Tryptophan exposure in various conformational isomers of bovine prothrombin fragment 1

Biochimica et Biophysica Acta, 667 ( 1 9 8 1 ) 3 5 - - 4 3

35

© E l s e v i e r / N o r t h - H o l l a n d B i o m e d i c a l Press

BBA 38594

TRYPTOPHAN EXPOSURE IN VARIOUS CONFORMATIONAL ISOMERS OF BOVINE PROTHROMBIN FRAGMENT 1 AN ACRYLAMIDE QUENCHING STUDY

H E N R Y C. M A R S H a, E L L E N M. G E O R G E a, K A R L A. K O E H L E R b a n d R I C H A R D G. H I S K E Y a

a Department of Chemistry and b Department of Pathology, University of North Carolina at Chapel Hill, Chapel Hill, NC 27514 (U.S.A.) (Received July 4th, 1980)

Key words: Prothrombin fragment 1; Tryptophan exposure; Fluorescence spectroscopy; Acrylamide quenching; (Bovine)

Summary In order to assess the importance of a variety of environmental factors on the structure of bovine prothrombin fragment 1, we have examined acrylamide quenching of fragment 1 intrinsic fluorescence. Tryptophan exposure, determined from Stern-Volmer plots, is heterogeneous with one or more of the three fragment 1 tryptophans being exposed to solvent. In the presence of Ca2÷ or Mg2+ even the most accessible tryptophan(s) are relatively buried. Only small differences in tryptophan exposure may exist between fragment 1-Ca 2÷ and fragment 1-Mg2+ complexes. Lowering pH, on the other hand, results in increased tryptophan exposure. Finally, structural isomers of fragment 1 which exist in the absence of metal ions have identical tryptophan exposure as determined by acrylamide quenching and fluorescence intensity.

Introduction The 7-carboxylation of ten specific glutamate residues in prothrombin has been shown to be necessary for cooperative Ca 2÷ binding [1,2] and for Ca 2+mediated binding to phospholipid surfaces by prothrombin. Both cooperative Ca2÷ binding [2--4] and Ca2÷-mediated lipid binding [5] have been localized in the amino-terminal 156 residues of prothrombin, fragment 1. Phospholipid binding has frequently been associated with maximal prothrombin activation rates in vitro. Ca2+ appears unique among metal cations in the magnitude of its ability to mediate binding of prothrombin or fragment 1 to lipid surfaces and Abbreviations: Fragment 1, N-terminal portion of prothrombin released on with thrombin; GdnHCI, guanidine hydrochloride.

treatment

of

prothxombin

36 to support normal prothrombin activation rates. A few other metal ions, particularly some of the trivalent lanthanides, have been shown to partially support lipid binding [6] and prothrombin activation [7]. Characterization of the interactions of Gla-containing proteins with metal ions, and ultimately of the metal ion-containing proteins with phospholipid is central to the understanding of the biologically important functions of these proteins. Metal ion binding to fragment 1 results in the perturbation of a wide variety of protein spectral characteristics [6,8--13]. We [14,15] and others have speculated that these spectral changes are a result of conformation changes in the fragment 1 molecule as a consequence of the binding of metal ions. At equilibrium, in the presence of optimal metal ion concentrations (in terms of the observed spectral change), few of the reported spectroscopic methods can distinguish between the complexes of fragment 1 with calcium or other metal ions. Thus a major focus of the studies reported here is to attempt to define similarities and differences between the Ca 2÷ and Mg2÷ complexes of bovine prothrombin fragment 1. Nelsestuen and co-workers [10,11], Prendergast and Mann [6], and we [16] have characterized the time dependence of the metal ion-induced quenching of the intrinsic fluorescence of bovine prothrombin fragment 1. Furthermore, Nelsestuen and co-workers [10,11] demonstrated that the calcium-induced quenching process is prerequisite to phospholipid binding. These data are readily interpreted in terms of the hypothesis that there exists, in the absence of metal ions, an equilibrium between two forms of bovine fragment 1, one of which can interact rapidly with Ca 2÷ and subsequently with phospholipid. The other form of fragment 1 cannot interact with Ca 2÷ in a manner which yields a phospholipid-binding form of the protein. We have further speculated that the interconversion of these two forms of fragment 1 involves the isomerization of a proline residue [16,17]. Kinetic studies [6,10,11,16] indicate that the two fragment 1 conformational isomers have identical fluorescence properties. Yet, their response to the presence of Ca 2÷ indicates that their conformations are certainly not identical. Thus, a second aim of the studies reported here has been the further evaluation of the structural similarity between the two structural isomers of bovine fragment 1 in the absence of metal ions. Lowering the pH, in the absence of metal ions, also leads to quenching of intrinsic fragment 1 fluorescence [15]. Similarly, lowering the pH increases the apparent a-helix content of fragment 1 to values similar to those achieved in the presence of metal ions [14]. The studies reported here allow comparison of pH-induced and metal ion-induced conformational changes in fragment 1. Solute quenching of intrinsic protein fluorescence can yield information regarding fluorophor exposure. It is a dynamic method in that the presence of quencher alters fluorescent lifetimes, and consequently steady-state intensities, without perturbing the electronic energy levels involved [18]. We have attempted to assess the relative exposure of tryptophan residues in various conformational states of bovine fragment 1, through the use of the neutral quencher, acrylamide. Preliminary studies of Nelsestuen et al. [19] utilizing iodide ion quenching of fragment 1 fluorescence suggest a difference in tryptophan exposure in fragment 1 in the presence and absence of Ca 2÷.

37 Materials and Methods Bovine fragment 1 was prepared essentially b y the m e t h o d of Mann [20]. Samples were prepared for fluorescence experiments b y initial dialysis against EDTA, exhaustive dialysis against the desired buffer, and centrifugation immediately after dialysis. Fluorescence measurements employed a Hitachi/Perkin Elmer MPF-2A fluorescence s p e c t r o p h o t o m e t e r with a temperature-controlled cell compartment. Acrylamide titrations employed a Micro-Metric Instrument Co., Model SB2 syringe microburet. Appropriate corrections were made for sample dilution. Protein concentrations were less than 0.07 mg/ml in order that the absorbance at 290 nm was near 0.05 and inner filter effects due to the protein absorbance could be neglected. Fluorescence intensities were corrected for acrylamide absorbance [21] as follows: Fcorrecte d = Fobsetved (10) -~bc/2

where F is the fluorescence intensity, e is the molar absorptivity of acrylamide, b is cell path-length, and c is the concentration of acrylamide. The molar absorptivity at 290 nm for acrylamide in 0.01 M Tris, 0.1 M KC1, pH 7.5 was determined experimentally on a Cary 17 UV-VIS spectrophotometer to have a value of e290 = 0.567. This value was also used in the studies at lower pH. Buffered solutions of the desired metal chlorides were added to yield a Ca 2÷ or Mg 2+ concentration of 2 mM. Results and Discussion Acrylamide titrations of bovine fragment 1 in the presence and absence of Ca 2÷ yield the Stern-Volmer plots shown in Fig. 1. The plots are ba3ed on the Stern-Volmer equation: Fo/F = i + Ksv[Q]

where Fo is fluorescence intensity in the absence of quencher, F is fluorescence intensity at some concentration of quencher, [Q] is the concentration of quencher, and Ksv is the Stern-Volmer quenching constant. The latter is related to the quenching process b y the equation: Ksv

=

~/~dTo

where kd is the diffusion-controlled bimolecular quenching constant, 7 is the efficiency of the quenching process, and To is the fluorescence lifetime in the absence o f quencher. The measured quenching rate constant, kq = ~'kd = K s v / To, is thus a reflection of quenching efficiency and f l u o r o p h o r accessibility. Eftink and Ghiron [22] have shown acrylamide to quench small indole derivatives predominantly b y a.collisional process. Furthermore, there are no obvious stereochemical requirements and the efficiency of the process was shown to be near unity (7 = 1) in b o t h aqueous and non-aqueous (dioxane) solutions. Therefore, acrylamide quenching of t r y p t o p h a n fluorescence should depend primarily

38

22

1 - -

o

22;

20

o

I

o 181'-

o

~

oo fo

ooo

-i-,6

• z~ ~'

~ ~e

~4~000 ~ 0 •

14

F2~ IOq

o!2

o!4 o!6 o8 --o ACRYLAMtDE CONCENTRATION.M

0 011 0.12 0!3 OJ4 0~5 ACRYLAMIDE CONCENTRATION, M

Fig. 1. S t e r n - V o l m e r p l o t f o r a c r y l a m i d e q u e n c h i n g o f b o v i n e p r o t h r o m b i n f r a g m e n t 1 intrinsic fluor e s c e n c e . T h e s y m b o l s h a v e t h e f o l l o w i n g m e a n i n g s : • f r a g m e n t 1 a l o n e ; o f r a g m e n t 1 in the presence of 2 r a m c a l c i u m ions. F o r details o f e x p e r i m e n t a l o r g a n i z a t i o n see Materials a n d M e t h o d s . Fig. 2. E f f e c t s of p H o n t h e S t e r n - V o l m e r p l o t s folc a c r y l a m i d e q u e n c h i n g o f f r a g m e n t 1. T h e s y m b o l s h a v e t h e f o l l o w i n g m e a n i n g s : o, r u n caxried o u t in 0 . 0 1 M c i t r a t e , 0.1 M p o t a s s i u m c h l o r i d e , p H 4.0; e, run carried o u t in 0.01 M c i t r a t e , 0.1 M p o t a s s i u m c h l o r i d e , p H 6.0; ~, run carried o u t in 0.01 M Tris, 0.1 M p o t a s s i u m c h l o r i d e , p H 7.5. F o r details of e x p e r i m e n t a l o r g a n i z a t i o n see Materials a n d M e t h o d s .

on the collisional frequency. For tryptophan in proteins, the frequency of encounter, kq, since 7 = 1, is a kinetic measure of the exposure of the residue. Stern-Volmer plots have been generated from the data for acrylamide quenching of the intrinsic fluorescence of bovine prothrombin fragment 1 under a variety of conditions. In Fig. 1 Stern-Volmer plots were obtained for acrylamide quenching of fragment 1 fluorescence in the absence of metal ions and in the presence of optimal concentrations of Ca 2+. In Fig. 2 the effect of pH on acrylamide-fragment 1 Stern-Volmer plots is explored from pH 7.5 to pH 4.0. Ksv values were calculated from linear least squares fits to the data obtained at acrylamide concentrations below 60 mM and are collected in Table I. Examination of the Stern-Volmer plots for fragment 1 (Fig. 1) reveals that: (1) the plots are non-linear with negative or concave downward deviation from linearity, and that (2) acrylamide is a considerably more effective quencher in TABLE I T h e K S V values are c a l c u l a t e d f r o m l e a s t - s q u a r e s l i n e a r fit a t [ a c r y l a m i d e ] < 6 0 raM. T h e h q r a n g e s are c o m p u t e d o v e r ~ a n d 7 m f r o m Prendergast et al. [ 2 3 ] , a n d a~sumes t h a t l i f e t i m e s are the same for Ca 2+ a n d Mg 2+ a n d t h a t l i f e t i m e s at p H 4.0 are the s a m e as in 6 M GdnHC1. Additions

Conditions

KSV (M -1 )

k q (range) (M - 1 • s - l ) X 1 0 - 9

None

p H 7.5

Ca 2+

p H 7.5

2.5 - - 4.9 1.8 - - 3.8 0 . 8 4 - - 1.8

Mg 2+

p H 7.5

None None

p H 6.0 p H 4.0

5.5 4.0 1.4 1,4 0.9 0.8 3.9 26

0 . 5 4 - - 1.2 0 . 4 8 - - 1.1 9.0 - - 1 1 . 1

39 the absence of calcium ions than in their presence. Non-linear Stern-Volmer plots typically arise either from static quenching processes (and hence reflect a binding process) or heterogeneity in the t r y p t o p h a n accessibility. Static quenching, the formation of dark, ground state complexes, reduces the fluorescence intensities without altering the excited state lifetimes. Such quenching generally results in positive deviations in Stern-Volmer plots [24]. These considerations as well as the observations of Prendergast et al. [23], have led us to conclude that the observed negative deviations in the fragment 1 Stern-Volmer plots for acrylamide quenching reflect heterogeneity in exposure of the three t r y p t o p h a n residues; that is, variation in kq. Thus, at low quencher concentrations, the more accessible tryptophans, with their higher observed Stern-Volmer constants, Ksv , are preferentially quenched. At higher quencher concentrations, easily quenched fluorescence having been depleted slopes reflect the quenching of the less accessible tryptophans. A n y positive deviations resulting from static quenching processes have apparently been overwhelmed. Prendergast et al. [23] observe differences in the values of intrinsic fragment 1 lifetimes observed b y phase and modulation. As expected in a system containing heterogenous fluorophores, the lifetimes determined by phase lag were shorter than those determined b y relative modulation [25]. The addition of 6 M guanidine hydrochloride tends to equalize the observed differences lending support to such a characterization of the system. A linear approximation to the fluorescence quenching data obtained at low quencher concentration (less than 60 mM) yields average Stern-Volmer constants, Ksv, weighted toward the more accessible fluorophors. Values of Ksv vary from 5 (M -1) in the absence of metal ions at pH 7.5 to values of 1.4 and 1.0 in the presence of calcium and magnesium ions, respectively, at pH 7.5. Eftink and Ghiron [22] have determined Ksv for acetyltryptophanamide in aqueous solution at r o o m temperature to be 17.5 (M-l). On lowering the fragm e n t 1 pH from 7.5 to 6.0 a small decrease in Ksv is apparent, while decreasing the pH to 4.0 results in a Ksv value of 26 (M-l). A range of quenching rate constants, kq, can be calculated over the intrinsic lifetime values (rv to Tm) reported b y Prendergast et al. [23]. Such values represent an approximation in light of the differing conditions of the lifetime and acrylamide quenching experiments (ionic strength, pH, excitation wavelength). These values are collected in Table I. Eftink and Ghiron [24] have determined the acrylamide quenching rate constants, kq, for a number of peptides and proteins both with single and multiple tryptophans. In the former case, values range from a b o u t 4 • 109 M -1 • s-1 for the exposed residues in adrenocorticotropin hormone and glucagon to 2 . 1 0 9 M "~ • s ~ for the buried residues in the human serum albumin-SDS complex. For multitryptophan containing proteins, values range from 2.7 • 109 M -~. s -1 for j3-trypsin in 6.7 M guanidine hydrochloride to 1.0 • 109 M -1 • s -1 for aldolase. The values of kq for bovine fragment 1 at pH 7.5 thus fall within normally observed ranges. In the absence of metal ion, one or more t r y p t o p h a n residues can be characterized as exposed with kq = (2.5m4.9) • 109 M -1 • s -1. In the presence of added Ca 2÷ or Mg 2÷, even the most accessible tryptophan(s) appears relatively buried with kq = (0.48--1.8) • 109 M -1 • s -1. Acrylamide quenching data confirms expectations based on the shifts in

40 I

1

•

I

I J5

rio

fo f

105

1.00 0

ACRYLAMIDE CONCENTRATION, raM_

Fig. 3. S t e r n - V o l m e r p l o t s f o r a c r y l a m i d e q u e n c h i n g of f r a g m e n t 1 in t h e p r e s e n c e of 2.0 m M Ca 2+ (o) or 2.0 m M Mg 2+ (o). E r r o r b a r s a s s u m e a ± 0 . 5 % e r r o r in d e t e r m i n i n g a p a r t i c u l a r f l u o r e s c e n c e i n t e n s i t y .

wavelength of m a x i m u m fluorescence emission (kmax) as previously reported by Marsh et al. [14] and Scott et al. [15], who observed a blue shift in the emission m a x i m u m of about 7--10 nm upon the addition of Ca 2÷ or Mg 2÷. Such observations suggest the movement of the fluorophor(s) from an exposed hydrophilic environment to a more sequestered, hydrophobic, one. Emission wavelength generally reflects microenvironment polarity which is usually, but n o t always, indicative o f solvent accessibility. The use of solute quenching as a direct measure o f fluorophor accessibility eliminates this ambiguity. Differences observed between the Ca 2÷- and Mg2+-induced conformations are within expected experimental error; t h a t is, t r y p t o p h a n exposure is not detectably dif. ferent (Fig. 3). This m a y appear to conflict with the different Xmax values reported. Using the empirical correlation between kq and Rr, ax developed by Eftink and Ghiron [24] a km~x of 330 nm (Mg 2÷) reflects a kQ of 0.4 • 109 M -1 • s -1, a kmax of 334 nm (Ca 2+) reflects a kq of (0.9--1) • 109 M -1 • s -1, a difference in kq of only about 0.5 • 109 M -1 • s -1. In light of heterogeneity in fragment 1 intrinsic lifetimes one is unable to unambiguously demonstrate such a small difference in k~. Intrinsic lifetimes for fragment 1 in the presence of Mg 2÷ have n o t been reported. Using the lifetime values obtained in the presence of Ca ~÷, kq values obtained with Mg 2÷ appear to be less than those with Ca 2÷ by (0.3--0.6). 109 M -1. s-1. The Ksv values for fragment 1 at pH 6.0 and 7.4 appear similar with somewhat higher values obtained at pH 7.5. The lack of intrinsic lifetime data at proper pH values prevents a direct comparison in terms o f kinetic accessibility. The Stern-Volmer constant, Ksv, at pH 4.0, on the other hand, is considerably larger reflecting at least one fully exposed tryptophan. Calculation of quenching rate constants Kq, based on fragment 1 lifetimes at pH 6.8 [23] yields a kq range of (10--20) • 109 M -1 • s-1. The quenching rate for indole is 7.1 • 109 M -1 • s-1 [22] and represents an upper boundary

41 for acrylamide quenching of indole derivatives. If fragment 1 lifetimes in 6 M guanidine hydrochloride are employed, calculated quenching rates, although still somewhat high, approach more plausible values. Thus, fragment 1 tryptophan lifetimes at low pH are probably 2.8 ns or greater and reflect protein denaturation and tryptophan exposure. Both Ca 2÷ and low pH (hydronium ions) have been shown to quench the intrinsic steady state fluorescence of fragment 1 [14]. Acrylamide quenching reveals, however, that the observed fluorescence quenching occurs through distinct protein conformational processes. (1) The quenching by Ca :÷ results from conformational changes in which tryptophan residues become sequestered, probably in a more hydrophobic environment as evidenced by a blue shift in the kmax, and possibly near some protein-bound collisional quencher as evidenced by shorter lifetimes. (2) Quenching of intrinsic fluorescence at low pH results from conformational changes (denaturation) in which tryptophan becomes exposed. The magnitude of the Stern-Volmer constant implicates longer lifetimes and the quenching therefore cannot be attributed to a simple collisional process by protons. Longer lifetimes and reduced steady state intensities suggest either a perturbation in the electronic energy levels resulting in reduced absorbance or a static quenching mechanism through the formation of a dark complex. The existence of a dark complex is further supported by the observation that estimated fragment 1 tryptophan lifetimes at pH 4, based upon the proportionality between quantum yield and lifetime, in the case of collisional quenching, yield unreasonably large values of k d. Alternatively, energy transfer from tyrosine in the folded conformation may result in unexpectedly high tryptophan absorbance which is lost upon unfolding at low pH. In addition to quenching of intrinsic fluorescence, titration of fragment 1 with acrylamide results in blue shifts in the emission km,x. In both the absence and presence of metal ions, the emission km~x at very high acrylamide concentrations (~>1 M) yields an apparent value of 326 nm. This is an apparent value in that a buffer blank has not been subtracted to eliminate contributions from the water Raman peak and comparison with other reported emission km~ values should be made only in light of this difference. The observed blue shift is consistent with preferential quenching of the accessible red fluorescence by acrylamide. Kinetic studies involving bovine fragment 1 are consistent with the following model: + C a 2+

A-~B.

"C

k --1

slow

---Ca2+

fast

S c h e m e I.

where A, B, and C are conformational isomers [16]. The fluorescence intensity of the A and B forms is the same since addition of excess EDTA to the quenched Ca2+-induced C conformation re-establishes fluorescence rapidly. If the fluorescence of A and B were different, the original (no Ca 2+) intensity would be re,established slowly reflecting the approach to equilibrium of A and

42 B with the first order rate of k 1 + k _ 1. Identical fluorescence intensities of A and B suggests that t r y p t o p h a n exposure in the two forms is the same. We attempted, however, to demonstrate a difference in tryptophan exposure in the A and B forms through acrylamide quenching. Fragment 1 was equilibrated in 2 mM Ca 2÷, 0.01 M Tris, 0.1 M KC1, pH 7.5. Acrylamide, 6 M in buffer, was added to a concentration of 0.12 M. The time dependence of the fluorescence intensity at 340 nm was followed upon the addition of EDTA to a concentration of 4 mM. Upon the addition of EDTA, an immediate increase in the fluorescence intensity was observed with no further detectable changes. Preferential quenching of either A or B would lead to slow changes as the B form relaxes to equilibrium with the A form. We therefore conclude that tryptophan exposure in the A and B conformers of fragment 1 is identical. Conclusions From the results reported here it is possible to conclude that tryptophan exposure in bovine prothrombin fragment 1 is heterogeneous with one or more of the three tryptophans of fragment 1 being exposed to solvent. In the presence of Ca 2+ or Mg 2÷ even the most accessible tryptophan(s) are relatively buried. The data indicate that at most only a small difference in tryptophan exposure exists between fragment 1-Ca 2+ and fragment 1-Mg 2÷ complexes. Distinct differences are observed between the effects of calcium ions and hydronium ions on fragment 1 structure. While, as noted above, calcium ions induce conformational changes in fragment 1 which lead to sequestration of tryptophan residues; low pH, in the absence of metal ions, has the effect of further exposing the t r y p t o p h a n residues. Finally, as evidenced by observed fluorescence intensity and acrylamide quenching, t r y p t o p h a n exposure in the isomeric forms o f fragment 1, in the absence of metal ions, appears to be identical. Some of these considerations are summarized in Scheme II, in which the conCo 2 *

A

B

C

IC* Scheme

II.

formational change involved in the isomerization of fragment 1 isomer A into isomer B is shown to have no effect on the environment of the relatively exposed t r y p t o p h a n residue. On the other hand, conversion of form B to C u p o n addition of calcium ions leads to a form of the protein in which the tryptophan residue is buried. Finally, the acrylamide quenching results reported

43 here suggest that in the presence of magnesium ions the fragment 1 tryptophan is possibly more buried than in the presence of calcium ions. The immunological studies of Madar et al. [26] further indicate that conformational differences exist between states C and 'C' when the latter involves Mg :÷, Eu 3÷, Mn :÷, and Gd 3+. Acknowledgement This study was supported in part by National Institutes of Health grants HL20161 and HL-23881. One of us (E.M.G.) was supported through the NSF Undergraduate Research Participation Program. References 1 Stenflo, J. ( 1 9 7 5 ) in P r o t h ~ o m b i n a n d R e l a t e d C o a g u l a t i o n F a c t o r s ( H e m k e r , H.C. a n d V e l t k a m p , J.J., eds.), pp. 1 5 2 - - 1 5 8 , L e i d e n University Press, L e i d e n 2 S t e n f l o , J. a n d G a n t o , P.O. ( 1 9 7 3 ) B i o c h e m . B i o p h y s . Res. C o m m u n . 50, 9 8 - - 1 0 4 3 H e n r i c k s e n , R . A . a n d J a c k s o n , C.M. ( 1 9 7 5 ) A r c h . B i o c h e m . B i o p h y s . 1 7 0 , 1 4 9 - - 1 5 9 4 B e n a r o u s , R., Elion, J. a n d Labie, D. ( 1 9 7 6 ) Biochimie 5 8 , 3 9 1 - - 3 9 4 5 Gitel, S.N., O w e n , W.G., E s m o n , C.T. a n d J a c k s o n , C.M. ( 1 9 7 3 ) Proc. Natl. A c a d . Sci. U.S.A. 70, 1344--1348 6 P r e n d e r g a s t , F . G . a n d M a n n , K.G. ( 1 9 7 7 ) J. Biol. Chem. 2 5 2 , 8 4 0 - - 8 5 0 7 Fttrie, B.C., M a n n , K.G. a n d Furie, B. ( 1 9 7 6 ) J. Biol. C h e m . 2 5 1 , 3 2 3 5 - - 3 2 4 1 8 B l o o m , J.W. a n d M a n n , K.G. ( 1 9 7 8 ) B i o c h e m i s t r y 17, 4 4 3 0 - - 4 4 3 8 9 Fuxle, B.D., B l u m e n s t e i n , M. a n d Furie, B. ( 1 9 7 9 ) J. Biol. C h e m . 2 5 4 , 1 2 5 2 1 - - 1 2 5 3 0 10 Nelsestuen, G.L. ( 1 9 7 6 ) J. Biol. C h e m . 2 5 1 , 5 6 4 8 - - 5 6 5 6 11 Neisestuen, G.L., B r o d e r i u s , M. a n d Martin, G. ( 1 9 7 6 ) J. Biol. C h e m . 2 5 1 , 6 8 3 1 - - 6 8 9 3 12 Brenckel, G.M., Peng, G.-W. a n d J a c k s o n , C.M. ( 1 9 8 0 ) in V i t a m i n K M e t a b o l i s m a n d V i t a m i n K - d e p e n d e n t Proteins (Suttie, J.W., ed.), pp. 5 4 - - 5 7 , University P a r k Press, B a l t i m o r e 13 Carlisle, T.L., Morita, T. a n d J a c k s o n , C.M. ( 1 9 8 0 ) in V i t a m i n K M e t a b o l i s m a n d V i t a m i n K-depend e n t Proteins (Sutti, J.W., ed.), p p . 5 8 - - 6 1 , University P a r k Press, B a l t i m o r e 14 Marsh, H.C., R o b e r t s o n , P., Jr., S c o t t , M.E., Koehler, K.A. a n d Hiskey, R.G. ( 1 9 7 9 ) J. Biol. Chem. 254, 10268--10275 15 S c o t t , M.E., K o e h l e r , K.A. a n d H i s k e y , R . G . ( 1 9 7 9 ) B i o c h e m . J. 1 7 7 , 8 7 9 - - 8 8 6 16 Marsh, H.C., S c o t t , M.E., Hiskey, R . G . a n d K o e h l e r , K.A. ( 1 9 7 9 ) B i o c h e m . J. 1 8 3 , 5 1 3 - - 5 1 7 17 Marsh, H.C., Boggs, N.T., III, R o b e r t s o n , P., Jr., Sarasua, M.M., S c o t t , M.E., t e n K o r t e n a a r , P.B.W., H e l p e r n , J.A., Pedersen, L.G., K o e h l e r , K.A. a n d Hiskey, R.G. ( 1 9 8 0 ) in V i t a m i n K M e t a b o l i s m a n d V i t a m i n K - d e p e n d e n t Proteins (Suttie, J.W., ed.), p p . 1 3 7 - - 1 4 9 , University P a r k Press, B a l t i m o r e 18 L e h r e r , S.S. a n d Leavis, P.C. ( 1 9 7 8 ) M e t h o d s E n z y m o l . 4 9 , 2 2 2 - - 2 3 6 19 Nelsestuen, G.L., Resnick, R.M., K i m , C.S. a n d P l e t c h e r , C. ( 1 9 8 0 ) in V i t a m i n K M e t a b o l i s m a n d V i t a m i n K - d e p e n d e n t P r o t e i n s (Sutti, J.W., ed.), p p . 2 8 - - 3 8 , University P a r k Press, B a l t i m o r e 20 M a n n , K.G. ( 1 9 7 6 ) M e t h o d s E n z y m o l . 45, 1 2 3 - - 1 5 6 21 Parker, C.A. ( 1 9 6 8 ) P h o t o l u m i n e s c e n c e of S o l u t i o n s , pp. 2 2 0 - - 2 2 2 , E i s e v i e r / N o r t h - H o l i a n d , New York 22 E f t i n k , M.R. a n d G h i r o n , C.A. (1'976) J. P h y s . C h e m . 80, 4 8 6 - - 4 9 3 23 P r e n d e r g a s t , F.G., B l o o m , J., D o w n i n g , M.R. a n d Mann, K.G. ( 1 9 8 0 ) in V i t a m i n K M e t a b o l i s m a n d V i t a m i n K - d e p e n d e n t P r o t e i n s (Sutti, J.W., ed.), PP. 3 9 - - 4 8 , University P a r k Press, B a l t i m o r e 24 E f t i n k , M.R. a n d G h i r o n , C.A. ( 1 9 7 6 ) B i o c h e m . 15, 672---680 25 S p e n c e r , R.D. a n d Weber, G. ( 1 9 6 9 ) Arm. N.Y. A c a d , Sci. 1 2 3 , 3 6 1 - - 3 7 6 26 Madar, D.A., Hall, T.J., Reisner, H.M., Hiskey, R.G. a n d K o e h l e r , K.A. ( 1 9 8 0 ) J. Biol. C h e m . 2 5 5 , 8599--8605