
`coefficients at X279
`Ea BSA = 667 The native BSA contained 10 equiv of fatty
`em
`acidmole of protein and the HSA 027 equivmole when
`measured by a modification Chen 1967 of the method of
`Dole 1956
`The alkyl
`ligands used were a special grade prepared for us
`by Mann Research Their purity has been described earlier
`Reynolds et al 1967 The laurate isotherm was determined
`with the C isotope New England Nuclear
`
`All measurements were made in phosphate buffers reagent
`grade at 0033 ionic strength at either pH 560 or 686
`Methods Binding isotherms were determined for HSA at
`2 and 25 by the equilibrium dialysis method used for BSA
`Reynolds et al 1967 The 14C isotope of laurate was used at
`23 Reynolds et al 1968 and assayed with a scintillation
`center
`
`in ionic
`
`all
`
`ORD measurements were made with a Jasco Model ORD
`UV 5 spectrophotometer
`Viscosity measurements were made in 50 stoke Cannon
`Fenske viscometers with flow times of 200600 sec with photo
`electric timing reproducible to =001 sec Measurements were
`made at 2 and 23 the temperature was controlled to better
`than ±001° All solutions were prepared with filtered Milli
`pore phosphate buffer
`Titration curves were obtained with a Radiometer Titrator
`TTT lc and Titrigraph SBR 2c using 8 ml of 0513
`protein
`solutions in 0033 xi KC1 and adding 0306 M 1ICI or NaOH
`in 0033 m KC1 there was no significant
`change
`strength during the titration since practically
`the
`acid
`added was combined and solutions were originally at the same
`strength No significant
`contribution to the
`ionic
`ionic
`strength was made by the protein Tanford 1961 p 468
`Fluorescence measurements of 01
`protein solutions at
`25° were made with an AmincoBowman spectrophotofluo
`rometer and a HewlettPackard XY recorder Model 7035B
`The sample in a 03 cm silica cell was excited with horizon
`tally polarized light to suppress scatter at 90 due to the excita
`tion beam Chen 1967a A 1P28 photomultiplier
`tube and
`gratings blazed for 300 nm were used with 05mm slits at
`and a 4 mm slit at
`two faces of the cell compartment
`the
`tube With
`this phototube gratings and
`photomultiplier
`the emission spectra were almost completely un
`polarizer
`light output sensitivity or absorp
`distorted by instrumental
`tion Chen 1967a
`The wavelength scales were calibrated
`and corrected except
`case of Figure 5 which uses
`consistent uncalibrated scales for expository purposes only
`Fluctuations
`intensity were minimized
`in the exciting light
`by closely regulating the power supply When necessary the
`photomultiplier gain or recorder amplification was altered to
`keep each set of spectra of the native protein approximately
`
`the
`
`in the
`
`superimposable
`translated to F the
`Equivalents of
`ligands present were
`molal ratio bound on the basis of the isotherms presented in
`isotherms for BSA
`Figure 1 combined with data from other
`lot number Reynolds
`determined earlier at 23° with a different
`et al 1967 Considerable variation in results with different
`V > 20 Anderson 1966 however
`lot numbers can occur at
`our own experiments show that
`this variability occurs sig
`only with ligands which unfold With myristyl
`nificantly
`and the problem does
`sulfate binding is almost quantitative
`
`obtained with a 1 mm slit
`Both excitation and emission spectra
`proved to be indistinguishable from those obtained with a 4 mm slit
`
`

`

`DIFFERENCES BETWEEN BOVINE AND HUMAN SERUM ALBUMINS
`
`co
`
`cr
`
`t 5
`
`Ce
`
`Ce
`
`00
`
`00
`
`CeCe
`
`Cr
`
`CCC
`
`C4
`
`Ce
`
`E c
`
`d
`
`cd
`
`PA
`
`743
`
`a
`
`en
`
`1
`
`1s
`
`a
`
`ts
`
`00
`
`<
`
`m
`m
`N 0000
`
`en co r Lr
`N N r s
`
`0
`
`10
`
`C
`
`00 0
`
`71
`
`r
`rs2ONN0
`O N 00
`
`71 N CP
`6
`
`6
`
`O
`
`en
`
`6
`
`C
`00 00 0
`
`00
`
`fNi
`
`i
`
`1
`
`1
`
`i
`
`1
`
`<1
`t4
`
`g
`
`5 5Ce mcz
`cdc5
`
`6 C4 5 8
`
`5 451
`
`c 1 0 >1 121+1
`
`riD
`
`x c
`
`n
`
`rID
`
`lID
`Pa
`
`lID
`
`z
`
`cn
`
`rn
`
`riD
`
`Ce
`
`X
`
`47
`
`cn
`
`X
`
`riD
`
`CA
`
`CA
`
`riDz
`
`riDx
`
`riD
`
`0 2
`
`1
`
`X 4
`
`4
`
`z
`
`Ce
`
`Ce
`
`riD
`
`0
`
`cd
`
`01
`
`not arise with dodecyl sulfate the other unfolder
`included
`in these experiments a new isotherm for BSA of the lot num
`ber used in this work included in Figure 1 proved to be mea
`surably different at high V from those published earlier Since
`we have also established that
`isotherms obtained with both
`proteins at both 2 and 25° do not differ within the accuracy
`of our measurements we have supplemented the data of Fig
`ure 1 with earlier data on the binding of the nonunfolders
`to BSA as required
`
`Results
`
`HSA Isotherms Portions of binding isotherms of HSA for
`the ligands studied were measured as in our earlier work with
`BSA Ray et al 1966 Reynolds et al 1967 The results of
`such measurements with four sulfate halfesters of varying
`chain lengthfrom C8 to C14 and with one long chain sul
`fonate are shown in Figure 14 For reference purposes binding
`isotherms of BSA lot 8167 with dodecyl sulfate at two pH
`values are included A few experiments at 25° established that
`differences between data obtained at 2 and 25° did not exist
`or were within the experimental error
`Although the binding isotherms for dodecyl sulfate of BSA
`and HSA are very similar with BSA a change
`in pH from
`56 to 686 has a relatively large effect on the unfolding
`region of the isotherms whereas with HSA the change in pH
`has only about half the effect
`With the exception of myristyl C14 sulfate the ligands in
`Figure 1 appear to bind to HSA to about
`the same extent as
`to BSA Myristyl sulfate is remarkable in appearing to unfold
`HSA at much lower equilibrium concentrations
`than with
`BSA Klotz or Scatchard plots of the low 77
`regions native
`the numbers of high affinity sites in
`protein indicate that
`HSA for each ligand are close to those in BSA and most of
`the association constants are usually very nearly the same in
`HSA as in BSA with the shorter chains however
`the associa
`tion constants are only one tenth to one fifth as great Table
`for the numbers of high affinity sites
`I gives the best values
`ligand and the corresponding as
`
`n for each nonunfolding
`sociation constants K in the native protein For comparison
`
`values for BSA Reynolds etal 1967 1970 are also included
`The values for the unfolder dodecyl sulfate were newly deter
`mined for the BSA lot 8167 used in the present study The
`constant Ui in
`for the unfolding
`table also contains values
`the case of unfolding ligands and m and J the numbers of
`sites in unfolded protein and their intrinsic association con
`U2 and L included refer
`stant The additional
`constants
`to
`reaction Decker and Foster
`a postulated second unfolding
`1967 Reynolds et al 1970 The values Ul and U2 are not
`critically determined The value for J in the case of myristyl
`to wide uncertainty
`since only a small
`sulfate is also subject
`part of the unfolded portion of the isotherm was obtained
`
`4
`
`15 values
`
`for BSA in the figures are usually from earlier papers and
`to a 6 70 decrease because 69000 rather than 65000
`are therefore subject
`in their calculation New 15 values for
`was used as the molecular weight
`
`BSA determined in this work are also high by 6 since for consistency
`they were similarly calculated 69000 was correctly used in calculating
`all values for HSA
`5 Data for laurate a fatty acid anion are taken from Goodman 1958
`for I < 6 and from our own data see Methods at v > 6 There is fairly
`6 to 9 between the two sets of
`good agreement in the overlap region 15
`data
`6 In the case of lot 8167 BSA only seven to eight high affinity sites
`for dodecyl sulfate were found in place of the eight
`to nine reported
`earlier for other lot numbers
`
`BIOCHEMISTRY
`
`VOL 10 NO 22 1 9 7 1
`
`4007
`
`

`

`STEINHARDT KRIJN AND LEIDY
`
`5
`
`6
`
`pH
`
`7
`
`FIGURE 4 Titration curves from pH 4 of HSA and BSA in 0033 NA
`KC1 with 0306 N NaOH in 0033 rvi KC1 at 250 Abscissa is negative
`
`from an initial
`M1233 is reduced
`stages is not observed
`value for the uncomplexed protein by nearly the same pro
`portion 57
`as with BSA as E rises from 0 to 10 With
`octyl and decyl sulfates which do not unfold there are no
`further changes With the two unfolders included in Figure 2
`where there was a plateau with BSA beginning at
`17 ca 10
`none is found with HSA
`M133 decreases monotonically
`with no hint of a pause at the level of P = 10 and only levels
`off at P = 60 where for BSA the second stage of change is
`give the same ORD results
`beginning The two unfolders
`The nonunfolders level off at different M233 values
`Viscosity Figure 3 displays the dependence of the reduced
`HSA solutions at 2° and pH 56 on the ex
`viscosity of 02
`to which the protein combines with decyl dodecyl and
`tent
`myristyl sulfates The results with decyl sulfate are similar to
`those found with BSA in experiments conducted at 01
`pro
`tein Reynolds et al 1967 The data obtained with dodecyl
`sulfate differ in three respects from those obtained with BSA
`shown as a broken curve7 a the viscosity goes through a
`minimum at a value of a below 108 b it rises substantially
`at a above 10 and c a further rise between a = 60 and 90
`brings the viscosity up to a plateau 86 where in the Reynolds
`et al data 1967 BSA showed a substantial
`further rise The
`results obtained with myristyl sulfate are in qualitative agree
`ment with those obtained with BSA at
`low 27 numbers the
`viscosities are lower than with dodecyl sulfate
`Titration Curves The titration curves
`shown in Figure 4
`
`is clear that
`
`were obtained by commencing the titration with base after
`bringing identical weight per cent solutions of both proteins
`in 0033 tvi NaCl to pH 40 with HC1 It
`to return
`the pH from 40 to 5560 5 equiv more base are required
`for HSA than for BSA This difference between
`the titra
`remains constant at higher pH values
`tion curves however
`up to at least pH 76 When 30 equiv of dodecyl sulfate are
`the difference between HSA and BSA rises from 5
`present
`equiv to 10 ie HSA gains a still greater buffering capacity
`than BSA under these conditions for incipient unfolding The
`into play by the presence of detergent
`extra groups brought
`have a midrange pK of 65
`
`Preliminary experiments show that when the initial acidi
`fication is much greater ie to pH 26 it
`is BSA rather than
`HSA that
`requires more base in returning to pH 8 Clearly
`the conformation changes accompanying acid unfolding N
`
`7 The results shown for BSA differ at high i
`from those published by
`Reynolds Cr at 1967 in that
`to show a sharp rise in
`lot 8167 fails
`t = 100 and thus resembles HSA at all
`v above about 40
`viscosity near
`low pH 38 and much
`8 A broad minimum is found with BSA only at
`lower ionic strength 0001 Reynolds et al 1970
`
`1
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`1
`
`ORDHSA
`
`65001
`
`4m1233
`
`7000
`
`e Myristyfsulfate
`o Dodecylsulfate
`O Decylsulfate
`o Octylsulfate
`BSA
`
`Dodecylsulfate
`
`80 90 00 10 eo
`
`70
`
`7500
`
`1IIIiIII
`
`f 0 20
`
`30 40 50 60
`
`HSA and BSA complexed
`FIGURE 2 Mean residue rotation of 01
`with various detergent sulfates as a function of molal ratio t at
`Cotton effect trough 233 nm pH 56 ionic strength 0033
`
`the
`
`because of solubility limitations The
`full equation used for
`two stage unfolding will be furnished by the authors on re
`quest
`Dispersion ORD
`Binding Dependent Optical Rotatory
`HSA shows a Cotton effect
`trough at 233 nm similar to that
`of BSA The two proteins have troughs of approximately the
`same depth about 8300° corresponding to a mean residue
`rotation M133 of about 7500° MRW = 114 How
`ever Figure 2 shows
`the reduction in the levorotation of
`HSA resulting from combination with a number of alkyl
`gands differs radically from the variation in the ORD of BSA
`described Reynolds et al 1967 With
`we have previously
`BSA there is an initial
`rises to about 10 almost
`identical effects are pro
`itself as v
`duced by complexes with Cg to C14 sulfates Among the sul
`fonates C8 alone produces a much smaller effect
`in terms of
`M133 at a given V and hexyl derivatives produce no effect
`at all No further changes occur as P rises beyond 10 except
`with the unfolders dodecyl and myristyl sulfates and my
`ristylsulfonate With the latter
`M1233 is greatly diminished
`at E values above about 50
`With HSA the division of changes
`
`that
`
`li
`
`reduction of about 7 due to binding
`
`in M1233 into three
`
`1
`
`I
`
`et al 2°
`
`230
`
`et 011967 230
`
`1
`
`Dodecyl sulfate
`o HSA
`2°
`BSA Reynolds
`0 HSA
`23°
`BSA lot 6167
`
`Myristylsulfate
`A HSA 2°
`BSA Reynolds
`Decylsulfate
`0 HSA 2°
`
`16
`
`14
`
`12
`
`1
`
`V 10
`
`gs 8
`
`0 60
`
`4 2
`
`80 II
`40
`
`20
`
`I
`
`60
`
`80
`
`100
`
`120
`
`140
`
`160
`
`FIGURE 3 Reduced viscosities of 02 HSA complexed with various
`detergent sulfates as a function of molal ratio t at pH 56 phos
`phate ionic strength 0033 2 and 230
`
`4008
`
`BIOCHEMISTRY
`
`VOL 10 NO 22 1 9 7 1
`
`

`

`DIFFERENCES BETWEEN BOVINE AND HUMAN SERUM ALBUMINS
`
`L
`
`16
`
`Excitation
`
`at
`
`285nm
`
`Ui° 14
`
`Ui
`
`fUi
`
`12
`
`Ct0 10
`u a
`
`Ui
`
`2
`
`J L
`
`ii
`et
`
`250
`
`300
`
`FIGURE 6 The emission spectra of BSA and HSA excited at 285 nm
`plotted on the same scale
`
`the spectra ob
`gand This procedure is possible because all
`ie
`tained with excitation above 295 nm are fully congruent
`superimposable by shifts along the wavelength axis after nor
`malization on a model emission spectrum of a protein trypto
`phan analog Nacetyltryptophanamide With HSA however
`the curves depart
`with its large tyrosine emission component
`and the blue shift mea
`slightly from complete congruency
`sured at Xm may be smaller than the blue shift measured at
`the red end after normalization the latter is used in this paper
`The discrepancy is usually under 20 and has been disre
`
`garded in this first
`investigation
`When the excitation wavelengths are <295 nm both tyro
`sine and tryptophan are excited and the emission of both may
`be affected by energy transfer from excited tyrosine to trypto
`phan Analysis of each of the emission spectra
`into tyrosine
`and tryptophan
`components is made possible by making use
`of
`1 The emission spectrum of a tryptophan peptide model
`compound Nacetyltryptophanamide in ethanol
`2 The emission spectra of HSA and BSAligand com
`is excited Xexe > 295 nm
`plexes when only the tryptophan
`between
`in emission
`this permits distinction
`these changes
`at particular wavelengths which are due to quenching or en
`hancement and those changes which are due to a shift
`toward
`the ultraviolet of the entire spectrum
`3 The emission spectra of HSA and BSAligand com
`and the tyrosine residues
`plexes when both the tryptophan
`are excited Xe < 295 nm Since the tryptophan
`contribu
`tion has been determined see 2 the contribution to the emis
`sion due to tyrosine can be obtained by subtraction
`from
`the steps in the procedure which differ
`Details of all
`the matrix method of Weber 19611 to which Figure 5B re
`fers are available on request from the authors
`Intensities Since BSA contains twice
`Relative Fluorescence
`as HSA a large difference in emission
`as much tryptophan
`intensity is to be expected however
`the ratio of intensities
`actually found when the fluorescence of both proteins is plot
`than 2 Figure 6
`ted with the same amplification is larger
`The ratio is close to 27 with either deionized or defatted pro
`tein when the excitation wavelength affects only the trypto
`phan Nevertheless the lifetimes of the excited states 61 and
`
`9 The emission spectrum of Nacetyltryptophanamide
`dissolved in
`is very similar to that of the tryptophan in BSA or HSA when
`ethanol
`is shifted 3 nm to the blue The emission spectrum of N
`the latter
`acetyltyrosinamide is almost independent of solvent
`
`BIOCHEMISTRY
`
`VOL 10 NO 22 1 9 7 1
`
`4009
`
`w
`9
`
`r
`
`ce0
`31
`
`BSA Dodecylsulfate
`
`EXCITATION 300nm
`
`EXCITATION 285nm
`
`BLUE SHIFT
`
`TYlirMIAIF
`
`Y111411i1119111
`
`PROTEIN
`EMISSION
`
`TYROSINE
`
`SLUE SHIFT4m
`
`PROTEIN EMISSION
`
`NORMALIZED MODEL
`CURVE RATIO
`R
`
`SEGMENT OF
`eLUESHIFTED1
`
`PROTEIN CURVE
`
`Ui
`
`PROTEIN
`LIGAND 4e41
`
`350
`A
`
`O30
`
`PROTEIN 1
`LIGAND4q
`
`RATIO r `n
`TRYPTOPHAN
`COMPONENT
`
`B
`
`FIGURE 5 Steps in the determination of the contributions of tyrosine
`and tryptophan to the total emission of BSA and its alkyl complexes
`and of distinguishing between the effects of shifts in km and quench
`ing or enhancement A Excitation at 295 to 305 nm B excitation
`at 280 to 290 nm The steps in the analysis are explained in the text
`
`F transition which may unmask prototropic groups are not
`rapidly reversible in at least one of the two proteins Lovrien
`and Tanford 1959 obtained identical
`titration curves
`1 sec
`and 5 min after mixing BSA with acid but apparently they
`titrate in the reverse direction also The results de
`did not
`to be compared with normal titration
`scribed above
`are not
`curves of these proteins Tanford 1955 Foster and Clark
`1962 Vijai and Foster 1967 since the titrations are preceded
`by an unmasking exposure to acid pH
`Analysis of the Fluorescence Spectra When native uncom
`bined HSA or BSA is excited by radiation with a wavelength
`near 280 nm the fluorescence
`contains emission by both tyro
`side chains Chen 1967b p 443 With
`sine and tryptophan
`BSA the tyrosine contribution is much less conspicuous than
`with HSA Tryptophan
`emission comes not only from side
`but also from
`chains excited by the external
`radiation
`tryptophan side chains excited by radiationless energy trans
`fer from excited tyrosine side chains Chen 1967b However
`when either HSA or BSA is excited by wavelengths between
`295 and 305 nm all of the tryptophan which emits is excited
`radiation tyrosine is not excited and thus
`by the incident
`neither emits nor
`transfers energy Therefore by exciting
`with such wavelengths only a pure tryptophan emission spec
`trum is obtained with the position of the energy maximum
`envi
`ronment When alkyl
`ligands are bound and tryptophan
`only is excited the spectrum changes
`and the entire fluorescence spectrum are blue shifted without
`distortion and b the intensity of
`the emission is either
`enhanced or diminished quenched Since only tryptophan
`in determining to what ex
`is excited there is no difficulty
`the changed
`emission represents a blue shift andor
`tent
`is readily determined from the ratio
`quenching
`The latter
`
`X dependent on the properties of the tryptophan
`in two ways a X
`
`of the intensities at X or at any other homologous part
`of X Since
`
`the emission spectrum The blue shift can be recognized
`of
`by direct
`the maxima are
`determination
`is more accurately determined by mul
`broad the blue shift
`tiplying the emission curve for the proteinligand complex by
`is required to make the maxima equal and
`whatever
`the long wave
`levels of
`then averaging at several different
`length portion of the two curves protein alone and proteinli
`
`factor
`
`

`

`BSA
`
`Xex 305
`
`HSA
`
`STEINHARDT
`BSA
`
`ex
`
`K RIJN AND LEIDY
`HSA
`
`305
`
`3
`
`2
`
`a
`
`300
`
`350
`
`400
`
`300
`
`350
`
`400
`
`300
`
`350
`
`400
`
`300
`
`350
`
`400
`
`290
`
`290
`
`1
`
`3
`
`c 2
`
`3
`
`a 0 U a
`
`4
`2a
`
`LY
`
`I
`
`300
`
`350
`
`400
`
`300
`
`350
`
`400
`
`nm
`FIGURE 7 The fluorescence of BSA and HSA cornplexed with various
`two different
`amounts of octyl sulfate Selected data obtained
`The wavelengths
`and
`excitation wavelengths
`scales are calibrated
`the spectra require practically no correction
`tivity of the measuring system
`
`at
`
`for differential sensi
`
`per
`
`is severely
`
`for
`
`as will be shown when
`59 sec are almost the same However
`HSA binds 4 to <60 equiv of most alkyl
`ligands its fluores
`cence is greatly enhanced The maximum enhanced
`fluores
`cence is almost exactly half of that of uncomplexed BSA so
`fluorescence
`that under
`these conditions average
`efficiency
`the same in the two proteins The
`tryptophan is about
`emission of both unfolded proteins
`tryptophan
`the enhanced HSA is quenched more than
`quenched but
`BSA thus at very high 0 both proteins have about
`the same
`ca 27 as in the uncombined
`ratio of tryptophan fluorescence
`result is considered further in the Dis
`state This unexpected
`cussion
`intensity given by 145 X 101 m BSA on
`The fluorescence
`excitation at 300 nm is very nearly the same as that given by
`3 X 105 m Nacetyltryptophanamide a suitable model
`tryptophan in a protein two tryptophans
`per molecule It
`the tryptophans in BSA
`is therefore tempting to conclude that
`are normal It
`the single tryptophan of HSA
`follows that
`the tyrosines in BSA were also nor
`is partially quenched If
`mal one would expect
`the intensity of the tyrosine emission
`from a 01 70 albumin solution to be similar to that of about
`24 X 104M Nacetyltyrosinamide when excited at 280 nm
`less than this is found in both BSA and HSA
`Considerably
`roughly equal to 1 < 10 m There are very nearly the same
`number of tyrosines in the two proteins and the tyrosine emis
`sion at 290300 nm with excitation at 280 in both appears
`in HSA is substantially
`to be the same although the latter
`higher than in BSA relative to that of tryptophan
`Spectra Some of
`the features of individual
`Fluorescence
`are shown in Figures 7 and 8 The peak
`fluorescence
`spectra
`or shoulder near 305 nm found in the spectra obtained with
`excitation at this wavelength is due to incomplete suppression
`of scattering of the exciting beam It
`is observed to increase
`
`10 This statement does not
`imply that
`the two tryptophans in BSA are the same
`in BSA has been reported
`The quantum efficiency
`of tryptophan
`as 0152 Teale 1960 Excitation was at 280 nm
`
`the fluorescence
`
`intensities of
`
`4010
`
`BIOCHEMISTRY
`
`VOL 10 NO 22 1 9 7 1
`
`300
`
`350
`
`400
`
`300
`
`350
`
`400
`
`X nm
`FIGURE 8 A selection of data obtained with dodecyl sulfate com
`plexes of BSA and HSA similar to those shown for octyl sulfate in
`Figure 7 Since binding large amounts of dodecyl sulfate causes the
`to unfold some of the curves included show the effects on
`proteins
`fluorescence severe tryptophan quenching of such unfolding
`
`when binding occurs or when defatted protein is used A vari
`able amount of aggregation may be responsible Andersson
`1966 The scattering is more prominent in experiments with
`HSA Excitation at shorter wavelengths at 280 285 and 290
`nm rarely shows a detectable shoulder
`the two proteins when octyl sulfate is
`Differences between
`bound are shown in Figure 7 When excited at 305 nm top
`row only tryptophan
`emits and the molecular
`basis of the
`differences is especially clear The BSA tryptophan emission is
`at 0 = 38 Very little additional quenching
`strongly quenched
`to 60 and
`the amount bound increases
`if any occurs
`as
`scarcely more when it rises to 37 nearly 300 equiv are required
`to attain 0 = 37 A progressive blue shift
`is clearly shown
`With HSA there is a blue shift but only a slight degree of
`quenching at o = 33 At larger amounts bound the trypto
`phan emission is enhanced When the exciting beam is at
`290 nm the analysis displayed in Figure 5 shows that
`in BSA
`is blue shifted and quenched With
`the tryptophan component
`HSA the blue shift
`is accompanied by considerable enhance
`ment at r = 22
`With dodecyl sulfate an unfolder Figure 8 a very similar
`situation prevails blue shifts accompany tryptophan enhance
`in HSA quenching in BSA until massive unfolding
`ment
`occurs 0 = 175 in this figure Tryptophan
`emission is then
`in both BSA and HSA upper row but
`strongly quenched
`this strong tryptophan
`quenching is accompanied by great
`in the case of HSA The curve ob
`tyrosine enhancement
`tained with 0 = 175 in the lower righthand panel
`is very simi
`lar to a tyrosine emission curve
`Comparisons with Indicidual Ligands Measurements
`have
`been made with the two proteins at many values of IT and with
`many ligands and the results analyzed into tryptophan
`blue
`shifts and quenching or enhancement
`and tyrosine quench
`ing or enhancement Summaries are presented in Figures 912
`Since o runs from 0 to over 175 and since some of the most
`significant changes occur at low values of 0 0 is represented on
`a logarithmic scale Since IT = 0 cannot be represented on the
`graphs the values for IT = 0 have been indicated by horizontal
`
`

`

`DIFFERENCES BETWEEN BOVINE AND HUMAN SERUM ALBUMINS
`
`DODECYLSULFON
`
`HSA
`
`BSA
`
`Excited
`
`o Excited
`
`285
`300
`
`_ C
`
`cr
`
`r
`
`10
`
`co
`
`0
`18e
`
`8
`
`L1
`
`it
`
`12
`
`0r
`
`0
`
`t Lob
`0
`
`I
`
`3 o
`
`sUi
`06
`
`LT
`
`04
`
`To
`= 02
`0
`
`5
`
`10
`
`15
`
`20
`
`log 0
`
`in HSA due to the
`FIGURE 11 Summary of fluorescence changes
`See Figure 9 figure legend for explana
`binding of dodecylsulfonate
`tion The abrupt drop in tryptophan fluorescence and the sharp rise
`in tyrosine fluorescence are similar to those of the unfolding ligands
`has been shown to be a nonunfolder of
`although dodecylsulfonate
`BSA
`
`t 201 OCTYLSULFATE
`10 I
`
`al
`
`0
`
`118
`s 16
`
`HSABSA
`
`Excited 285
`o Excited 300
`
`er
`
`12
`
`l
`
`I 0 11Lic
`Oa
`14
`04
`
`I
`
`U §
`
`8E
`
`3u
`
`>LLI
`
`E 021
`L
`10
`
`05
`
`10
`
`15
`
`20
`
`log
`
`FIGURE 9 Summary of the effects on the fluorescence of BSA and
`HSA of binding various amounts of octyl sulfate Closed circles rep
`resent BSA and open circles represent HSA The blue shifts top of
`figure were obtained with Xexe close to 300 nm The quenching
`BSA or enhancement HSA of tryptophan is shown for two exci
`ca 300 and ca 285 nm Data obtained with the
`tation wavelengths
`which
`wavelength
`excitation
`excited
`tyrosine as well as
`are denoted by an attached diagonal The data for
`tryptophan
`tyrosine emission were all obtained with Xmax near 285 nm Unlike
`the tryptophan data they are not given on a scale relative to the
`uncombined protein see text
`
`latter
`
`1
`
`in
`
`ledger lines at the values characteristic of the uncombined pro
`tein 0 in the case of the blue shift
`in the case of relative
`tryptophan emission
`Although the results obtained with each ligand differ
`the more important distinctions
`can be displayed by
`detail
`contrasting the data obtained for both proteins with one li
`
`gand that does not unfold octyl sulfate and one that unfolds
`dodecyl sulfate
`a Figure 9 octyl sulfate There is a progressive blue shift
`in both proteins from e = 2
`fluorescence
`of the tryptophan
`is greater in HSA than in BSA The changes
`to almost 20 It
`which occur between 27 = 10 to 20 and e = 40 to 60 are small
`in the range V = 2 to 8 and
`compared to those which occur
`in the case of the unfolders at V over 60
`Even the smallest amounts bound to BSA result
`in strong
`27 exceeds 40
`quenching which does not
`increase further until
`When tyrosine is excited Xe 285 nm the apparent quenching
`
`DODECYLSULFATE
`0156
`pH 686
`o
`o
`Excited
`o Excited
`
`285
`300
`
`HSA
`
`041
`
`20
`
`U2
`
`1 10
`
`o
`2 18
`
`p 16
`
`w 14
`
`3 o
`
`w
`
`12
`
`LO
`
`as
`
`06
`
`3
`
`4 l
`
`a 04
`cc
`I a2
`
`05
`
`10
`
`log
`
`15
`
`20
`
`in HSA brought
`FIGURE 12 Comparison of the fluoresce