Protein-protein interactions in blood serum

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Request accessible transcript

Transcript (if available)

Content PRO TEIN-PROTEIN INTERACTIONS IN BLOOD SIRUM A Dissertation Presented to the Faculty of the Graduate School University of Southern California In Partial Rilfillment of the Requirements for the Degree Doctor of Philosophy by Harry N. Barnet June 1953 F 0 r3io f' - I T h is dissertation, w ritte n by Harry.. N._. Barnet.............. under the d ire ctio n of..IkLs. F a c u lty C o m m itte e , on Studies, and a p p ro v e d by a ll its members, has been presented to and accepted by the C o u n c il on G radu ate S tu d y and Research, in p a r tia l f u l fillm e n t o f requirem ents f o r the degree of D O C T O R O F P H I L O S O P H Y _ .. Committee on Studies .. \ s—y Chairman ACKNOWLEDGMENTS It Is with sincere gratitude and appreciation that I acknowledge the invaluable guidance, encouragement and advice of Dr* Eloise Jameson during my postgraduate train ing and during the course of this investigation* In addition I great fully acknowledge the use of her personal materials and apparatus used throughout this investigation* I would like to express my appreciation to Mr* and Mrs* John J. Elmore for their support of this work* My sincerest appreciation to my wife, Nathalee, for her assistance in the preparation of this manuscript* TABLE OP CONTENTS PAGE HISTORICAL INTRODUCTION...................... 1 STATEMENT OP THE PROBLEM AND PLAN OP ATTACK........ 17 MATERIALS AND METHODS.............................. 18 Electrophoresis Apparatus .................... 18 The cell and cell holding frame.......... 18 Optical system • • • • * ........ • . • • 26 Power source . . ........... 26 Thermostat • • • • . .................... 28 Protein Solutions for Overlay Studies ........ 28 Serum • • • • • • • • .................... 28 Electrophoretically equilibrated serum protein solutions • . • • ..........» • . 28 Fractionated protein solutions • ••••• 29 Supernatant from heat coagulated serum • • 29 Electrophoretic Media • • .................... 30 Ringer’s solution • ••••» ............ 30 Basic Ringer’s solution, pH 8,0 • • • 30 Modified Ringer’s solution, pH 8.0 • • 30 Veronal buffers ••••» ................ 31 Veronal buffer ionic strength 0.11, pH 8.0........................ 31 V PAGE Veronal buffer ionic strength 0.02, pH 8.0........................ 31 Veronal buffer ionic strength 0.2, pH 8.0 • ••»••••••••* 31 Electrophoretic Technique ••••••••••• 32 Overlay technique •«*••••• ........ 32 Preparation of Protein Solutions •••••.•• 36 Electrophoretic separation and equilibration ................... 36 Dilution................................ 36 Dialysis .« ........ ••....».«• 36 Analysis of Protein Solutions «•••••••* 37 pH measurement ••••••••• ........ 37 Protein concentration determinations • • • 38 Copper sulfate specific gravity method •«•••••• ........... 38 Biuret method 38 Reagent 38 Method................ . . . . 39 Conductivity determination ••..•••• 39 Enlargement and Calculation of Patterns • • • • 41 Ultracentrifugal Isolation Technique •••••• 41 EXPERIMENTAL RESULTS.............................. 43 Vi PAGE Investigation of Boundary Anomalies •••••• 44 Formation of epsilon boundaries in veronal buffer ••••••••••»••♦ 45 Formation of epsilon boundaries in RingerT s solution ••••••••••••• 49 Interactions in Artificial Mixtures of Purified Proteins •••• ...... 50 Albumin and beta metal-combining globulin ♦ 50 Albumin and gamma globulin ......... • 51 Interactions Between Fractionated Purified Proteins and Native Serum Proteins 54 Single pure proteins 55 Fractionated mixtures ••••• .......... 55 Interactions Between Electrophoretically Separated Serum Proteins and Native Serum Proteins ............. * * .................. 63 Fractions containing albumin alone or with globulin ........... . . ♦ 63 Fractions containing globulin alone * . . • 68 Studies on the Non-Protein Factors of Serum Contributing to Formation of the "Complex” . . . 68 Experiments with the supernatant from heat coagulated serum ......... • 68 vii PAGE Inorganic elements of the serum as completing factors • • • • ........... 72 Organic elements of the serum as completing factors • ••••••••••. 72 Chemical and Physical Influences on the Formation of the t t Complet” • •••••••••• 73 Ionic strength • ••••••••••••• 73 Dilution and dialysis • •••••••*•» 74 Isolation Etperiments • • • • ......... • • • » 74 Isolation of the X-protein ........ 74 DISCUSSION....................................... 78 SUMMARY AND CONCLUSIONS ........................... 97 BIBLIOGRAPHY........................................102 LIST OP TABLES TABLE PAGE I* Pattern Area Composition and Total Protein Concentration of Upper and Lower Solutions in Experiments Where Cohn Fractionated Proteins Were Overlaid Upon More Concentrated Solutions of the Same • 46 II* Pattern Area Composition and Total Protein Concentration of Upper and Lower Solutions in Experiments Where Single Cohn Fractionated Proteins Were Overlaid Upon Serum •••••• 56 III* Pattern Area Composition and Total Protein Concentration of Upper and Lower Solutions in Experiments Where Mixed Cohn Fractionated Proteins Were Overlaid Upon Serum • *•••• 59 IV. Pattern Area Composition and Total Protein Concentration of Upper and Lower Solutions in Experiments Where Electrophoretically Separated Serum Proteins Were Overlaid on Serum ••••• ............ ••••••••• 64 LIST OF FIGURES FIGURE PAGE 1* Original Microcell Frame and Electrode Support . . . . ............... ♦ ......... 20 2. Modified Microcell Frame and Electrode Support . ......... . • . . . • • 21 3* Magnified View of Modified Microcell Frame • • 23 4* Plastic Microcell Blank • . » . ........... 27 5. Diagrammatic Sketch of Tiered Mierocell Arrangement . • ...................... . . . . 33 6. Diagrammatic Sketch of the Upper and Lower Cells Before and After Electrophoresis of an Albumin Solution Superimposed Upon Undiluted Serum * ..................................... 34 7. Standardization Curve for Albumin Using the Biuret Method............................ • 40 8. Electrophoretic Patterns of Cohn Fraction V Albumin Superimposed Upon More Concentrated Solutions of the Same In Veronal Buffer and Ringer’s Solution pH 8.0 ..••••••••* 47 9. Electrophoretic Patterns of Mixtures of Cohn Fractionated Proteins Superimposed Upon More Concentrated Solutions of the Same in Veronal X FIGURE PAGE Buffer and Ringer1 s Solution pH 8*0 • ••••• 52 10* Electrophoretic Patterns of Solutions of Single Cohn Fractionated Proteins Superimposed Upon Undiluted Serum in Veronal Buffer pH 8.0 ♦ 57 11* Electrophoretic Patterns of Mixtures of Dialyzed Cohn Fractionated Proteins Super imposed Upon Undiluted Serum in Veronal Buffer pH 8.0............................ 61 12. Electrophoretic Patterns of Electro- phoretically Separated Proteins Superimposed Upon Undiluted Serum in Ringerr s Solution pH 8.0 • • • • 66 13. Electrophoretic Patterns Resulting from Superimposing the Gomplexing Factor Alone or with Albumin Upon Undiluted Serum in Ringer’s Solution pH8.0...... ........... 70 14. Electrophoretic Patterns Resulting from Super imposing Electrophoretically Equilibrated Albumin Plus Alpha and Beta Globulins Upon Undiluted Serum in Veronal Buffer at Different Ionic Strengths « • • • • • ............ 75 15. Electrophoretic Patterns of X-Protein Layer • . 77 HISTORICAL INTRODUCTION The role of blood serum proteins in the physio logical activity of the animal body is extremely complex* No less than twenty-nine separate and distinct types of protein molecules have been isolated from human blood serum by recently developed fractionation techniques (1)• Each of these proteins has been found to have different properties and presumably each a different physiological fhnction* With all of this knowledge we do not, however, have as yet the more fundamental insight into the behavior of the molecules as they occur In the natural physiological state, where they are much more difficult to study , particularly as regards the physical-chemical properties* It is not unreasonable to assume that fractionation techniques and other procedures could in the course of chemical and physical treatment considerably alter the isolated protein molecules which are chemically and physically quite labile* It is quite certain that inter actions between simpler protein molecules could result in the formation of complexes which would be dissociated by most fractionation procedures* Therefore, one of our more urgent needs is to devise techniques for studying the physical and chemical properties of proteins in the most 2 natural environment possible. To date these techniques are very few. Many techniques have of course been developed for the study of the physical and chemical properties of proteins of biological fluids in an unnatural environment. The most Important of these, viz., electrophoresis, ultra- centrifugation, osmotic pressure studies, vlscometry, etc., have, however, for the most part employed the use of dilute solutions, buffers and dialysis techniques in an effort to control pH, minimize the effects of Donnan equilibria and eliminate other of the numerous complicating factors. In the field of electrophoresis the generally accepted practice has been that of diluting the protein-containing solution with a suitable buffer to a concentration far less than physiological and then dialyzlng the diluted sample for a long period of time (twelve to forty-eight hours). The original state of the biological fluid or serum has thus been greatly altered, but nevertheless this procedure has yielded much of. the desired information. The development of electrophoresis has led to many important discoveries regarding the physical-chemical constitution and physiological and biological significance of blood serum proteins. As early as 1905 Hardy (2) reported an electrophoretic Investigation of blood serum 3 using the moving boundary technique for studying protein solutions which was first employed by Reuss many years earlier (3)* The experiments of Hardy,which are now classical in the field of protein physical chemistry, provided much of our basic knowledge of the chemical and physical properties of blood serum proteins* Although scattered evidence appeared in the literature subsequent to the pioneer work of Hardy, it was not until 1935 when Tlsellus (4) Introduced the use of low temperature and an improved apparatus that the now great body of electro* phoretic evidence from studies of blood serum began to appear in the literature. Tiselius introduced the idea of conducting electrophoresis at the temperature of maximum density of the medium (salt solution) in order to reduce convection to a minimum. To provide most efficient cooling the impractical cylindrical U-tube was replaced with a rectangular glass cell (described in the section on Methods)# Tlsellus at this same time reported an extensive Investigation of blood serum using this newly developed technique. Although most of his studies were carried out on diluted blood serum, he reported some investigations of undiluted serum. Tlsellus overlaid the undiluted serum with slightly diluted serum and found an unexplainably higher content of gamma globulin in the undiluted serum 4 patterns (compared with the diluted serum pattern) • He proposed that this was due to a neutralization of the charge on the globulin in the undiluted state, the neutralizing substance being dissociated with increasing dilution. Ho farther studies of this phenomenon were ever reported by Tiselius or for that matter any other Investigator until 1947 when Jameson (5) reported a study of the effect of protein concentration on the electrophoretic pattern. Practically all investigators had ascribed the unexplained fractions to artifacts of the system and had failed to find any immobile fraction, since they conducted their analyses with diluted solutions and did not use the method of over laying one so lutlon upon another. Jameson confirmed the earlier findings of Tlsellus and presented much additional evidence regarding the reality of the concentration effect and its relation to the large Immobile fraction* Extending Tlsellus1 earlier work, Jameson was able to study a wide range of dilutions of serum, Including undiluted, 4:1, 3:2 (sometimes 1:1), 2:3 and 1:4 parts of serum to buffer respectively. Patterns were obtained from the upper and lower solutions of a two cell system in which the more dilute solution was superimposed as an overlying solution upon the lower cell containing the more concentrated solu tion. Fov example, two parts of serum diluted with three 5 parts of buffer was overlaid on a solution of serum containing four parts of serum diluted with one part of buffer (designated 3-2/4-1) * It was found upon comparison of the lower solution patterns that the undiluted and only slightly diluted human and rat sera contained the large Immobile fraction, previously mentioned, in addition to the globulins and albumin* Rirther it was discovered that the upper solution patterns contained much more albumin while in these patterns the Immobile fraction had dis appeared* Similarly but to a lesser degree the globulin fractions were found to Increase in the dilute serum patterns at the expense of the immobile fraction when compared to the undiluted or more concentrated serum patterns* The Increases were found to be absolute as measured by the total area under the peaks of the respective fractions* Studying, in addition to serum, the effect of dilution on chicken plasma, Jameson found that after fibrinogen had coagulated, the gamma globulin fraction increased* When corresponding dilution patterns of plasma and serum were compared, no immobile fraction was observed in the serum while the gamma globulin concentrations of these patterns were observed to be much Increased over the concentration of this fraction in the plasma* A few additional studies of the effect of protein 6 concentration and Ionic strength on the electrophoretic distribution pattern of the components of serum have been carried out at concentrations up to four per cent by Svensson (6) and as high as three per cent by Armstrong, et al. (7) and Ferlmann and Kaufman (8). However, in these studies the patterns were obtained by running the serum directly against buffer and not as a superimposed overlying system, and thus they are not strictly comparable to studies of Tiselius or Jameson cited above. Svensson observed an Increase in the area of the albumin fraction with increasing concentrations of serum but his patterns showed no Immobile fraction or any significant changes in the globulin peaks. An extensive study of boundary anomaly phenomena herein reported by Svensson will be discussed in detail later, since many investigators have considered the large immobile fraction merely a boundary anomaly. This necessitates the careful consideration of anomalous salt boundaries in any interpretation to be made. Armstrong and co-workers (7) made an Interesting study of serum and artificial mixtures of purified proteins obtained by the cold alcohol fractionation method of Cohn et al. (1) in an attempt to correlate the fractions obtained with control electrophoretic analyses. A systematic study of the effect (on the patterns) of certain variables 7 Including protein concentration, Ionic strength and specific buffer ion interactions was carried out* It was found by using albumin with only one of the globulins or with fibrinogen as a two component system that the devia tion of the apparent electrophoretic distribution of albumin from the known relative concentration increased with increasing values of the ratio of protein concentra tion to ionic strength* The deviations, the authors cited, were in agreement with those predicted using the theoretical treatment of the subject by Dole (9) and Svensson (6)• A similar study using plasma (at concentrations not greater than two and one-half per cent) showed the same results with respect to albumin distribution with only very small deviations in the other components of the plasma* The effect of specific buffer ions on the distribution pattern was found to be confined primarily to a slight shift in mobilities of certain components resulting in better or poorer separation of peaks* Ihis comprehensive study, however, gave no Information regarding any irregular behavior of the gamma globulin fraction which might suggest the existence of an immobile fraction as seen by Tiselius and Jameson* A prior study to Armstrong et al* (7) by Perlmann and Kaufman (8) dealing with the effect of protein 8 concentration and ionic strength on the electrophoretic pattern obtained distribution data essentially the same as the later investigators and thus yielded little additional information. In addition to this small amount of published electrophoretic data which forms a historical basis for the present problem, a considerable number of unpublished experiments, using the overlying solution method previously described, have been carried out by this Investigator in collaboration with Dr* Jameson as an extension of the earlier work of Jameson (5)* These experiments, which Include extensive dilution studies of normal and patho~ logical human sera as well as several animal species, have yielded strong evidence for the reality of the existence of an immobile fraction In undiluted serum* Most of the latter experiments have been performed with a microcell In the same manner as described earlier (5) using the macro- cell* Eowever, the microcell apparatus made possible the simultaneous running of a complete dilution series using three dilutions of serum run in a three tiered cell (to be described in detail in the section on Methods)• The three solutions routinely used were undiluted serum, serum diluted one part in three with Ringer*s solution and serum diluted two parts in three with Ringer1s solution as the 9 most dilute* Thus the patterns of the most dilute serum were obtained along with the undiluted serum In the same experiment* The insertion of the intermediate dilution (one part in three with Ringer's solution) reduced the large protein concentration gradients which would have otherwise existed between the diluted and undiluted serum and which made pattern interpretation very difficult, as was the ease with the macrocell experiments* In addition it provided another dilution for comparison. Comparisons were made of the patterns obtained from the boundary between Ringer's solution and the most dilute serum (•'diluted* pattern) with those obtained from the boundary between the serum diluted one part in three with Ringer's solution and undiluted serum ("undiluted* pattern)* In normal human sera the large immobile fraction was observed In the patterns of the boundaries of the "undiluted* serum as In the previous studies (4, 5) while In the patterns of the "diluted* sera this fraction was usually entirely missing. The albumin was found to increase in the "diluted" compared to the "undiluted* serum patterns as did the globulin fractions to a lesser degree* When, however, the same comparison was made of patterns using certain patho- logical sera, the Increase was found to be primarily in the alpha and beta globulin fractions with the albumin increase 10 assuming a role of secondary Importance# Some pathological sera did not show much Increase in any fraction by the same method of comparison analysis but were found to have the immobile fraction present in the "diluted” pattern# Studies of animal species have likewise shown an Irregular behavior of the serum on dilution as Indicated by the changes In the components# In addition to the experiments Just described, numerous experiments on the effect of physiological factors, chemo-therapeutic agents and physical influences on the normal serum have been performed# Many of the effects were very striking and have indicated further the real signi ficance of the immobile fraction and its possible Important role In blood physiology and chemistry# The data suggest that the immobile fraction is a protein-protein complex that is dissociated Into its components by dilution# It therefore seemed that what was most needed was a more basic approach to a better understanding of the nature of this fraction for which we have proposed the name "complex”. In view of the fact that a complex consisting of interacting proteins is a possible explanation for the immobile boundary, the discussion would not be complete were it not to include the topic of protein interaction phenomena# The postulated "complex” suggested by the 11 dilution studies probably represents a protein-protein complex structure weakly held together by a metal or other biological polyvalent ion, Van der Waalfs forces, other weak electrostatic long range forces, or a coating of some foreign substance which can be removed by dialysis. Protein-protein interaction Is not new in the field of protein chemistry and was recognized by some of the earliest investigators in the field* The salting out of proteins by neutral salts is In essence dependent upon the phenomenon of protein-protein interaction* The recent development of cold alcohol fractionation of Cohn and co-workers (1) employs the interaction of proteins with zinc ion to effect one phase of their separation* As previously mentioned Armstrong et al* (7), studying artificial mixtures of serum proteins separated by the fractionation technique of Cohn et al* (1), found a considerable degree of protein-protein interaction depending upon the protein concentration, ionic strength, pH, etc* In early phase rule and electrophoretic studies on fractions obtained by the salting out of proteins using potassium citrate, Jameson (10) showed the inter action of serum globulins unequivocally* It was shown that the gamma globulin fractions represented single phases (according to the phase rule) but must be reversible mobile equilibrium systems of the serum globulins* 12 The recent work of Goldwasser and Putnam (11) deals with the Interaction of albumin with nucleic acid. These authors found in their electrophoretic studies a fraction which eould only be accounted for as a complex resulting from the interaction of the protein with the nucleic acid* It should also be noted that these authors cite the particularly interesting fact that interaction of these substances has taken place even though both are charged negatively at the pH studied (pH 5-6)• Equally remarkable evidence for the binding of anions by negatively charged proteins may be found in the literature (12)* Humerous noneleetrophoretlc techniques have been used in the study of the interaction of serum proteins* Recent studies of the osmotic pressure of protein solutions has yielded additional evidence for the interaction of blood serum proteins. Oneley (13) has demonstrated the interaction of beta and gamma globulins by osmotic pressure measurements, as has Scatchard (14) using gelatin polymers* Ultracentrifugal studies by several investigators including Pedersen (15) and McFarlane (16) have suggested interaction of proteins of the serum with a dilution effect similar to that seen in the electrophoretic dilution studies* Artificial mixtures of pure albumin and globulins exhibited changes in relative concentrations on dilution which 13 MeParian© considered to be a reversible equilibrium inde pendent of* pH and nature of the salts present. Pedersen (17) isolated a protein complex by flotation in the ultra centrifuge. This complex which he called the "TC-protein* was composed primarily of albumin and beta globulin. The entire field of immunological specificity is of course intimately concerned with protein-protein inter action. Many reactions of antigen with antibody involve the interaction and binding of protein to protein (antibody to the antigen) in a highly specific manner based on charge, bond angles and stereochemical configuration of some type as described by Pressman and Siegel (18), a protein-protein interaction phenomenon of great importance. Much information regarding the nature of the ionic charge and distribution of charge on proteins has been gained in recent years through studies of protein-ion interaction. Extensive investigations on the binding of inorganic ions to proteins by Klotz (19, 20) and others (21) has led to a basic insight into the numbers of anionic and cationic binding sites on the molecule and considerable information regarding the binding energy of these ions. Morawetz and Hughes (22) and Klotz et al. (12) have studied the binding of organic anions including synthetic organic polyelectrolytes such as polyvinyl and polyacrylic acid 14 derivatives and dyes such as methyl red or azo dyes, leading to an understanding of polyion interaction with proteins. A biological reaction intimately associated with ion binding and in certain instances protein-protein interaction (as for example, in the hydrolysis of the peptide bond by proteolytic enzymes) may be found in enzyme activity. In this case there could conceivably be a protein-ion-protein complex linkage of a type involving both Ionic binding and protein electrostatic or hydrogen bonding. Since, as previously mentioned, the proposed "complex* boundary is found in the same region as the false or anomalous salt boundary, a discussion of this phenomenon is obviously necessary. Tiselius (23) and Tiselius and Kabat (24) first discovered the false resting or slowly moving boundary on the ascending limb of the cell. This boundary arising from the charge on the colloid and called the delta boundary, has a counterpart In the descending limb called the epsilon boundary and similarly arises as a result of the charge on the colloid. Discovered by Longsworth and Maclnnes (25) , the epsilon anomaly, which like the delta anomaly is Immobile or only very slightly mobile, is always much smaller than the delta. For this reason the descending side was chosen for the present study 15 of the “complex" existing in the same region, Svensson^ extensive treatment of the theory of true and false moving boundaries (6, 26) presented considerable experimental evidence to substantiate the theory. A similar treatment of the theory of moving boundary systems formed by strong electrolytes published concurrently by Dole (9) provided important corroboration of the theory, Svensson stated that the anomalies can be made to disappear or become sufficiently small provided the ratio of salt (buffer Ion) to protein concentration Is high enough. He studied the effect of systematic variations of anions and cations, and ionic strength of buffers with various dilu tions of serum and pure proteins or mixtures thereof conducted at various Ionic strengths. In a study using serum diluted with an equal volume of buffer, the highest protein concentration used by him, Svensson found that upon increasing the salt concentration (NaCl) of his buffer over the range of 0.01M to 0.37M there was a progressive decrease In the delta boundary to a minimum value of five per cent at the highest salt concentration used. Ho data were given for the epsilon boundary in this study but other studies by the same investigator at lower protein concentrations indicate that the epsilon anomaly would be very small or negligible at the higher concentrations of salt (probably 16 0.15M or greater). It may be estimated from his data that at a salt concentration equivalent to Ringer’s solution (0.156M as used in the unpublished dilution studies described above) there should exist a delta gradient representing approximately ten per cent of the pattern area* This estimate may be misleading, however, since higher protein concentrations might behave quite differ ently, as well as the fact that anomalous boundaries arising from the electrophoresis of a protein solution superimposed upon another by overlaying have never been considered sufficiently, and remain completely undefined with respect to formation of false boundaries* Other investigators (25, 27) have likewise considered boundary anomalies but all studies have been at low protein concen tration and thus provide little additional evidence. The observations reviewed above should serve to orient the reader In the field most pertinent to the proposed problem below, namely, the phenomenon of protein- protein interaction* In addition it is hoped that the importance of the problem in relation to the role of proteins In many biological phenomena may be realized. STATEMENT OP THE PROBLEM AND PLAN OF ATTACK In the course of the Investigations of undiluted blood serum (dilution studies) described In the previous section, It became increasingly clear that further insight into the nature of the immobile fraction of the undiluted serum was needed* Since the mobility of the immobile fraction or “complex” is such that it lies in the region of the starting boundary where an anomalous boundary would be located if present and since the “complex" only exists in the undiluted serum where it cannot be isolated by conventional electrophoretic procedures, it was highly important that a true boundary be proven to exist repre senting a real colloid gradient. It was the object of this study to demonstrate that the “complex® does exist, and to investigate some of the factors responsible for its existence* Studies were made using the method of superimposing one protein solution upon another as a descending electro phoretic system. This technique was employed with various protein solutions and native serum under various physical and chemical conditions* In addition, studies of a protein complex, isolated by ultracentrifugation, were carried out to obtain supporting evidence. MATERIALS AND METHODS Electrophoresis Apparatus The cell and cell holding frame * The chosen method of study of the proposed complex was based upon the pre liminary investigations involving electrophoretic dilution studies described in the Introduction* Since these studies employ an apparatus of an unconventional nature, it is necessary to describe it somewhat in detail* One phase of the problem undertaken involved the electrophoresis of two solutions simultaneously, one of which is superimposed upon the other as an overlying solution* This type of experiment requires the use of a tiered cell arrangement as in the dilution studies discussed. The choiee of microcells for general use in place of the macrocell was made for several reasons* First, the microcell is more adaptable to work with higher colloid concentrations because the decreased length of light path through the cell permits the formation of an optical image with lower well separated and more easily interpretable peaks at the higher concentrations* Second, the amounts of solutions required, both colloid and buffer or other media, are only a fraction of those required in the macrocell* Third, the ease of handling and the lower cost of each cell is a great advantage* The microcell 19 is, however, unfortunately so constructed that the width of the cell flange in relation to the height of the cell is much less compared to either short or long macrocells# When used as a tier arrangement, there is a great tendency for separation of the flanges while filling and during shifting of cells to form boundaries. It was, therefore, necessary to construct a cell holding frame which could provide the required flange interface stability* As a starting point for the modification of the apparatus, there was available a standard micro-electro phoretic cell support, Figure 1, obtained from the Klett Manufacturing Company. This consisted of a stainless steel frame (A); a cell holding frame (B) mounted on two cross members (C) of adjustable height on the frame legs (L); brackets for holding the electrode vessels (D) and two geared plunger devices of adjustable height (P) for displacing the cell* These general features of the original frame are those which were altered or modified to make the apparatus adaptable to the use of one, two or three cells at a time, as shown in Figure 2* The original frame bottom support, Figure 1, consisted of two horizontal shelf plates (H) which were adjustable on the cross members (C) for leveling. However, this was Inadequate since It was extremely difficult to 2 0 FIGURE 1 ORIGINAL MICROCELL FRAME AND ELECTRODE SUPPORT 21 FIGURE 2 MODIFIED MICROCELL ERAME AND ELECTRODE SUPPORT 22 get both sides level and free from tilt* To eliminate this, the bottom plate was replaced with a single machined piece of aTuminum metal (h) , Figure 3 (magnified view of cell frame, Figure 2), having a well for a rubber gasket and the bottom piece of the cell* This arrangement proved superior to the original arrangement, as it Is more stable and easier to level by the adjusting screws (hs) . The side posts (p), Figure 1, were extended to the proper height (pr), Figure 3, for three cells. These constitute the vertical support for the accessory parts required to stabilize the cell at each flange level. The front flange guides (g) were constructed of two parts. The reason for this is that the cell flanges may not be identical and therefore permit a fixed system. The rear flange guides (f), Figure 3, were made so that they remain in a fixed position and are of a permanent nature. As may be seen, these guides consist of a flat piece of aluminum plate fastened to the rear side posts (pf) with screws. Two accessary spring clamps (ac) fastened to front flange guides (f), Figure 3, provided as part of the original apparatus (ac), Figure 1, were used during filling to help stabilize the cell but were removed prior to placing the apparatus in the thermostat. Thus, the bracing system for controlling front to rear tilt parallel to the optical = n r R E * __ L A S S — 'VV FIGURE 3 MAGNIFIED VIEW OF MODIFIED MICROCELL FRAME 24 axis of the system was complete. However, as mentioned previously, the ratio of flange width to cell height was of such nature that any possible leverage whatsoever in either direction tended to break apart the cell at the flanges and, therefore, the tilt, away from the optical axis vertically, had to be eliminated. To achieve this, It was necessary to devise a side bracing, which presented a very difficult problem because of the necessity of shifting cells to make the boundaries* The only way of easily attaining horizontal tilt stability was to use a downward vertical thrust by means of a pressure system via springs (s) fastened to the top front and rear flange guide members, the horizontal part of the spring resting on the flange of the top piece. Set screws (t) were arranged for the adjusting of the vertical downward thrust and hence stability. The springs were made from spring steel and tempered to proper hardness by oil bath quenching. It was also necessary to again minimize any effect of tilting to the side due to an uneven horizontal thrust as encountered in the rack and pinion plunger system (rp) , Figure 3. The substitution of a vertical plate (vp), with recessed edges on the top and bottom to fit between the flanges, for the plunger head (ph) served to provide horizontal thrust at the flange edges of each cell. This eliminated the 25 possibility of breaking the vertical glass cell wall. These plates (vp) have been found useful in adding stabil ity when placed in position against the flange during the filling of the cell. Trial runs using the apparatus as modified above showed the arrangement to be satisfactory in all respects except for the use of the top cell piece (tp) of the type shown in Figure 1, which was found to be inadequate. It was very difficult, using rubber tubing, to connect this type of a top piece with the electrode vessels, and thus was an additional point In the procedure tending to cause leaks. In addition, this type of cell which is open at the top must be fitted with rubber tubing to prevent any splash of water from the water bath. There Is also danger of overflow and electrical leaks around the top, A special top was therefore designed so that the thrust for insertion of the rubber connecting tubes (rt), Figure 2, is all downward and thus reduces horizontal tilting effects. This top piece was made from the standard microcell top piece supplied by the Pyrocell Manufacturing Company, The ends of the top piece connecting tubes (ct), Figure 3, originally equipped with ball joint connections, had to be removed and rubber pressure tubing substituted as the connecting link with the equilibrating tubes (et), Figure 2, 26 This arrangement affords great flexibility and ease of adjustment of electrode vessels* Should one choose to run a single cell experiment for comparison with the conventional electrophoretic analyses found in the literature, a blank simulated cell piece shown In Figure 4 may be substituted for the cells removed. The top Is reeessed to hold the bottom cell of the apparatus* All cells used were of the standard two cc. micro cell apparatus type supplied by either Pyrocell or Klett Manufacturing Companies. Optical system* All patterns were recorded on thirty-five millimeter Eastman Linagraph Ortho 420 safety film using the Longsworth schlieren scanning optical system (28)• This system was mounted on adjustable shafts attached to the optical bench such that special adjust ments could be made when in certain instances the third or top cell section had to be observed. Power source* Power source for the electrophoretic runs was a full wave rectified power supply with an approximate rating of fifty milliamps at 500 volts, using a Sola line voltage stabilizer. Current fluctuations under standard conditions did not exceed 0.2 milliamps (10-30 milliamp range)• 27 FIGURE 4 PLASTIC MICROCELL BLANK 28 Thermostat* The free electrophoresis experiments (overlay experiments) were conducted in a water thermostat at a temperature of 1~2°C. Protein Solutions for Overlay Studies Serum* Human serum was obtained through individual blood donations and from discarded pooled serum of the U. S. C. Student Health Service and Bio-Science Labora tories of Culver City, California, courtesy of Dr. Orville J. Golub and Dr. Milton Segalove. All serum samples were kept refrigerated and used as soon as possible. No serum was used which was more than five days old or that had any indication of cloudiness due to bacterial decomposition or other aging process. No preservative was used on any serum due to the possible effect of a foreign agent. If possible no hemolyzed serum was used. Specific treatment of the serum is described in the section on Experimental Results. Electrophoretically equilibrated serum protein solutions. Solutions of serum proteins or mixtures thereof were obtained by electrophoretic separation of the serum into the desired buffer or Ringer1s solution (see section on Experimental Results). Using the multitiered cell, ease of separation was effected through disalignment of the section or sections containing the separated protein 29 fraction or mixture from either ascending or descending limbs* Equilibrated solutions of albumin; albumin plus alpha and beta globulin (sometimes plus gamma globulin); gamma globulin; and gamma, beta and alpha globulins were obtained in this manner in the required medium* Fractionated protein solutions* Electrophoretically equilibrated solutions of purified proteins were obtained from solutions of the pure proteins or mixtures in the same manner as with serum* Cohn Fractions used as a source of these proteins were generously donated by the Cutter Laboratories of Berkeley, California and Dr* W. B. Wallace of the Los Angeles County Hospital. The composition of the fractions used may be found in the Tables included in the section on Experimental Results, as the composition of the overlying solution when used alone* Non-electrophoretically equilibrated protein solutions were obtained by dissolving the dry Cohn Fractions directly in the desired buffer with subsequent adjustment of the pH and dialyzed for at least twelve hours with continuous stirring. Diluted aliquots were made when necessary prior to dialysis* Supernatant from heat coagulated serum* Supernatant liquid was obtained from coagulum of normal human serum heated fifteen minutes in boiling water* Entrained fluid 30 was expressed from the protein coagulum using a syringe equipped with a gauze filter pad to prevent clogging of the nozzle. The centrifuged clear solution was kept refrig erated until used, at which time the pH was adjusted to 8.0 with 0.1N HC1 (pH prior to the adjustment ca. 9.5). One non-dialyzable protein component having the mobility of an alpha globulin and which may be a mucoprotein is shown in the electrophoretic pattern of this material (see section on Experimental Results). The Inorganic components of this solution were obtained by ashing of the supernatant liquid with sulfuric acid in a porcelain erucible. The white residue (approximately 15 mg. per ml. of solution) containing a small amount of carbonaceous material was heated at less than dull red heat to prevent metal sulfate decomposition. Electrophoretic Media Ringer* a solution. 3-* Basic Ringer* s solution, pH 8.0. This solu tion was prepared by adding 8.60 grams NaCl, 0.30 grams KC1 and 0.44 grams CaClg^SHgO to one liter of distilled water, making up to volume after adjusting to pH 8.0 with N NaOH. The ionic strength was 0.156M* 2. Modified Ringer1s solution, pH 8.0. Several 31 experimental modifications of the basic Ringer solution were made as indicated in the section on Experimental Results. Eor preparation of these solutions the modifying material was added directly to the basic Ringer solution with subsequent adjustment of pH if necessary to 8.0. Veronal buffers. Veronal buffer ionic strength 0.11, pH 8.0. This buffer contained 4.12 gm. sodium diethyl barbiturate, 4.64 gm. NaCl and 0.44 gm. CaCl2*2H20 per liter of solution. The buffer was adjusted to pH 8.0 with N HC1 prior to making up to volume in distilled water. 2. Veronal buffer ionic strength 0.02, pH 8.0. This buffer was the same as the veronal buffer ionic strength 0.11 except for the sodium chloride which was left out entirely. 3* Veronal buffer ionic strength 0.2, pH 8.0. This buffer was the same in all respects as veronal buffer ionic strength 0.11 except that two times as much sodium chloride was added. 32 Electrophoretic Technique Overlay technique. The overlay studies were carried out in the multitiered cell, A typical overlay experi mental setup is represented diagrammatically in Figure 5, and is included for purposes of orienting the reader with respect to the dynamic aspects of superimposing one solution upon another. This orientation is very important as will be found later in interpreting the electrophoretic patterns which form the bulk of the data. Filling of the cell consists of arranging the protein solutions so that the bottom piece compartment and the descending limb of the lowest center section (III), Figure 5, contains the serum or protein solution of higher concentration while the overlying solution is placed in the same limb of the mid center section (II), Figure 5, Ringer’s solution or buffer is placed in the ascending limb of all three sections as well as the top section (I), Figure 5, of the descending limb. Having filled the cell in the conventional manner (29) using the disalignment technique, alignment of the center sections as shown in the diagram of the two lower sections of the descending limb (a), Figure 6, results in the formation of two initial descending colloid boundaries, the ascending boundary being disregarded. One descending boundary representing that of 2 JSCEND/MG TOP CELL DESCEA/D/A/G M/D CELL E B O T r O M CEI L UPPER SOLUT/OA/ J Z SEPUM OR LOWER SOLUT/OM HGURE S DIAGRAMMATIC SKETCH G P TIERED MICROCELL A R R A N G E M E N T 34 C START/A/6 BOUNDARY ALBUMIN STARTING BOUNDARY SERUM OE5CEA/D/A/G S/DE BEEORE ELECTROPHORESIS SOON OAR Y ALB. C Y P < ALB ALB. FROM TOP CEIL r-a S A M E S ID E A P T E R ELEC TR O PH O R E S/S a FIGURE 6 DIAGRAMMATIC SKETCH CP THE UPPER AND LOWER CELLS BEFORE AND AFTER ELECTROPHORESIS OF AN ALBUMIN SOLUTION SUPERIMPOSED UPON UNDILUTED SERUM 35 the overlying solution is formed in the mid descending compartment (II), Figure 5, between the electrolyte medium of the top section (I), Figure 5, and the protein solution of the mid compartment (II), Figure 5, A second descending colloid boundary is formed In the lower descending compartment (III), Figure 5, between the solution of this compartment and the overlying solution in the mid cell* The lower solution boundary represents a boundary formed between two protein solutions and is thus unique In that the resulting pattern obtained after migration and separation of the proteins across this boundary will be dependent upon the difference in concentration between the protein in the underlying solution and Its identical counterpart In the overlying solution. The schematic diagram (b), Figure 6, showing the boundaries present in both overlying and lower solutions at the end of migration, represents that distribution which would result were a solution of albumin alone (upper cell) overlaid on un diluted serum (lower cell)* Thus if the concentration of the albumin in the overlying solution is identical with the concentration of the albumin in the serum, no gradient of albumin would be present In the serum (lower cell) and the albumin band would be absent (assuming the albumins to be identical molecules)• Similarly, the presence of any 56 globulin in the overlying solution would diminish the apparent concentration or concentration gradient of its counterpart in the serum proportionately with its concen tration difference between the two solutions* Considerable emphasis is placed on this concept due to the extreme difficulty of interpretation of patterns to be discussed* Filling of the cell was facilitated through the use of a special syringe needle made from a length of stainless steel tubing of sufficient length to reach all cell tier levels* The flange lubricant found most suitable was undiluted white vaseline which is solid at room tempera ture* Any less viscous grease was found inadequate to prevent flange separation in the tiered cell* Preparation of Protein Solutions Electrophoretio separation and equilibration* This technique is described above in the discussion of protein solutions* All electrophoretically equilibrated solutions were used directly without dialysis unless indicated* Dilution* Where dilution of a protein solution is indicated, it was done by direct addition of the diluent with or without dialysis as indicated* Dialysis* Dialysis of protein solution against large volumes of Ringer1s solution or buffer was carried 37 out using Visking sausage casing tied tightly with minimal air space# The resulting bag containing the protein solution was then bound to one end of a glass rod which served as the shaft of an air stirrer. The stirrer was then mounted with a liter Erlenmeyer flask, containing the dialysi s medium, in the thermostat water bath and the bag then slowly rotated in the dialysis medium for at least twelve hours. This arrangement which served as an excellent stirring device, at the same time effected rapid equilibrium of the solutions at a low temperature. Dialysis of the overlying solution against the underlying solution or serum was carried out using the same membrane as above with a marble or glass bead added to the bag. The bag containing either overlying or underlying solution was placed in a glass test tube of a slightly greater diameter than the bag and a small (approximately equal) volume of the other solution was introduced into the test tube. The test tube was then stoppered tightly with a rubber stopper and the system shaken frequently for at least twelve hours, keeping the tube refrigerated* Analysis of Protein Solutlons pH measurement. All pH measurements were made with a glass electrode pH meter, Beckman Model G - type. Headings of the solutions of the upper and lower compartments of the 38 descending limb (overlying and underlying solutions) were taken both prior to tilling the cell and after making the run to check for any major pH changes. The range of pH for upper and lower solutions in unbuffered media at the end of electrophoresis was approximately 7.7 - 8.1 while that of buffered media was 7.9 - 8.1* Protein concentration determinations. Copper sulfate specific gravity method. Some determinations of undiluted serum protein concentrations were made using the copper sulfate specific gravity method of Phillips et al. (30). This method was checked by comparison with the biuret method described below and found to check to within to.2 per cent. 2. Biuret method. The biuret method of Kingsley (31) was used for all solutions except the few serum determinations by the method above. Procedure for the biuret method was as follows: a. Reagent. The biuret reagent consisted of a solution of 1.1 per cent cupric chloride in 30.0 per cent sodium hydroxide (w/v)# 39 b. Method* (1) An aliquot of sample was diluted with distilled water to a concen tration of 0.2 to 0.6 per cent (2-6 mg. per ml.). (2) An equal volume of reagent and sample were then mixed thoroughly and allowed to stand thirty minutes before reading. (3) Solutions were read on either a Klett colorimeter or Beckman model DU Spectrophotometer at 540 m/4. (4) Standard solutions of Cohn Fraction V albumin (approximately 98# pure) were run simultaneously for com parison. A standard curve is shown in Figure 7. Standard solutions used were checked by both micro- Kjeldahl and dry weight methods, which agreed to within 0.01#. Conductivity determination. Conductivity measure ment was made with a Leeds and Northrup conductivity cell and resistance bridge calibrated with standard potassium chloride solutions. All measurements were made at the same 4© S 4 5 2 1 © 0 5 1© 15 Mg* Album in per m l* FIGURE 7 STANDARDIZATION CURVE T O ALBUMIN USING THE BIURET M E T H O D Na. 6 1 D 5 , U n i v e r s i t y B o o k s t o r e , L n s A n d e l e s 41 temperature as that of the electrophoretic runs (1-2°C). Enlargement and Calculation of Patterns Patterns recorded on thirty-five millimeter film negative were projected and enlarged approximately five times and a tracing made of the enlarged projection on graph paper* Areas were calculated by triangulation in preference to the use of a planimeter because the height of the peaks are not adapted well to planimetry* Ultracentrifugal Isolation Technique The ultracentrifugal preparations were obtained using a Spinco preparative ultracentrifuge* The procedure was the technique of Pedersen (17) for the preparation of the X-protein but modified by using a sucrose layer at the top of the serum* This permitted the X-protein complex layer to be completely separated from the serum by levita tion through the sucrose and avoided any possible contami nation of the X-protein by the protein components of the serum* The sucrose layer consisted of a solution of greater density than the X-protein (sucrose sp* gr* 1*08; X-protein sp* gr* 1.04)* Rotor speed was 40,000 rpm. for a period of five hours. Approximate centrifugal force was 100,000 times gravity at the center of the centrifuge tube* The top layer containing the X-protein complex was drawn 42 off using a square end syringe needle* EXPERIMENTAL RESQLTS In presenting the tabular data for the experiments to be described below, both the areas of the components in the upper or superimposed solution as well as the component areas of the underlying solution or serum, as the case may be, are included. The necessity for presenting both data lies in the unique situation which exists in the technique of superimposing one protein solution upon another, TJpon observation of the data obtained from the patterns of the underlying solution, it must be remembered that the upper solution in all Instances contained at least one of the protein fractions present in the underlying solution or serum. If the protein concentration of the fraction In the solution lying above Is at the identical concentration with its corresponding fraction In the underlying solution, no peak representing that fraction is to be expected In the pattern of the lower solution, assuming no difference In mobility of the two boundaries. Accordingly with Increas ing differences In the concentration of the same protein fraction between the upper and lower solutions (that Is, decreasing the concentration of the fraction In the upper solution relative to the lower) the areas representing these differences Increase, These facts must be taken into 44 account In any interpretation to be given the data* Patterns are prints made from enlarged tracings re-photographed. Every attempt was made to make these patterns as accurate as possible. In every case the pattern presented Is typical of two or more separate runs under identical conditions. Investigation of Boundary Anomalies The prediction of the extent of false boundary formation based on the Svensson (6) or Dole (9) theories or the more recent extension of these theories to weak elec trolytes by Alberty (32) has been considered by this investigator. However, after careful examination of these theories, it was decided that no Immediate practical solution could be obtained without considerably more Information as to the type, nature, concentration, pH, equivalent weight, conductance, etc. of the ionic species present. Not all this information is possible to obtain in light of our present day knowledge of proteins and protein solutions. Rirthermore, in order to obtain some Informa tion, certain undesirable conditions such as dialysis, buffer salts, etc. would have to be imposed upon the system. Since, as previously mentioned, the proposed ^complex" lies in the same region as the epsilon or false boundary, it was imperative that conditions for minimal 45 anomalous boundary be established experimentally. To achieve this, it seemed most logical to use a system in which no interaction would probably occur under the condi tions to be used in the overlay studies. For these studies, albumin seemed the most practical and was, in the studies to be described, assumed not to interact* In this phase of the investigation solutions of purified albumin (Cohn Fraction V) at one concentration were overlaid on a more concentrated solution of the same fraction under various experimental conditions. Table I lists the pattern area data for the patterns seen in Figure 8. All patterns shown are representative patterns of a given type of experiment and have in most cases been repeated several times for confirmation. The pattern of the overlying solution is shown on the left hand side of each pair of patterns while that of the underlying solution is on the right as is the convention to be used for all overlay experiments to be described* Formation of epsilon boundaries in veronal buffer* Pattern (d), Figure 8, shows the results from overlaying a solution of Cohn Fraction V albumin dissolved in veronal buffer pH 8.0, ionic strength 0.1 on a more concentrated solution of the same, both solutions having been dialyzed against the same buffer. Concentrations of the overlying 46 TABLE I PATTERN AREA COMPOSITION AND TOTAL PROTEIN CONCENTRATION OP UPPER AND LOWER SOLUTIONS IN EXPERIMENTS WHERE COHN FRACTIONATED PROTEINS WERE OVERLAID UPON MORE CONCENTRATED SOLUTIONS OF THE SAME Exp. No. Medium Fractions in Upper and Lower Solutions Upper Solution . Component Areasi/ Total Protein Cone. mi. . Lower Solution. , Component Areasi/ Total Protein Cone. m % Alb. Beta Gamma € Alb. Beta Gamma Complex plus € A455 Ringer Cohn V, Uy 1467 283 4.1 900 600$/ 7.3 A456 Veronal Cohn V, !&/ 1695 184 3.4 400 125 5.5 A459 Ringer Cohn V(ES)£/, U 1145 82 3.0 718 44 4.5 A460 Ringer Cohn V(ES), U 1403 0 4.0 1145 0 5.9 A474 Ringer Cohn V(ES), D 1000 82 550 202 A476 Veronal Cohn V(ES), D 1100 187 630 133 A475 Ringer Cohn V+IV- 7-2D(ES), U 1382 447 104 3.7 521 96/186(X)£/ 230 5.8 A433 Veronal Cohn V+IV-7-2D, D 1360 350 205 4.5 624 138 270 6.8 A454 Veronal Cohn V+IV-7-2D, D 1473 363 167 4.0 570 84/79 (X) 228 5.6 A477 Ringer Cohn V+II(ES), U 847 481 3.0 362 169 89 5.0 A432 Veronal Cohn V+II, D 1426 515 4.0 862 262 285 7.0 A451 Veronal Cohn V+II, D 1392 503 4.7 761 168 314 7.2 1/ Arbitrary Units. 2/ Undialyzed. 3/ An epsilon boundary since no Interaction of albumin with Itself Is assumed. 4/ Dialyzed. 5/ This Fraction V was electrophoretlcally separated prior to running. 6/ The (X) denotes the extra "X* fraction in this run. Area of "X* is the denominator of the fraction (see Figure 9a)• 47 FT GORE 8 ELECTROPHORETIC PATTERNS OF COHN FRACTION V ALBUMIN SUPER IMPOSED UPON MORE CONCENTRATED SOLUTIONS OF THE SAME IN VERONAL BUFFER AND RINGER’S SOLUTION pH 8.0 (a) Electrophoretically equilibrated In Ringer’s solution, pH 8.0, undialyzed. (b) Electrophoretically equilibrated in Ringer’s solution, pH 8.0, dialyzed. (c) Dissolved In Ringer’s solution, pH 8.0, undialyzed. (d) Dissolved In Veronal buffer, pH 8.0, dialyzed. (e) Electrophoretically equilibrated in Veronal buffer, pH 8.0, dialyzed. ZL If £L Li A m EXPERIMENT A 4 6 0 EXPERIMENT EXPERIMENT A4ES c f/ surc a EXPERIMENT EXPERIMENT A 474 i A4SC A 476 e 00 49 and lower solutions were 3.5 and 5.2 per cent respectively. Epsilon boundaries of a size comparable to those reported in the literature are seen. When the same buffer is used but the albumin solution is first electrophoretically equilibrated against the buffer as described above (section on Materials and Methods) and then dialyzed against the buffer, an epsilon boundary of approximately the same size (slightly smaller) is observed (e), Figure 8. Range of area of the epsilon boundary in arbitrary units as listed in Table I is 125 to 200 units for this series. Fbrmation of epsilon boundaries in Ringer1 s solution. To determine the effect of RingerT s solution upon the formation of epsilon boundaries, a series of experiments were performed using the non-buffered medium of Ringer1s solution together with Cohn Fraction V albumin. The result of electrophoretically equilibrating a solution of the albumin into Ringer1 s solution and then overlaying a diluted aliquot (four per cent) of this solution on the original more concentrated solution (six per cent) is shown in pattern (a), Figure 8. The maximum anomalous boundary in either solution is only 80 units, while in some patterns not shown the epsilon boundary is scarcely visible, by far the smallest anomaly to be observed. Dialysis of the upper and lower solutions of the electrophoretically equilibrated albumin used to obtain (a), Figure 8, resulted in the pattern (b), Figure 8, containing appreciable anomalous boundaries as expected* Areas of these boundaries were of the same order of magnitude as those in (e) , Figure 8. A third type of experiment consisted of dissolving the albumin directly in Ringer’s solution, adjusting the pH and using it in an overlay experiment without equilibration or dialysis. The pattern (c), Figure 8, is the result of imposing such conditions and as may be seen, leads to very large anomalous boundaries in both solutions, particularly the lower solution. Experiments with pure beta metal-combining globulin (Cohn Fraction IV-7-2D) and gamma globulin (Cohn Fraction II) under the conditions of electrophoretic equilibration without dialysis yielded patterns containing no appreciable epsilon boundary in either upper or lower solutions. These patterns are not shown as they were merely run for control purposes to demonstrate the fact that absence of the anomalous boundaries under these conditions was non specific for albumin. Interactions in Artificial Mixtures of Purified Proteins. Albumin Mid beta metal-comblnlng globulin. Repre- sentative patterns resulting from overlaying artificial mixtures of albumin (Cohn Fraction V) and beta metal- 51 combining globulin (Cohn Fraction IV-7-2D) on more concen trated solutions of the same under various conditions are shown in patterns (a) and (b), Figure 9* Table I includes the data for this series of overlay studies. When mixed electrophoretically equilibrated proteins (albumin 75 per cent and beta globulin 25 per cent) in Ringer* s solution are overlaid by a diluted aliquot of the mixture, the pattern (a), Figure 9, is obtained. The probable contributions of the “complex” and of the epsilon boundary to the area of 230 units will be discussed in detail in the general discussion of results. The ”X” boundary which could conceivably represent a “complex” but might be either part of the beta globulin or an anomalous boundary, will similarly be considered later. A typical pattern (b), Figure 9, results from the overlay studies of mixtures of the same proteins which were subjected only to dialysis after dissolving in veronal buffer pH 8.0, ionic strength 0.1. A significant “complex” plus epsilon boundary is observed under these conditions but probably in this instance, it represents a large amount of epsilon boundary together with only a slight amount of “complex", if any. Albumin and gamma globulin. The interaction of albumin and gamma globulin was next studied using Cohn 52 FIGURE 9 ELECTROPHORETIC PATTERNS OF MIXTURES OF COHN FRACTIONATED PROTEINS SUPERIMPOSED UPON MORE CONCENTRATED SOLUTIONS OF THE SAME IN VERONAL BUFFER AND RINGER’S SOLUTION pH 8*0 (a) Cohn Fractions V plus IV-7-2D electrophoret ically equilibrated in Ringer1s solution, pH 8.0, undialyzed, (b) Cohn Fractions V plus IV-7-2D dissolved In veronal buffer, pH 8*0, dialyzed, (c) Cohn Fractions V plus II electrophoretically equilibrated in Ringer*s solution, pH 8,0, undialyzed, (d) Cohn Fractions V plus II dissolved in veronal buffer, pH 8.0, dialyzed. 53 I* O i FIGURE 54 factions II and V* Representative patterns (c) and (d) , Figure 9, result from overlaying artificial mixtures of albumin (Cohn Fraction V) and gamma globulin (Cohn fraction II) on more concentrated solutions of the same under conditions as described Immediately above for albumin and beta globulin. Table I Includes the data for this series of overlay studies, Electrophoretically equilibrated mixtures of these proteins in RingerTs solution upon overlaying, yielded the typical pattern (c), Figure 9, containing the very small "complex” plus epsilon boundary as an immobile shoulder on the gamma fraction. This quite obviously represents little if any interaction, contrary to the finding In the case of beta globulin and albumin. The same mixture when dialyzed against veronal buffer pH 8,0, ionic strength 0.1, yielded a pattern (d), figure 9, with an Immobile boundary area only slightly larger than the size expected In veronal buffer (expected 200 units, observed 280 units). This could represent the inclusion of a small "complex" but it is doubtful that It does. Interactions Between Fractionated Purified Proteins and Native Serum Proteins In a series of studies on the Interaction of 55 fractionated proteins with the proteins of serum in their native state, solutions of single pure proteins and/or mixed subfractions or mixtures of both obtained by Cohn fractionation techniques were treated as described for "fractionated protein solutions" (section on Materials and Methods) and overlaid on undiluted serum* All protein solutions of this series were dialyzed against veronal buffer pH 8.0, ionic strength 0.1 as were the serum samples which were the underlying solutions in each instance. Single pure proteins. Table II lists the data for this series which includes pure albumin, beta metal- combining globulin and gamma globulin solutions overlaid on undiluted human serum. Figure 10 shows typical patterns obtained using respectively albumin (Cohn Fraction V), beta globulin (Cohn Fraction IV-7-2D) and gamma globulin (Cohn Fraction II) at two concentrations as the overlying solutions. No significant degree of interaction was observed in any case except with albumin (a), Figure 10, with a "complex" plus epsilon boundary of 310 and 318 units in each of two separate runs, which may represent slight interaction. Fractionated mixtures. Table III gives the data for this series of experiments using solutions of mixed subfractions with or without added pure protein fractions TABLE II 56 PATTERN AREA COMPOSITION AND TOTAE PROTEIN CONCENTRATION OP UPPER AND LOWER SOLUTIONS IN EXPERIMENTS WHERE SINGLE COHN FRACTIONATED PROTEINS WERE OVERLAID UPON SERUM Exp. No. Medium Fractions In Upper Solutions Upper Solution . Component Areasi/ Total Protein Cone. . - m _______ Serum Component Areasi/ Total Protein Cone. m . Alb. Alpha Beta Gamma Alb. Alpha Beta Gamma Complex plus e A355 Veronal Elect. Sep*d.2/, d2/ 703 590 104 325 327 189 A373 Veronal Cohn V, D 608 2.8 530 214 483 417 318 6.6 A374 Veronal Cohn V, D 890 3.6 245 123 584 395 310 6.6 A 404 Veronal Cohn IV-1, D 258 90 0.6 1427 56 253 614 7.5 A415 Veronal Cohn IV-1, D 100 35 1620 345 652 6.3 A 387 Veronal Cohn IV-7-2D, D 637 1.5 1163 317 119 A401 Veronal Cohn II, D 746 1.2 1958 444 266 7.4 A413 Veronal Cohn II, D 596 1.2 1729 215 221 6.3 A414 Veronal Cohn II, D 287 0.8 1163 134 269 174 6.3 1/ Arbitrary Units. 2/ This albumin was electrophoretically separated from serum. 3/ Dialyzed. 57 PTGURE 10 ELECTROPHORETIC PATTERNS OP SOLUTIONS OP SINGLE COHN PRACTIONATED PROTEINS SUPERIMPOSED UPON UNDILUTED SERUM IN VERONAL BUFFER pH 8,0 (a) Upper solution - Cohn Fraction V albumin* (b) Upper solution - Cohn Fraction IV-7-2D beta metal-combining globulin* (c) Upper solution - Cohn Fraction II gamma globulin. (d) Upper solution - Cohn Fraction II gamma globulin* 58 £ X P £ R t M £ A /T £ X P £ R /M £ N T £ X P £ R 1 M £ N T £ X P £ R !M £ A /r A 374 A 387 A 4/3 A4/4- F !G U R £ /O 59 TABLE III PATTERN AREA COMPOSITION AND TOTAL PROTEIN CONCENTRATION OP UPPER AND LOWER S O L U T I O N S IN EXPERIMENTS WHERE M I X E D COHN FRACTIONATED PROTEINS WERE OVERLAID UPON SERUM E x p . N o . Medium fractions in Upper , Solutions!/ Upper Solution. Component Areas*/ Total Protein Cone . . . m A .. Serum Component Areas2/ Total Protein Cone. mi Alb. Alpha Beta Gamma Alb. Alpha Beta Gamma Complex plus € A388 Veronal Cohn V+IV-7 1 4 0 4 3 2 7 4 . 5 1 0 8 2 8 0 2 2 3 4 9 7 7 . 5 A 3 8 9 Veronal Cohn V+IV-7 1 0 5 8 1 5 8 4 . 8 2 8 6 3 6 2 3 3 2 3 7 7 8 . 0 A 3 9 0 Veronal Cohn V+IV-7^ 5 0 7 1 2 8 1 . 8 8 2 6 4 2 1 4 2 2 6 2 . 8 A 4 0 5 Veronal Cohn V+IV-4 737 94 9 0 3 . 0 681 1 37 282 356 368 7.5 A 4 0 6 Veronal Cohn IV-4 472 374 324 3 . 2 841 190 2 36 7.5 A 3 9 3 Veronal Cohn V + I I 1 5 5 0 5 7 1 3 3 1 4 3 4 7 . 4 A 3 9 6 Veronal Cohn V + I I 9 8 5 2 8 4 3 . 4 6 6 1 4 3 3 3 3 8 3 9 7 7 . 4 A 4 0 8 Veronal Cohn V + I I 1 0 4 7 5 3 5 4 . 5 3 5 9 6 8 4 0 8 1 / All solutions dialyzed against veronal b u f f e r . 2 / Arbitrary u n i t s . 3 / Both upper solution and serum were diluted two parts in three with b u f f e r . 60 and pure protein mixtures as the overlying solutions. Pattern (a), Figure 11, was obtained by overlaying serum with a mixture of albumin (Cohn Fraction V) and beta metal-combining globulin (Cohn Fraction IV-7-2D)• The large "complex” plus epsilon boundary (some 377 units) probably Includes a considerable amount of "complex” component representing a significant degree of interaction. This is as would be expected from the other studies with albumin and beta globulin mixtures (see previous section on pure protein overlays). Pattern (b), Figure 11, represents the series of overlay experiments In which albumin (Cohn Fraction V) and gamma globulin (Cohn Fraction II) were overlaid on serum, yielding a typical pattern with a "complex" plus epsilon boundary of some 400 pattern units. This result was unexpected on the basis of the lack of Interaction in the pure protein over pure protein studies (c) and (d) , Figure 9, but does not preclude the possibility of an Interaction under these conditions due to the presence of serum. When as shown in (c) , Figure 11, either albumin (Cohn Fraction V) plus subfraction IV-4 (a mixture of albumin, alpha and beta globulins, see Table III) or fraction IV-4 alone (not illustrated) were overlaid on 61 PI GORE 11 ELECTROPHORETIC PATTERNS OP MIXTURES OP DIALYZED COHN FRACTIONATED PROTEINS SUPERIMPOSED UPON UNDILUTED SERUM IN VERONAL BUFFER pH 8.0 (a) Upper solution - Cohn Fractions V plus IV-7-2D, albumin plus beta metal-combining globulin* (b) Upper solution - Cohn Fractions V plus II, albumin plus gamma globulin. (c) Upper solution - Cohn Fractions V plus IV-4, albumin plus alpha and beta globulins. u 62 A E X P E R /M EA/r A 3 8 9 E X P E R /M E N T A 3 9 6 E X P E R /M E A /T A 4-O S F IG U R E // 63 serum, the "complex* plus epsilon boundary indicated a definite "complex* present (area 368 units in the figure illustrated (c), Figure 11, with albumin plus IV-4, but only 236 in the pattern not shown, using a lower albumin concentration, see Table III, Experiment A406)♦ Two experiments were attempted using subfraction IV-1 but this fraction proved impractical for use due to Its high degree of Insolubility. Interactions Between Electrophoretically Separated Serum Proteins and Native Serum Proteins These series of overlay experiments were performed with solutions of proteins obtained by the electrophoretic separation of protein fractions from undiluted serum into Ringer’s solution and in every case Involved the direct overlaying of the separated fraction or fractions on the original serum without dialysis except where indicated. Results may be found in Table IV. Fractions containing albumin alone or with globulin. Figure 12 shows the patterns resulting when albumin and albumin plus alpha and beta globulins with or without gamma globulin are overlaid on undiluted serum. A large "complex" plus epsilon boundary Is seen when albumin alone is the overlying solution (a), Figure 12, as Is the case with albumin plus the globulins (b) , Figure 12, where the 64 TABLE IT PATTERN AREA COMPOSITION HD TOTAL PROTEIH CONCENTRATION QP UPPER AHD LOWER SOLUTIONS IN EXPMMEN78 WHERE ELEMSOPHOREHCALEr SEPARATED SERUM PROTEINS ME OVERLAID ON SERUM E x p . N o * Medium Treatment Upper Solution . Component A r e a s ! / T o t a l Protein C o n e , m i Serum Component A r e a s ! / T o t a l Protein Cone* m i A l b . Alpha Beta Gamma A l b . Alpha Beta Gamma Complex p l u s € A 3 3 9 Ringer Undialyzed 655 782 590 566 623 A 3 6 7 Ringer D i a l y z e d 6 0 7 9 0 2 2 3 9 3 3 2 5 8 0 A 3 4 2 Ringer Undialyzed 541 93 1 339 %&! A 3 4 1 Ringer Undialyzed 1 1 7 4 3 3 1 1 9 4 3 8 7 3 2 4 A 3 5 8 Ringer Undialyzed 876 184 162 Tfa^ 100 343 790 A 4 6 1 Ringer Undialyzed 1 0 4 2 9 7 1 0 8 3 * 4 4 7 6 T h 6 4 1 9 5 2 7 . 2 A 3 6 8 Ringer D i a l y z e d 886 1 4 5 1 0 4 2 , 7 9 7 4 3 3 1 8 6 3 3 6 221 5 . 6 A 4 6 3 Ringer D i a l y z e d 1 0 5 0 1 3 4 7 9 2 * 7 9 4 1 3 0 2 3 3 3 8 8 2 9 8 5 . 5 A 3 4 0 Ringer Undialyzed 6 2 163 283 1 7 3 7 2 6 7 2 5 6 2 1 6 A 3 4 4 Ringer U n d i a l y z e d 2 2 6 3 9 5 4 3 5 9 6 9 5 8 9 1 8 4 A 4 6 6 Ringer Undialyzed 243 288 607 1.8 1 9 2 5 1 5 9 2 8 6 3 4 6 7 . 5 A 4 8 6 Ringer Undialyzed y ( 1 8 0 ) 1 4 8 0 T h 7 9 0 7 0 0 2 / A 4 8 7 Ringer Undialyzed y 1 8 5 0 2 7 8 5 1 5 7 2 5 A 5 0 3 Ringer Undialyzed 7 0 2 ^ / 8 3 9 1 4 0 4 7 0 5 6 6 5 5 0 A 5 0 4 Ringer Undialyzed 78 i! 1 0 6 7 3 1 2 5 0 0 1 8 0 65 TABLE IV (continued) PATTERN AREA COMPOSITION AND TOTAL PROTEIN CONCENTRATION OP UPPER AND LOWER SOLUTIONS IN EXPERIMENTS WHERE ELECTROPHORETIC ALLY SEPARATED SERUM PROTEINS WERE OVERLAID ON SERUM 1/ Arbitrary units* 2/ Includes the complex, which was very poorly separated* 3/ Th Indicates thread, uncalculated. 4/ This upper solution consisted of the supernatant from heat coagulated serum and contained an unknown fraction with an alpha mobility. 5/ Area of “X** peak adjacent to albumin in (a) , Figure 13.. 6/ RingerTs solution used as the upper solution. Control to A486. 7/ Albumin plus supernatant from heat coagulated serum. 8/ Albumin In Ringer’s solution was used in this experiment as a control to experiment A503. The same result was obtained with dialyzed super natant from heat coagulated serum plus albumin as the upper solution* 66 FIGURE 12 ELECTROPHORETIC PATTERNS OP ELECTROPHORETIC ALLY SEPARATED PROTEINS SUPERIMPOSED UPON UNDILUTED SERUM IN RINGER’S SOLUTION pH 8.0 (a) Upper solution - albumin, undialyzed. (b) Upper solution - albumin plus alpha and beta globulins, undialyzed. (c) Upper solution - albumin plus alpha and beta globulins, dialyzed. (d) Upper solution - gamma, beta and alpha globulins alone, undialyzed. (e) Upper solution - Ringer’s solution, undialyzed. EXPER IM EN T A 3 3 9 a EXPERIMENT , 4 46/ b EXPERIMENT A sea c FIGURE IE EXPERIMENT A 3 4 4 d EXPERIMENT A 46 7 e o> 68 largest immobile boundary of the entire study was observed (920, 790 units)• When, however, the latter systems (overlying solution and serum) were first dialyzed against Ringer’s solution, a marked reduction in the size of the Immobile boundary occured as shown In the pattern (c), Figure 12, with a concurrent rise In the area of the albumin fraction (see Table IV). Fractions containing globulin alone. Only one fraction of globulin was used in this series of experiments# Its composition, Table IV, consisted of alpha, beta and gamma globulin. Overlaid on undiluted serum this fraction yielded a typical pattern (d), Figure 12, showing no Immobile fraction whatsoever. The pattern (e), Figure 12, is that which results when Ringer1s solution Is used as the upper solution and serum, the lower solution (control)• Studies on the Non-Protein Factors of Serum Contributing To Formation of the Complex Experiments with the supernatant from heat coagulated serum. At this point of experimentation a careful examination of the data was made and It was tentatively concluded that in addition to the required protein components of the ncomplex”, namely albumin and a globulin, some other dialyzable factor must have been 69 present* One of the many possible methods of approach to determine the nature of the dialyzable component was to coagulate the serum by heat and if the unknown factor was heat stabile, study the supernatant (used interchangeably throughout the text with the term '•mother liquor") from the coagulum. To test whether the mother liquor was active, i*e., still contained the unknown agent, an untreated sample (except for pH adjustment) containing added albumin (electrophoretically equilibrated) was overlaid on undiluted serum in Ringerfs solution pH 8*0. The result of this experiment shown by the pattern (c), Figure 13, revealed the presence of a large neomplext t plus epsilon boundary indicating a large amount of interaction* The control to this experiment (a) , Figure 13, is the result obtained when undialyzed mother liquor from the heat coagulation is overlaid on serum. A pattern showing no immobile boundary, but containing a large unknown fraction "X* adjacent to the albumin with a mobility approximately equal to that of alpha globulin was observed* Implications of this pattern will be discussed in a later section but it should be noted that if interaction were to take place under these conditions, this would be the logical position for a "complex1 * to occur, since albumin is present from this point down coexisting with the other 70 FI GORE 13 ELECTROPHORETIC PATTERNS RESULTING PROM SUPERIMPOSING THE COMPLETING FACTOR ALONE OR WITH ALBUMIN UPON UNDILUTED SERUM IN RINGER1 S SOLUTION pH 8.0 (a) Upper solution - undialyzed supernatant from heat coagulated serum. An identical pattern was obtained using undialyzed Ringer’s solution with added glycine. (b) Upper solution - undialyzed Ringer’s solution. (c) Upper solution - undialyzed supernatant from heat coagulated serum plus equilibrated Cohn Fraction V albumin. An Identical pattern was obtained using an undialyzed upper solution of glycine added to Ringer’s solution contain ing equilibrated albumin. (d) Upper solution - equilibrated albumin in Ringer’s solution, undialyzed. An identical pattern was obtained using the dialyzed upper solution from (c). EXPERIMENT A 4 6 6 P EXPERIMENT EXPERIMENT A S03 c FIGURE 13 A 4-67 b h i EXPERIMENT A S O * d 72 proteins of the serum. Dialysis of the mother liquor completely abolished the effect as shown by the result of overlaying dialyzed mother liquor plus albumin on serum (d) , Figure 13, a pattern such as would result if albumin alone were overlaid on serum. The small peak in the mother liquor, probably a mucoprotein, was not removed by dialysis and was thus ruled out as a possible complexing factor. Data for this series of experiments are found in Table IV, Inorganic elements of the serum as complexing factors. Ashing of the serum to test whether inorganic components of the supernatant from the heated serum coagulum were active gave negative results as shown by the addition of the ash to Ringer1s solution (not illustrated). The addition of Mg** ion to Ringer1 s solution likewise proved that Ion inactive at physiological concentrations or slightly greater. Organic elements of the serum as complexing factors. Though this group includes a great many substances, only the Ionic substanees were considered. One of the larger groups present of this type, the amino acids and peptides seemed reasonably possible as active factors. Accordingly glycine was added to a Ringer’s solution (glycine concen tration 0,25 per cent) containing albumin and the resulting 73 pattern (e) , Figure 13, showed the presence of a large immobile boundary (“complex" plus epsilon) and was for all intents and purposes identical with the pattern for mother liquor plus albumin overlaid on serum* The control experiment of glycine in Ringer’s solution (glycine concen tration 0*25 per cent) overlaid on serum gave a pattern identical to that obtained using undialyzed mother liquor overlaid on serum (a), Figure 13* No other substances of the serum were examined for activity although quite conceivably such substances as the lipids per se could play an important role in the complex- ing of serum proteins and there is good evidence that lipoproteins form complexes with other serum proteins (17, 33, 34). Chemical and Physical Influences on the Formation of the "Complex" Ionic strength. A study of the effect of ionic strength upon the formation of the "complex" was carried out using the experimental system as represented by (a), Figure 12, (albumin plus alpha and beta globulins overlaid on undiluted serum) but under special buffered conditions. In addition to electrophoretic equilibration of the upper solution with veronal buffer at pH 8.0 and ionic strengths of 0.02, 0.11 and 0.20 respectively, the upper solution 74 was dialyzed against the serum (lower solution) for further equilibration (see section on Materials and Methods)* Results of these experiments are shown in Figure 14* Dilutjon and dialysis* The effects of both dilution and dialysis were not separately studied but the influence of these physical effects are actually included as parts of specific experiments cited above and will be considered in the data to be discussed in detail in the next section* Isolation Experiments Isolation of the X-protein* Employing the method of Pedersen (17) described previously (in the section on Materials and Methods) the X-protein layer was isolated from normal human serum* Only two samples out of four used gave positive results confirming Pedersen* s work* Those two samples of X-protein from the same person yielded an electrophoretic pattern, Figure 15, containing definite amounts of albumin together with beta globulin as described by Pedersen (17)* 75 FIGURE 14 ELECTROPHORETIC PATTERNS RESULTING PROM SUPERIMPOSING ELECTROPHORETIC ALLY EQUILIBRATED ALBUMIN PLUS ALPHA AND BETA GLOBULINS UPON UNDILUTED SERUM IN VERONAL BUFFER AT DIFFERENT IONIC STRENGTHS (a) Veronal buffer ionic strength O.IM. (b) Veronal buffer ionic strength 0.02M* (c) Veronal buffer ionic strength 0*2M# 76 FIGURE I4> ft I t £ A lb <X+fi Ave %7otal Area yAlbuM/*, zz Beta.* AI p h * GartuWd-hg (Ar Descending Side Ascending Side FIGURE 15 ELECTROPHORETIC PATTERNS OF X-PROTEIN LAYER DISCUSSION The first experimental work undertaken was that dealing with the establishment of conditions for minimal epsilon boundary anomaly in the overlay studies, such that the fewest possible interpretative complications be encountered* The problem of epsilon boundaries in the pattern Interpretation is particularly important since the well known boundary anomaly usually seen in the descending limb, the epsilon anomaly, Is always found In the same region as the Immobile boundary representing the proposed serum ^complex”* The latter, which could possibly be the same as the X-protein complex separated by ultracentri- fugation following Pedersen's technique (17), depends at present for its acceptance upon pattern interpretation* Assuming no Interaction of albumin with itself when a more dilute solution of albumin is overlaid on a more concen trated one, experiments were carried out to determine under what conditions the epsilon anomaly would have the smallest pattern area* It was found that when a solution of albumin first electrophoretically equilibrated with Ringer's solution (by electrophoretically moving it out Into the Ringer's solution) was overlaid on a more concentrated solution of the same without any dialysis or other 79 treatment prior to using, the smallest epsilon boundaries, sometimes scarcely visible, were observed (a), Figure 8* When, however, an albumin solution from the same source was, in addition, pretreated by dialysis against Ringer*s solution, the resulting pattern (b), Figure 8, contained the customary epsilon boundary. A similar experiment to that just described but using veronal buffer throughout including the initial equilibration and dialysis yielded a pattern very similar to that with Ringer1s solution and dialysis. Dissolving the dry protein in veronal buffer with subsequent dialysis likewise resulted in the usual epsilon boundaries (d), Figure 8. From these experiments, it was concluded that electrophoretic equilibration without dialysis would give the most suitable conditions for minimal epsilon boundary when overlaying pure protein solutions on themselves. The largest area for an epsilon boundary in the lower solution under optimal conditions was some forty units (arbitrary units to be used through out) for four experiments and for both upper and lower solutions represents a range of only zero to eighty units, while after dialysis the maximum area was as great as two hundred pattern units. For purposes of interpretation, this investigator has therefore assumed that the maximum epsilon boundary in the lower solution for overlay 80 experiments involving electrophoretic equilibration of the overlying solution without dialysis will not exceed one hundred pattern units at the very outside* Anything greater than this will then represent the “complex* area* Likewise for any experiment involving dialysis in the pretreatment of the solutions used, an area of two hundred area units has been set as the maximum value for the anomalous boundary, anything greater representing the “complex* fraction area* These limits, however, do not preclude the possibility of larger anomalous boundaries occuring with other systems such as serum overlays, especially since it is very difficult to determine what conditions are actually the best when dealing with undiluted serum* It should be noted here, however, that when undiluted serum Is overlaid with Ringer1 s solution, a pattern (e), Figure 12, results showing no epsilon boundary at all. It was of particularly great Interest that neither the large protein concentration gradient between the overlying protein solution and the salt solution above it nor the superimposed protein gradient between the upper and lower solutions in the overlay experiments necessarily gave rise to large anomalous boundaries as has been, to a certain extent, Implied in the literature. It is of course quite possible, as 81 suggested by Svensson (6) especially in cases of complex mixtures like serum, that the inorganic salt anomalies may be masked by the large protein gradients at the higher concentrations of protein. Using the above defined limits for epsilon boundaries under the prescribed conditions, patterns obtained by over laying solution mixtures of pure Cohn fractionated proteins on more concentrated solutions of the same could then be interpreted with some degree of certainty* Since previous experimental work suggested the presence of albumin and one or more of the globulins in the "complex*, mixtures of these proteins were used* When a mixture of Cohn Fractions V plus IV-7-2B (albumin and beta globulin) was overlaid on a more concentrated solution of the same after first equilibrating electrophoretieally, a pattern containing an immobile boundary of some 250 units was observed (a), Figure 9. Under these conditions very little epsilon boundary is expected and considerable interaction of the proteins of the two solutions is indicated. This, moreover, is substantiated by the fact that the albumin and beta boundaries in the lower solution are of a smaller area than predicted by the concentration differential. When albumin was overlaid on albumin, the areas of the respective peaks were proportional to the concentration difference as 82 expected since no interaction had taken place. Rirther support to this contention is provided by the fact that the mobilities of the albumin and beta globulin in the overlying solution are slower than their mobilities in the lower solution. This would result in a net dilution of the upper solution {increased volume swept out between respective boundaries) and would make the lower boundary appear larger (increased gradient) in that pattern, which is the opposite effect to that observed. This finding is in accord with evidence to be considered later in this section. Pattern (b), Figure 9, resulting from overlaying of the same mixture not electrophoretically equilibrated but dlalyzed against veronal buffer showed the presence of an immobile boundary significantly larger than the expected epsilon boundary, but not as much in excess as in the Ringer’s solution run above. From the relative concen trations of albumin and the mobilities of the albumin in the upper and lower solutions, there is good indication that interaction has occured under these conditions because the mobility of the albumin is slowed to an even greater extent in the upper solution in this experiment than in the case of the Ringer* s experiment. Thus the expected albumin concentration over and above the albumin contained in the • ’complex1 * is less than the observed albumin area in 83 the lower solution* When this type of study was extended to a mixture of Cohn Fractions V and II, albumin plus gamma globulin, the results were not as clear cut* The reason for this is that when dealing with solutions containing gamma globulin, the poor separation of the immobile fraction from the gamma boundary makes pattern interpretation very difficult. ITsing the electrophoretically equilibrated solution mixtures overlaid on more concentrated solutions of the same resulted In a pattern (c) , Figure 9, having not much more than the expected area based on the protein concentration differential and the mobility of the albumin in the upper solution relative to the lower* This Indicates an inter action which is certainly far from obvious, possibly due to Incomplete separation of the gamma and "complex" plus epsilon boundaries. The results obtained using solutions non-equilibrated but dialyzed against veronal buffer lend further Indication of interaction* The pattern obtained (d), Figure 9, shows an Immobile boundary of considerably greater area (320 units average) than the expected epsilon anomaly, which considered In conjunction with the size of the lower albumin and the relative mobilities suggests an interaction. It must be concluded from the above experiments 84 representing the interaction of pure proteins in buffer or salt solution that an incomplete degree of interaction takes place in these artificial systems which are far from physiological* The results with serum used as an under lying solution appear to be far more clear cut as might be expected since if we are dealing with a physiological protein equilibrium, the presence of certain serum elements, as will be seen later, is very probably necessary for interaction* The next phase of the problem was to study the interactions of purified fractionated proteins with proteins in native serum. The patterns shown in Figure 10 represent the results of overlaying single pure Cohn Fractions on undiluted serum, both solutions dialyzed against veronal buffer. In summary it may be said that only in the case of albumin (a), Figure 10, was there any significant interaction, epsilon boundaries of the usual size being obtained with beta and gamma globulin alone. The albumin interaction produced an immobile boundary of some 310 units. Moreover, later work does support the finding that albumin interacts with serum and is in fact, a necessary constituent for the formation of the "complex* In the presence of a complexing agent. When mixtures of the pure proteins or mixed subfractions all containing 85 albumin are overlaid on undiluted serum, both solutions dialyzed against Ringer’s solution, the resulting patterns, Figure 11, all show the presence of very significant ’ •complexes’ * in excess of the expected epsilon boundary# It is noted that all of the ’ ’ complexes* are of the same order of magnitude and that the addition of even small amounts of a globulin to the albumin results in a much increased “complex’ * especially in the case of beta globulin (a), Figure 11. Though the significance is not understood, it should also be noted that In any case where a “complex” has been observed using beta globulin in combination with albumin, the “complex” is always well separated from the gamma globulin and has a small but definite cationic mobility at the alkaline pH 8.0. These results Indicated considerable difference in the behavior of fractionated proteins toward underlying proteins In native serum compared to their behavior toward interacting with themselves when used as underlying fractions. Such proteins are presumably still native but in all probability do differ in certain respects from the same molecules as they existed in the serum originally. Perhaps an important difference in their reactivity is due to the absence of certain other proteins which must influence their behavior to some degree. 86 If on© compares the pattern (c), Figure 11, (Cohn Fractions V plus IV-4 overlaid on serum) with (b) , Figure 12, an electrophoretically equilibrated mixture of the same fractions separated from fresh serum into RingerT s solution and overlaid on serum with approximately the same total protein concentrations in each overlay, the very much larger "complex* is seen in the case of the serum proteins. The latter experiment, confirmed several times, resulted in the largest "complex* boundary of any experi ment to be reported (920 units)• A very similar pattern (a), Figure 12, to this was likewise obtained from solutions of separated (equilibrated) albumin alone. The Immobile boundary in this case contained 620 units. When, however, a solution of electrophoretically separated (equilibrated) globulins (alpha, beta and gamma) in Ringer1s solution were overlaid on serum, a pattern (d), Figure 12, resulted having no immobile boundary at all. This finding together with the evidence above using fractionated proteins overlaid on serum has led to the definite conclusion that albumin is necessary for the formation of the "complex". Striking results were obtained when the overlying solution and serum from the experiment with the largest "complex* (b) , Figure 12, were both dialyzed against RingerT s solution. The observed pattern 87 (c) , Figure 12, showed a greatly diminished "complex* boundary and a greatly Increased albumin boundary area (compare albumin thread (b) , Figure 12, with albumin peak (c), Figure 12)♦ The latter observation further substan tiated the role of albumin in the ’ ’complex* since the increase after dialysis was absolute in nature* The increased area could not be explained on the basis of differing mobilities of the over and underlying albumins or by dilution of the overlying albumin solution due to dialysis. It is also of great significance that when equilibrated globulins are overlaid on serum (d), Figure 12, there seems to be no visible epsilon boundary. This observation further supports the idea that under these conditions (equilibration in Ringer’s solution) the super imposed colloid gradient produces only a very small epsilon boundary or none at all. In contrast to this after dialysis against veronal, the epsilon boundaries are of the usual size even when no Interaction occurs, as when pure Cohn Fractions are overlaid on serum. The effect of running undiluted and undialyzed serum overlaid with RingerTs solution only is seen in (e), Figure 12. No epsilon or any other boundary anomaly is seen. Dilution by the Ringer’s solution apparently breaks up the "complex* completely as does dialysis, to be discussed below. 88 Since dialysis greatly reduced the size of the ®eomplex* in the experiment and its control, cited above, it was suspected that a factor responsible at least in part for the formation of the “complex* may have been present In the serum and have dialyzed out. As has been described in the section on Experimental Results, the supernatant from heat coagulated serum was used to isolate or at least localize the unknown complexing factor. To test for its complexing activity, a sample of this fluid was overlaid on undiluted serum. The unexpected pattern (a), Figure 13, resulted, having an unknown fraction *Xn present adjacent to the albumin boundary of the serum. This pattern is, however, not altogether unexpected if one considers what might be taking place under these conditions# It is certainly conceivable that a “complex* could form at this point rather than in the usual position under these conditions. In this case from the albumin gradient down ward all essential elements exist (albumin and globulins) and the essential complexing agent Is continuously being supplied by the superimposed heated serum supernatant. In the usual case where albumin Is present in the overlying solution together with the complexing factor, the “complex* Is formed at the first globulin gradient (gamma). When this supernatant, dialyzed against Ringer’s solution, Is 89 overlaid on undiluted serum not dialyzed, a pattern (b), Figure 13, results which is identical with that obtained when running undiluted serum overlaid by Ringer1s solution (e), Figure 12, showing no immobile or "X" boundaries whatsoever* These results confirm the presence of a heat stable, dlalyzable component in the serum and supernatant fluid. Further investigations with the supernatant fluid showed that when albumin was added to that liquid, the pattern (e) , Figure 13, having a large "complex1 * boundary was obtained. This result would lend strong support to the reality of the hypothetical "X-complex" in the pattern (a), Figure 13, obtained when the supernatant fluid alone was overlaid on undiluted serum. As a control to the experi- ♦ ment just described, the supernatant fluid was first dialyzed against Ringer1s solution with albumin added. The resulting overlay on serum yielded the pattern (d), Figure 13, which contained only the usual small "complex" produced by albumin. Attention was then turned to the isolation of the complexing agent, if possible, from the heated serum supernatant. This fluid, assuming no heat inactivation of the unknown agent, would contain that agent along with other dialyzable components of the serum. Using the direct approach to at least localize the complexing factor, the 90 heated supernatant was ashed with sulfuric acid to obtain inorganic cations* Ash consisting of the stable metal sulfates (including most of the metals of the serum since decomposition was kept to a minimum) was added to Ringer1s solution both in the presence and absence of albumin* The results were negative in each Instance, as was also the case •when Mg** Ion (considered as the most likely inorganic complexing agent) was added directly to Ringer1s solution in the presence and absence of albumin. Similar attempts to demonstrate the role of a cation by their removal from the heated supernatant using an ion exchange resin produced negative results. Since probably the largest group of organic anionic constituents of the serum (or heated supernatant) consists of the amino acids and peptides, these were next considered. In view of the fact that some evidence of binding of amino acids to proteins has appeared in the literature (55) as well as some reports of the complexing of metal ions to amino acids to form cationic complexes (56), these sub stances could conceivably mediate In the complexing of two proteins. Glycine was chosen for the first of these studies and was used as described in the section on Experimental Results. When this amino acid was added to Ringer1s solution at a concentration equivalent to the 91 total free amino acid content of the serum and overlaid on undiluted serum, a pattern as in (a), Figure 13, was obtained: that is to say for all intents and purposes, the overlay solution behaved just as undialyzed heated super natant, and there appeared a large shoulder separating fi*om the albumin as with fraction ®X® in (a), Figure 13* When albumin was added to the Ringer1 s solution containing glycine and was overlaid on serum, a large "complex® identical to (c), Figure 13, appeared just as had occured under the same conditions with the supernatant from the heat coagulum* Whether the substitutional effect of glycine is specific is indeterminable from the data and remains to be decided by further experimentation. Perhaps the effect is that of a non-specific amino acid or peptide which could be shown by a number of these compounds. Ionic strength studies were carried out in veronal buffer at pH 8.0 and varying ETaCl concentrations. The results of these studies are shown in Figure 14. Pattern (a), Figure 14, shows the results obtained when a solution of albumin, alpha and beta globulin separated from serum into veronal buffer, ionic strength 0.11, was overlaid on undiluted serum. This same solution had been previously dialyzed against the buffer (see section on Experimental Results for details) for maximal equilibration. A "complex® 92 of moderately large size is seen. When the same experi ment is repeated in every detail in veronal buffers of ionic strengths 0.02 and 0.20, the patterns (b) and (c), Figure 14, respectively, are obtained. These results indicate that a physiological range of salt concentration is important for maintenance of the "complex” although more complete experimentation is required to establish this point unequivocally. In an effort to present some less interpretative type of evidence for a "complex”, isolation of the X- protein complex, first described by Pedersen (17), was attempted. Results confirmed the findings of Pedersen that the oily layer obtained from the ultracentrifugation of at least one sample of human serum in the present study showed the presence of albumin and alpha plus beta globulins with small amounts of gamma globulin. This experiment was repeated three times. Two other samples of human serum showed alpha plus beta globulins to be the primary constit uent of the oily layer with only traces of albumin and the other globulins. The possibility of contamination of the oily layer by protein constituents of the serum was prac tically nil since the serum was layered under sucrose with the result that the oily layer was separated from the serum by the sucrose. These findings, though very important 9 3 supporting evidence for the "complex*, should be taken with certain reservation. The purpose for this is that there is no particular reason to assume that the X-protein complex is necessarily the same as that found in the electrophoretic studies. It may be concluded, however, that the X-protein bears a close relationship to the electrophoretic "complex" as revealed by the constituents into which it dissociated upon dialysis. The fact that dialysis did dissociate the X-protein complex indicates that it likewise may be held together by some complexing factor as in the case of the electrophoretic "complex*. A survey of the results of this investigation to gether with a consideration of previously reported evidence permits a few generalizations to be made regarding the nature of the serum "complex". The "complex" is apparently composed of albumin plus one or more of the globulins and at a pH of 8.0 has no mobility or a very slight cationic one. The formation of the "complex* requires the presence of a heat stable, dialyzable component for which glycine may be substituted. Dilution dissociates the "complex* as does dialysis (implied above) indicating the weak nature of the association of the protein components to form the "complex". Whether the "complex* preexists In the serum could not be determined from the data. The serum proteins 94 appear to Interact to form a "complex” in artificial systems to a limited extent. The interaction of albumin with beta globulin appears to be quite clear eut while gamma globulin and albumin do not show evidence of Inter action. It is likely that this type of "complex* formation Is not the same as that found in the undiluted serum studies. However, It is highly significant that an inter action was shown to take place using the purified fraction ated proteins since this evidence indicates a ready Inter action between the proteins of the serum over a wide range of conditions. When physiological conditions exist presumably the greatest degree of Interaction would occur although this Is not necessarily true. It Is this Investigator1s conception that the "complex" Is probably not a stolchiometrically related compound but a labile complex consisting of associated or aggregated protein molecules having an ephemeral relationship brought about by the forces previously described. It is obvious from the results discussed that some very interesting points remain to be clarified by further experimentation. The actual isolation of the complexing factor should be relatively easy to achieve In light of the findings with glycine. Specificity for glycine should first be tested using other amino acids or related 95 compounds* In addition the interaction of the Cohn Fractions should be tested in the presence of glycine or other complexing factors. Isolation of the “complex® in the presence of albumin plus the complexing factor is, of course, a very important experiment to be carried out. This may be done by removing all the solution in a des cending superimposed system from the upper solution albumin band down through the “complex® boundary to the albumin band in the lower solution. Re-electrophoresis of this solution in the conventional manner should then result in an albumin concentration greater than that predicted from the known original concentration of free albumin In the upper solution by an amount of albumin released from the “complex® (due to the dilution effect of electro- phoreslng directly against Ringerfs solution). Although this Is somewhat indirect evidence, it is especially Interesting since it should give considerable Information about the amounts of the proteins which go to make up the “complex®. Under very accurately controlled conditions, it is conceivable that dissociation constants for the protein constituents could be calculated. Further evidence regarding the effect of pH and specific buffer Ions upon the formation of the “complex® Is needed as is more information leading to an understanding of the relationship 96 of the X-protein to the "complex* described in this investigation. Finally, the extreme importance of the "complex* to physiological activity has been overlooked in the discussion of the results. This phase of the over-all problem is of course the most important but will have to remain obscure until more fundamental relationships have been worked out using systems which are as applicable to well known physiological principles as is possible. SUMMARY AND CONCLUSIONS Both early and recent literature pertaining to the proposed problem of the interaction of blood serum proteins was reviewed in light of known physical chemical methods for studying physiological systems and from a standpoint of known physiological activity wherein interactions are to be found* An extensive treatment was given not only to the published work but also to considerable unpublished background material, which formed a logical basis for the problem* The problem undertaken was an investigation of the immobile boundary observed In the electrophoretic patterns of undiluted serum overlaid with a diluted serum which appeared to be a protein "complex*. It was the object of this study to demonstrate that the "complex* does exist and to investigate some of the factors responsible for its existence. Studies were made using the method of super imposing one protein solution upon another as a descending electrophoretic system* This technique was employed with various protein solutions and an undiluted serum under various physical and chemical conditions. In addition studies of a protein complex, isolated by ultracentrifu gation were carried out* 98 The special electrophoretic apparatus used for the investigation was described and illustrated in detail as was the technique for which it was designed, namely, the superimposing of one protein solution upon another of higher concentration. Special attention was given to describing the protein solutions as well as their preparation and treatment for use In the overlay studies. An extensive investigation of the epsilon boundary anomaly occuring In the same region as the proposed "complex" was carried out. Albumin solutions overlaid on more concentrated solutions of the same under varying conditions revealed that electrophoretic equilibration of the protein solution with the medium prior to over laying and without subsequent dialysis against the medium, gave rise to the smallest epsilon boundaries In the upper and lower solutions. Dialysis of equilibrated solutions against veronal buffer or Ringer* s solution gave rise to the usually reported epsilon boundaries. It was concluded that neither a large protein concentra tion gradient with the medium nor the gradient between protein at one concentration superimposed upon another of higher concentration necessarily gave rise to anomalous boundaries. Electrophoretic equilibration apparently establishes an equilibrium with the environment which 99 Is not disturbed by the dynamics of the electrophoresis. Using the conditions established for minimal epsilon boundaries, the interaction of Cohn fractionated serum proteins was studied. Albumin was shown to interact with beta metal-combining globulin, giving rise to a significant Immobile "complex'* boundary, while no sig nificant Interaction was detected using gamma globulin and albumin. The interaction of Cohn fractionated proteins with proteins as they exist in the serum was demonstrated by superimposing solutions of the former on undiluted serum. In this Instance Interaction was found to be considerably greater than in the completely artificial system, presumably due to the absence of certain serum elements. It was found that albumin in the presence of alpha, beta or gamma globulin gave rise to moderately large "complex" boundaries. The interaction of electrophoretically separated (equilibrated) serum proteins, isolated from fresh serum samples, with proteins as they exist in the serum was studied by overlaying the former on the latter. These systems gave rise to very large "complex" boundaries. When albumin plus alpha and beta globulin were overlaid upon serum, the greatest amount of interaction of the entire investigation was observed. Albumin alone showed 100 moderate interaction while globulins alone showed no interaction whatsoever* When undiluted serum was run directly against Ringer’s solution in the conventional manner, no "complex” was observed* It was concluded from this observation together with the fact that no “complex" was observed in any of the upper solution patterns that dilution dis sociates the "complex". Dialysis was also shown to dissociate the "complex"* Upon dialysis of the upper solution and serum of the experiment showing the greatest "complex" boundary area, a much reduced "complex" boundary was observed together with a large increase in the serum albumin boundary which prior to dialysis appeared very small. These results led to the conclusion that a dialyzable component was present which was in part responsible for the formation of the "complex". The supernatant from heat coagulated serum was found to contain a dialyzable complexing agent and it was shown that glycine was able to substitute for this agent, however, its specificity was not determined* In organic elements of the serum were shown to be inactive as complexing factors* The X-protein complex of Pedersen was isolated with 101 the ultracentrifuge using a special technique to avoid contamination of the oil layer (containing the X-protein) by the serum proteins. This complex was shown by electro phoretic analysis to contain albumin and beta globulin with a very snail amount of gamma globulin. The possible relationship of the X-protein to the electrophoretic "complex* was discussed. It was concluded that a "complex* exists and that it is composed of albumin and one or more of the globulins. Its formation is dependent upon the presence of some heat stable dialyzable component (for which glycine may be substituted) • The "complex* is immobile at pH 8.0 or has a very slight cationic mobility. Dialysis and dilution dissociate the "complex*. Some general observations regarding the "complex* have been made and a prospectus for future experimentation included. BIBLIO GRAPHS' 103 1* Cohn, E.J., Gurd, F.R.N., Surgenor, D.M., Barnes, B.A., Brown, R.K. , Derouaux, G. , Gillespie, J.M., Kahnt, F.W., Lever, W. P., Liu, C.H., Mittelman, D., Mouton, R.P., Schmid, K., and Uroma, E*, J. Am* Chem. Soc., 72, 465 (1950). 2. Hardy, W.B., J. Physiol., 33, 17 (1905). 3. Reuss, F.P., Memoires Soc. imperiale naturalistes Moskau, 2, 327 (1809)• 4. Tiselius, A., Trans. Paraday Soc., 33, 524 (1937). 5. Jameson, E., Arch. Biochem., 15, 389 (1947). 6. Svensson, H., Arkiv. Kemi. Mineral. Geol., 22A, No. 10 (1946). 7* Armstrong, S.H., Jr., Budka, M.J.E., and Morrison, K.C., J. Am. Chem. Soc., 69, 416 (1947). 8. Perlmann, G.E., and Kaufman, D., J. Am. Chem. Soc., 67, 638 (1945). 9. Dole, V.P., J. Am. Chem. Soc., 67, 1119 (1945). 10. Jameson, E., and Alverez-Tostado, C., J. Am. Chem. Soc., 65, 459 (1943). 11. Goldwasser, E., and Putnam, F.W., J. Phys. & Colloid Chem., 54, 79 (1950). 12. Klotz, I.M., Gelewitz, E.W., and Urquhart, J.M., J. Am. Chem. Soc., 74, 209 (1952). 13. Oncley, J.L., Ellenhogen, E., Gitlin, D., and Gurd, F.R.N., J. Phys. Chem*, 56, 85 (1952). 14. Scatchard, G., Am. Scientist, 40, 61 (1952). 15. Pedersen, K.O., Ultracentrifugal studies on serum and fractions, Uppsala (1945). 16. McFarlane, A.S., Biochem. J., 29, 660 (1935). 17. Pedersen, K.O., J. Phys. & Colloid Chem., 51, 156 (1947). 104 18* Pressman, D., and Siegel, M., J* Am. Chem* Soc., 75, 686 (1953). 19. Klotz, I.M., Gold Spring Harbor Symposia Quant. Biol*, 14, 97 (1949). 20. Klotz, I.M., Arch. Biochem., 9, 109 (1946). 21. Longsworth, L.G., and Jacobsen, C.P., J. Phys. & Colloid Chem., 53, 126 (1949). 22. Morawetz, H., and Hughes, W.L., Jr., J. Phys. Chem., 56, 64 (1952). 23. Tiselius, A., Biochem. J., 31, 1464 (1937). 24. Tiselius, A., and Kabat, E.A., J. Exp. Med., 69, 119 (1939). 25. Longsworth, L.G., and Maclnnes, D.A., Chem. Rev., 24, 271 (1939). 26. Svensson, H., Arkiv. Kemi Mineral. Geol., 17A, No. 14 (1943). f 27. Moore, D.H., and Lynn, J., J. Biol. Chem., 141, 819 (1941). 28. Longsworth, L.G., J. Am. Chem. Soc., 61, 529 (1939). 29. Greenberg, D.M., Amino acids and proteins, Springfield (1951). 30. Phillips, R.A., Van Slyke, D.D., Dole, V.P., Emerson, K., Jr., Hamilton, P.B., and Archibald, R.M., Bull. TJ. S. Army Med. Dep., 71, 66 (1943). 31. Kingsley, G.R., J. Lab. Clin. Med., 27 , 840 (1942). 32. Alberty, R.A., Abstracts of Meetings of the American Chemical Society, Los Angeles, 1953. 33. Macheboeuf, M., and Rebeyrotte, P., Discussions FSaraday Soc., No. 6, 62 (1949). 34. Lindgen, P.T., Elliot , H.A., and Gofraan, J.W., J. Phys. & Colloid Chem., 55, 80 (1951). 105 35* Tullis, J.L., Blood cells and plasma proteins, New York (1953). 36. Monk, C.B., Trans. Fkraday Soc., 47. 285 (1951). UMI Number: DP21549 All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. U M I Dissertation Publishing UMI DP21549 Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author. Microform Edition © ProQuest LLC. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 ProQuest Que ^ _

Asset Metadata

Creator Barnet, Harry Nathan (author)

Core Title Protein-protein interactions in blood serum

Contributor Digitized by ProQuest (provenance)

School Graduate School

Degree Doctor of Philosophy

Degree Program Biochemistry

Degree Conferral Date 1953-06

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag chemistry, biochemistry,OAI-PMH Harvest

Language English

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c17-1522

Unique identifier UC11347895

Identifier DP21549.pdf (filename),usctheses-c17-1522 (legacy record id)

Legacy Identifier DP21549.pdf

Dmrecord 1522

Document Type Dissertation

Rights Barnet, Harry N.

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA