AN INVESTIGATION OF THE ELECTROPHORETIC PATTERNS OF FISH SERUM AND PLASMA PRGTEINS. WITH SPECIAL REFERENCE TO THE INFLUENCE (IF A DIURNAI. OXYGEN PULS£ ON SERUM PROTEINS Thad: Io: The boom of M. S. MICHIGAN STATE UNIVERSITY Gerald Ray Bouck €963 WWIm1wmumwu\m'mml ABSTRACT AN INVESTIGATION OF THE ELECTROPHORETIC PATTERNS OF FISH SERUM AND PLASMA PROTEINS, WITH SPECIAL REFERENCE TO THE INFLUENCE OF A 'DIURNAL OXYGEN PULSE ON SERUM PROTEINS by Gerald Ray Bouck Conditions for separation and evaluation of fish plasma and/or serum proteins were investigated. Specific recommen- dations for utilizing the Beckman/Spinco model R paper elec- trophoresis system are listed. Certain confidence limits for this system are described. A survey was made of the electrophoretic serum and/or plasma protein fractions of twenty-three species of fish, one species of immature aquatic insect, and one Species of freshwater crayfish. Each pattern was found to be suffici- ently unique to differentiate between Species and in one case, sub-species were differentiated. Fundamental differences between the serum protein fractions of man and fish, as well as between fish species are described. As shown by electrophoretic separation, the number of protein fractions, their positions, and their conjugation with polysaccharides and lipids differ widely. The iso-electric protein in fish serum appears to be similar to mammalian gamma globulin. A rank order of iso- electric quantities in twenty-three species of fish revealed that "pollution sensitive” fish have less of this fraction than do "pollution tolerant” fish. Gerald Ray Bouck Bluegills (Lepomis macrochirus), largemouth bass (Micropterus salmoides), and yellow bullheads (Ictaluris natalus) were subjected to a diurnal low-oxygen pluse in an artificial stream. Serum patterns of bluegills and bass were significantly changed while the pattern of bullheads did not change. This conforms to an empirical estimate of their pollution tolerance. Other changes noted were (1) a signifi- cant change in the distribution of serum glycoproteins in bluegills and bullheads, (2) a change in the relationship of certain protein fractions to body weight or length, (3) a change in protein composition as shown by the conjugation of protein with polysaccharides in bluegills and bullheads. _During the periods of low dissolved oxygen, the fish vomited indigested food, increased ventilation activity, reduced swimming activity and lost normal body color. QMQM AN INVESTIGATION OF THE ELECTROPHORETIC PATTERNS OF FISH SERUM AND PLASMA PROTEINS, WITH SPECIAL REFERENCE TO THE INFLUENCE OF A DIURNAL OXYGEN PULSE ON SERUM PROTEINS By Gerald.Ray Bouck A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Fisheries and Wildlife 1963 ACKNOWLEDGMENTS The writer wishes to express his sincere-appreciation to Dr. Robert C. Ball for his stimulation and guidance. Thanks are also extended to the writer's fellow graduate students who unselfishly assisted at various times in the collection of the Specimens utilized in this project. In particular, he wishes to thank Mr. Jack Wood, Mr. Darrel King, Mr. Kenneth Linton, Mr. Robin Vannote, and Mr. Benjamin Tuller. Lastly, tha writer wishes to express his thanks to Dr. Masaru Fujiya of the Inland Sea Fisheries Research Station, Ujina, Hiroshima, Japan, for his helpful suggestions. 11 INTRODUCTION TABLE OF CONTENTS ELECTROPHORETIC METHODS. . . . . . . . . . Electrophoretic Migration of Protein on Paper Electrophoretic Migrations on Cellulose Acetate Evaluating Electrophoretically Separated Serum Proteins Evaluation of the Electrophoretic Method. Methods for the Collection of Blood ELECTROPHORETIC PATTERNS OF FISH AND INVERTEBRATES. ARTIFICIAL STREAM SYSTEM . . . . . . . . . The Influence of a Diurnal Oxygen-Pulse on Bluegills Influence of Bullheads Influence of a Diurnal Oxygen- Pulse on Largemouth Bass. SUMMARY LITERATURE CITED APPENDICES Appendix I Appendix II Appendix III a Diurnal Oxygen- Pulse on Yellow Raw Data from Male Bluegills, Yellow Bullheads, and Largemouth Bass Expos to a Diurnal Oxygen— Pulse and Tested Electrophoretically Staining Methods for Serum and/or. Plasma Proteins, Glycoproteins, and Lipoproteins after Paper Electro- phoresis . Staining Solution Used to Stain Protein, Glyc0protein, and Lipo- protein in Serum and/or Plasma. iii Page I’\) O 76 88 97 106 118 122 ed 123 128 131 TABLE 10. 11. LIST OF TABLES Summary of data obtained from electrophoretic analyses of sixteen portions of pooled yellow bullhead serum . . . . . . . Summary of data obtained from fourteen repeated scannings of the same strip containing rain— bow trout serum protein. Summary of water chemistry during the degassing periods for bluegills Summary of the influence of a diurnal oxygen- pulse on the distribution of proteins in the serum of male bluegills Coefficients of correlation for the distribu- tion of serum proteins to total body length and body weight of male bluegills . Summary of the influence of a diurnal oxygen- pulse on the distribution of glycoproteins in the serum of male bluegills Influence of a diurnal low oxygen pulse on the relationship of glycoprotein to protein in male bluegill serum Comparison of the distribution of protein in the serum and plasma fractions of bluegills collected at different locations and months Summary of the water chemistry during the degassing periods for yellow bullheads. Summary of the influence of a diurnal oxygen- pulse on the distribution of proteins in the serum of male yellow bullheads Coefficients of correlation for the distribu- tion of serum proteins to total body length and body weight for male yellow bullheads. iv PAGE 26 27 77 81 83 86 87 89 91 92 TABLE 12. 14. 15. 16. 17. Summary of the influence of a diurnal oxygen- pulse on the distribution of glycoproteins in the serum of male yellow bullheads Influence of a diurnal low oxygen—pulse on the relationship of glycoproteins to protein in male yellow bullheads . . . . . . Summary of the influence of a diurnal oxygen— pulse on the distribution of lipoproteins in the serum of male yellow bullheads Summary of the water chemistry data collected during the degassing periods for largemouth bass Summary of the influence of a diurnal oxygen- pulse on the distribution of proteins in the serum of male largemouth bass. Coefficients of correlation for the distribu- tion of serum proteins to total body length and body weight of male largemouth bass PAGE 95 96 98 100 103 105 FIGURE 11. 12. 13. LIST OF FIGURES An electrophoresis bridge to modify Spinco/ beckman {Durram type) electrophoresis cells for the use of cellulose acetate sabstrates . Uptake of bromphenol blue by bovine beta- globulin as evaluated by scanning in a Spinco Analytrol . . . . . . . . . . Relationship of empirical pollution tolerance to apparent blood volume . . . . . . . Electrophoretic pattern of human plasma proteinSg O O 0 O O O O O O O O O Electrophoretic pattern of the Bowfin (Amia calva) . . . . Electrophoretic pattern of the plasma proteins of the Rainbow trout (Salmo gairdnerii gairdnerii). Electrophoretic pattern of the plasma proteins of the Kamloops trout \Salmo gairdnerii kamloops) . . . . . . . . . . . . ElectrOphoretic pattern of the serum proteins . f “1 . ' of the Brown trout tbalmo trutta) ElectrOphoretic pattern of the serum proteins of the Brook trout (Salvenius fontinalis Electrophoretic pattern of the serum proteins of the Western grass pickerel {Esox americanus vermiculatus) Electrophoretic pattern of the Northern Pike Esox (Esox lucius) ElectrOphoretic pattern of serum proteins of the goldfish {Carassius auratus). Electrophoretic pattern of plasma proteins of the Carp (Cyprinus carpio). vi PAGE 16 39 to Al A3 AA FIGURE PAGE 14. Electrophoretic pattern of serum proteins of the Lake Chubsucker (Erimyzon sucetta) . . A9 l5. Electrophoretic pattern of plasma proteins of the Northern Hog Sucker (Hypentelium nigricans). . . . . . . . . . . . 5O 16. Electrophoretic pattern of the serum proteins of the Black Bullhead (Ictalurus melas) . . 51 17. Electrophoretklpattern of serum proteins of the Yellow Bullhead (Ictalurus natalus) . . 52 18. Electrophoretic pattern of serum proteins of the Brown Bullhead (Ictalurus nebulosus). . 53 19. Electrophoretic pattern of serum proteins of Channel Catfish (Ictalurus punctatus). . . 51L 20. Electrophoretic pattern of plasma proteins of the Rock Bass (Ambloplites rupestiris) . . 55 21. Electrophoretic pattern of serum proteins of the Green Sunfish (Lepomis cyanellus). . . 56 22. Electrophorfit‘ ic pattern of serum proteins of the Pumpk inseed (Lepomis gibbosus) . . . 57 23. Electrophoretic pattern of serum proteins of the Bluegill (Lepomis macrochirus). . . . 58 2A. Electrophoretic pattern of the serum proteins of the Smallmouth Bass (Micropterus dolomieui). . . . . . . . . . . . 59 25. Electrophoretic pattern of the serum proteins of the Largemouth Bass (Micropterus 7 salmoides). . . . . . . . . . . . 6O 26. Electrophoretic pattern of serum proteins of , the Black Crappie (Pomoxis nigromaculatus) . Cl 27. Electrophoretic pattern of plasma proteins of the Yellow Perch (Perca flavescens) . . . 62 28. Electrophoretic pattern of plasma proteins of the freshwater crayfish (Orconectes _ propinquus) . . . . . . . . . . . 63 vii FIGURE PAGE 29. Electrophoretic pattern of plasma proteins of . the Dobson fly larvae (Corydalus cornutus) . 6A 30. Rank—order of fish species according to their mean per cent of iso-electric protein in their lasma and/or serum at pH 8.6 (vernol buffer? . . . 67 31° Re-circulating artificial stream for controlling dissolved oxygen content, temperature, and photo period . . . . . . . . . . . 72 32. Positional relationship of proteins, glyco— proteins, and lipoproteins in the serum of the bluegill . . . . . . . . . . . 79 33. Positional relationship of proteins, glyco— proteins, and lipoproteins in the serum of the yellow bullhead . . . . . . . . . 9A 3A. Positional relationship of proteins, glyco- proteins, and lipOproteins in the serum of the largemouth bass . . . . . . . . . 102 35. Hemoglobin contamination of the serum of pumpkinseed sunfish which was exposed to copper toxicity . . . . . . . . . . 107 36. Influence of sub—lethal cupric copper on the electrophoretic patterns of the serum of green sunfish . . . . . . . . . . . lll viii INTRODUCTION The blood of mammals has been recognized as a sensitive indicator of their physiological condition. Although there are notable differences between the blood of fish and mammals, many similarities exist. Both contain erythrocytes, leucocytes, and platelets or thrombrocytes. These are carried in a liquid menstruum and circulate in a closed circulatory system. This menstruum contains a mixture of electrolytes, amino acids, non—cellular proteins and other basic organic materials. The fluid portion of the blood is termed plasma. If whole blood is allowed to clot, fibrinogen and other coagulation factors are transformed into a fibrin reticulum which retracts and expresses the remaining fluid portions. The remaining fluid portion is termed serum and is essentially plasma with the fibrinogen and coagulation factors removed. The non-cellular proteins of the serum are believed to be produced by the reticulo—endothelial tissue cells. These non-cellular proteins of man's serum have been exten— sively investigated by many researchers and one hundred twenty— six of these studies have been summarized by Ehrmantraut (1958). Some conclusions from these studies have universal applica- tions and are germane to this study. That component of the plasma and/or serum which is non- mobile at pH 8.6 in barbital buffer during paper electro- phoresis is termed gamma globulin and is widely believed to possess great antibody functions. However, it has been demonstrated that antibodies exist in other fractions of the serum and/or plasma. Elevation of the quantity of this fraction and the general reduction of other protein frac- tions of the serum continues the "stress" pattern described by Dunn and Pearce (1961). However, the drastic alteration of any fraction is indicative of physiological change. Some other functions of man s plasma and/or serum proteins are mentioned by Weil (1959). Conjugation and subsequent transportation of various organic and inorganic substances such as hormones, metallic ions, lipids, and polysaccharides materials to plasma and/or serum protein occurs. Conjugation with these proteins appears to greatly increase the solubility and enhances the tranSportation of these materials to areas of utilization. Weil (1959) also points out that blood proteins form hydrostatic pressure to force fluids into the tissues on the arterial side and main- tain osmotic pressure by drawing fluid from the tissues into the blood on the venous Side of the capillaries. The latter is a function which he asserts could not be accomplished by electrolytes alone. Laurell (1961) points out that many metallic cations are chelated by these proteins, thus preventing their loss in the bile or urine. In example, transferrin (= siderophilin) chelates iron and may serve as a supply of iron for hemoglobin production during anemia and other stress periods. Prosser and Brown (1961) have discussed the ability of blood proteins to act as a buffer against pH changes and this aSpect is too well recognized to bear further discussion here. Whether these relationships are true for the blood of inframammalian vertebrates and invertebrates is being studied by many investigators at this time. Some evidence indicates that in general these relationships are valid for fish. Brachet (1957) and Anker (1961) have reported that RNA (ribose nucleic acids) has a high ability to synthesize proteins. A reduction of RNA would decrease the quantity of proteins produced by that organism and this change in protein bio-synthesis might be reflected in the serum protein fractions. Fujiya (1961b) has shown that kraft pulp mill waste causes serious reduction of RNA in the cells of the biliary duct system, hepatic and pancreatic systems of Sparus macrocephalus. Neuhold and Sigler (1960) have demonstrated that fluoride intoxication causes significant changes in the electrOphoretic patterns of carp (Cyprinus carpio). Fujiya (1961a) also has demonstrated that exposure to sublethal concentrations of copper sulfite, lead acetate, hydroxide ions, and hydrogen ions causes drastic changes in the electro— phoretic serum protein patterns of carp and other fish. These changes correspond well to the "stress" pattern of Dunn and Pearce (gp,_§it,) and appear to be due to changes in protein metabolism. 11 Fine and Drilhon (1960, 1961) have demonstrated that an antibody function is present in the serum of certain fish. From the previous statements one can infer that any significant change in the composition of the serum proteins is an event indicative of serious physiological ramifications. The increased emphasis on pollution control demands the continual search for better methods of detecting and evaluating pollution. One of the greatest needs in this area is a method for evaluating the influence of chronic, sub—lethal pollution conditions. Use of the electrophoretic ‘technique on serum proteins promisesto be useful in this area and as such constitutes the basis for evaluation of stress due to low oxygen-tension in this study. Low oxygen—tension stress represents one of the most common conditions associated with polluted water. In its most subtle form it appears as a diurnal pulse beginning a reduction late in the afternoon and reaching its lowest level prior to sunrise. During daylight hours, such water may show normal oxygen tensions and under certain conditions may even become super-saturated. Tarzwell (1957) has listed a numerical criterion for oxygen content of water as follows: not lower than 3 ppm at any time and not lower than 5 ppm for more than 8 hours in any 24 hour period. If one interprets this criterion for oxygen content of water literally, it is possible to allow a diurnal oxygen pulse which drOps rapidly to 3 ppm for eight hours and then returns to normal. It was decided to simulate such conditions in an. artificial stream and determine if these conditions produced changes in the serum proteins of fish. It was believed that the use of warm—water fish would produce information of greatest interest to pollution biologists. Other objectives of this study were to refine and evaluate the techniques of paper electrOphoresis and survey the electrOphoretic patterns of several Species of fish. ELECTROPHORETIC METHODS Electrophoretic Migration of Protein on Paper Proteins are colloids and as such have an electrically charged surface which by mutual repulsion prevents their coalescense. Proteins are also amphiproctic and can be made to vary their external charge by varying the pH of their medium. By placing proteins in a buffered solution of prOper pH and molar concentration, one can induce molecular migra- tion by subjecting them to an electrical field. These protein molecules migrate according to their electro-chemical prOperties and will separate into groups or zones of electo- chemically similar molecules. Paper electrophoresis consists of soaking filter paper strips in such a buffered solution, applying a protein sample near the center of the strip, and then applying sufficient current to maintain an electrical field on it. After migra- tion and subsequent separation of the various protein fractions, a protein denaturing agent is used to irreversably afix the fractions in their position on the paper. Location reagents are then applied to the paper strip to locate the protein fractions, and other fractions which have been separated. The literature contains many Specific instructions for the electrOphoretic separation and evaluation of human plasma and/or serum. However, these instructions are only of limited value for studies of fish since they were designed for a single species with samples containing high quantities of albumin and because they may apply only to one particular electrophoresis cell, type of buffer, and/or type of sub- strate. The general lack of a large quantity of protein in the fastest moving fraction of fish plasma or serum is a definite disadvantage. This fraction experiences the greatest amount of chromatographic capture by the paper and the farther it is caused to migrate, the more it will be absorbed. This difficulty appears to be a "saturation" phenomenon. Once the paper is saturated by the first fraction, it does not appear to capture proteins from the following fractions. This reduced quantity in the first-migrating fraction can be con- siderably lower than the original quantity. Therefore, it is best to use procedures which produce maximum zone separation with minimum total migration. Samples may be migrated either over night or for short periods, provided the above conditions are fulfilled. The conditions which achieved very good separation of fish plasma and/or serum protein in the Spinco/Beckman "Durram- type" electrOphoresis cell are as follows: 1. Use Schleicher and Schuell #2043a paper strips, eight per electrOphoretic run. 2. Barbital buffer (Vernol buffer) is used to wet the strips. It consists of 1.66 g barbituric acid (crystals) 12.76 g sodium diethyl barbitone ‘IIlIlII' l “lfu‘l[‘=l{ III'I' (powder) and sufficient distilled water to make one liter (Smith, 1960). 3. Use constant current at 0.5 MA/cm width of strip or 12.0 MA for eight strips. 4. Migration time: eight hours. The use of these procedures will yield consistently better results than the conditions listed in Beckman Technical Bulletin No. 6095A (1961). In some cases it may be more convenient to migrate the samples over night. To do this, the above procedures apply but the current is changed to 0.104 MA/cm width of strip or 2.5 MA for eight strips. A migration time of 14 hours gave best results for these con- ditions. To obtain maximum reproducibility of migration distance for each fraction, constant voltage must be used. However, the use of constant voltage requires refrigeration to off- set the constantly increasing current and its heat production. If refrigeration is not available, it appears to be necessary to Operate constant-voltage separations at voltages below 150 volts. Otherwise the heat produced will cause considerable evaporation off the strips and result in streaked or blurred patterns. Temperature greatly influences conductivity. At low temperatures, the conductivity will be low and result in lower current requirements to produce a given voltage. At higher temperatures, the conductivity will be higher and higher current will be required to produce the same voltage. 9 Four different buffer solutions were tried and it was found that the previously listed buffer solution best met the needs of this investigation. Borate, phosphate, and barbital-acetate buffers yield poor separation of fractions (resolution) and different appearing patterns with generally fewer fractions than Smith's I22-.El£-) buffer. With this in mind, one type of buffer Should be used on all strips. Although commercially prepared buffer powder is available and needs only the addition of distilled water to ready it for use, generally poorer results were obtained when it was utilized. This product was found to yield widely different conductivities and in some cases, various pH values. It is believed that these differences were associated with the generally poor resolution and fraction separation obtained when this buffer was used to separate serum proteins of fish. The protein fractions of fish serum with intermediate migration- speed are particularly difficult to separate when this buffer is used. Therefore, it appears necessary that each investi- gator weigh and mix the ingredients for each buffer solution. A considerable saving of the buffer solution can be realized if it is used in the following manner: (1) Make two liters of buffer solution, place in cells and on the strips in the usual manner. (2) Make a third liter of buffer solution and wet the paper strips with this solution on suc- cessive electr0phoretic runs. It will not be necessary to add further buffer solution to the cell compartments for several days. The easiest way to wet these successive strips 10 is to place them on the folding rack of the electrophoresis cell, place the rack in a dishpan, and add buffer solution to each strip in a gentle stream from a pipette. Allow the excess to drain off the strips or blot the strips gently, before replacing the folding-rack and its paper strips in the electr0phoresis cell. Approximately one hour should be allowed for the moisture content of the strips to equilibrate with the air in the cell. During this time, the level of buffer within the anode and cathode buffer compartments should be allowed to equilibrate. If the buffer level is higher in one com- partment than in the other compartment, capillary flowage in the paper will occur. Such flowage will smear and distort the fractions and the strip will be ruined. Continual re—use of the buffer solution on the strips is not recommended. However, the re—use of the buffer solu- tion in the cells is allowable provided the conductivity of the buffer solution is not greatly altered by evaporation. Evaporation can be greatly reduced by keeping the cells covered at all times. It should be noted that the buffer solution is attacked by microrganisms and should be discarded when any change is apparent. Some authors recommend the addition of sufficient merthiolate to obtain a concentration of l;100,000, to control such microrganisms. Sodium azide is also used for this purpose. The volume of the sample for protein analysis, lipo- protein analysis and g1y00protein analysis should be determined 11 for each Species under investigation. In general, 10 micro- liters of either serum or plasma are sufficient for protein and lipoprotein analysis, if the total protein concentration is in the range of 3-5 g/100 ml. However, the gly00protein content of fish serum and/or plasma is considerably lower than that of human serum. Therefore, a sample of 40—50 microliters is required for the analysis of glycoprotein in fish serum and/or plasma. In applying such large quantities of fluid to the already wet strip, one should be certain that the extra fluid does not run or smear. It is wise to apply such samples in units of 10 microliters and allow one to two minutes for the strip to assimilate this added fluid before adding more of the sample. It cannot be overly stressed that the sample must be evenly distributed on the strip and that the sample must not extend to the edges of the strip. If these conditions are not obtained, the sample usually will smear and blur or otherwise ruin the strip for evaluation purposes. The following procedures have proved to be of value in applying samples to the strips: (1) Always keep the Spinco applicator s wires in a horizontal plane. (2) Always apply the sample to dry applicator wires in such quantities that small beads of nearly equal volume are evenly distributed along the applicator wires. (3) Apply the sample to the paper with a quick thrust on the applicator depresser button. (4) The applicator should be held in contact with the paper for 20 seconds and then wiped dry for the next sample. Avoid uneven l2 pressure on the glass rods supporting the wet paper strips. Do not press the applicator wires so heavily that the glass rod is bent, as this may loosen the wet paper strips from their contact with the wicks. Staining methods for paper electrophoresis. After the electrophoretic separation is completed, the strips are removed from the cells by fully extending the folding rack and then lifting the extended rack from its support. The rack is then placed in an oven preheated to 110° C. for 30 minutes to denature the proteins. Peterson and Strong (1953) and Mertz (1959) point out that serum proteins are of the globular-type and have a coiled molecule which is profusely cross-linked within itself. Certain salts and heat have the ability to break these cross- bonds and allow the molecule to un—coil. Hence, after de- naturation, globular proteins have a structure similar to fibrous proteins. The extent of conversion of protein structure @enaturation)is related to the temperature and time of exposure to this heat. Henry, Golub, and Sobel (1957) have demonstrated that serum proteins on strips which were denatured at different temperatures have different uptakes of bromphenol blue. Brackenridge (1960) seems to have overcome this problem when using cellulose acetate as a substrate, by denaturing the protein with 30% sulfosalicylic acid. With this in mind, strict conformity to a single temperature of fixation should III! III) ’II I in: IIIIIIIIII I I 13 be the rule. Strips treated at other temperatures should not be compared to each other. It is interesting to note that of the serum proteins, only albumin has been reported by Smith (op. cit.) to complex with bromphenol blue before denaturation. Smith ( describes a simple test for the presence of albumin: a crystal of bromphenol blue (the smaller the better) is added to the serum sample on the strip. The albumin will retain some of the dye and the rest of the dye will migrate freely. At the end of the separation all free bromphenol blue will usually have migrated to the end of the strip. Thus, if albumin is present, two zones of dye can be distinguished. This method was used to determine if albumin was present in the serum of largemouth bass. This serum did not complex with bromphenol blue. If bromphenol blue is used to label a mammalian serum (i e. bovine), the migration of albumin is readily observed. The use of such a standard allows migrations to be ended at a definite point. Such migrations are easily reproduced. The alcoholic bromphenol blue method of staining the protein on the strips was used throughout this study. It consisted of a 10 minute pre-rinse in methanol to remove the buffer, a 30 minute bath in 0.1% bromphenol clue (acid salt) in methanol, followed by three, 5 minute rinses in 5% acetic acid. Although the stain appears to be good indefinitely, it was discarded whenever the pH rose above 4 (color change to blue) or after five series of strips have been stained in it. 14 After the final rinse in dilute acetic acid, the strips are removed and laid on filter paper for light blottingn The strips are then replaced on the rack and the rack iS-placed in the oven for 15 minutes. This leaves an-alkaline~residue of protein on the strips and assists in-color development. To develop maximum color intensity, the strips are then exposed to an ammonia vapor bath for 15 minutes,-but notlonger than 30 minutes. The strips are then ready to be evaluated. Saccharides conjugated with proteins are-termed glyco- proteins, Staining for glycoproteins is accomplished on separate strips with Shiff reagent after oxidizing the closed aldehyde groups in the saccharide-conjugate with periodic acid. All other substances which would be stained by fuchin- sulfite are removed in the ethanol pre-rinse. Lipids conjugated with proteins are termed lipoproteins. Staining for lipoproteins is accomplished on separate strips by a heat-saturated ethanol solution of "011 red 0." All other lipids are removed by the warm ethanol. It should be noted that in many species freezing apparently breaks the lipid to protein bond and frozen protein samples cannot be evaluated for lipoproteins. For a more complete description of the reagents and the staining procedures, the reader is referred to either Beckman Technical Bulletin 6095A (pp, cit.) or Appendix II. ElectrOphoretic Migrations on Cellulose Acetate Cellulose acetate has certain advantages over paper as a substrate for electrophoretic separations. This material 15 does not retard the migration of protein to the extent that paper does. Tailing and streaking of the protein fraction are reduced to a minimum. The total time for migration and staining are reduced to approximately one—fourth that-required for paper strips. In addition to these advantages, usually more fractions are separated and the uptake of certain-dyes appears to be truly quantitative to the amount of protein on the strip. Also, these strips can be evaluated by elution easier than paper strips. These advantages are not obtained without some dis- advantages. Cellulose acetate is very expensive and costs nearly $0.50/strip. These strips are easily torn and tend to break if bent in a short radius. The absorption of the buffer solution and other solutions into the strip must be accomplished with great care. Evaporation of water off the strip by joule heat production tends to be a problem of great magnitude and it is recommended that separations on cellulose acetate be performed with an ambient temperature of less than 35° C. Dialyzing the protein sample is advised due to the small quantity of buffer solution on the strip, but has not been utilized in this investigation. All contact with low weight alcohols must be avoided since these alcohols cause the strips to curl beyond repair. Utilization of this material was not possible or justi- fied in the previously described electrophoresis cell until favorable modifications and Optimum migration conditions could be ascertained. Figure 1 depicts a modified 16 .mmpmpquSm opwpoom omoHSHHoo Mo om: esp pom maaoo mfimoposaoppooam Awake Empgsav CmEXOom\oocHQm mmHUoE op mMUHSQ mHmSSOQQOSpooHo :< .H mhswfim l7 ya“ I! la )6 x ROD .GLHSS 18 electrophoresis bridge which allows the use of cellulose acetate strips on the Spinco/Beckman "Durram" type cell. Eight strips are accommodated as before. Very good separation is obtained using the following conditions: 1. Cellulose acetate strips, 2.5 cm x 12.0 cm. (May be purchased from Gellman Instruments Company, Chelsea, Michigan.) 2. Barbital buffer previously described. 3. Constant current as previously described. 4. Migration time: two hours. The lower current obtained by these conditions results in lowered heat production and the results are comparable to those obtained on paper. The buffer must be absorbed into the strips by care- fully floating them on the surface of the buffer solution. As the buffer is absorbed, the strips will turn light gray and any areas which have not absorbed it will remain white. The strip should not be plunged into the buffer until all such white areas have disappeared. Otherwise, air will be trapped and the strips will be ruined. At no time is it permissable to handle the strips with fingers. Micros00pe slide cover-glass forceps are useful for handling these strips. Smaller sample volumes are used for protein analysis and in general 3-5 microliters provides more than enough serum for protein analysis. This must be applied approximately l9 one-fourth of the bridge distance away from the anode, as the protein will migrate toward the cathode. The glass supporting— rods are conveniently located in this position. Again, the sample must not extend to the edges of the strip. Staining methods for cellulose acetate strips. After the migration is completed, the wet strips are laid on filter paper and placed for 15 minutes in an oven preheated to 110° C. This temperature will cause the strips to curl slightly and this curling should be ignored at this time. Various dyes may be used to locate the protein on the strip. A solution of Ponceau "S" described by Smith (22. £13.) is the stain of choice in this investigation. It con- sists of 0.2% Ponceau "S" in 3% trichloracetic acid. The uptake of this dye is stated by Smith (pp; 313.) to be quantitative to the amount of protein on the strip. Brackenridge (22;.SlE-) using Light Green in an involved staining procedure was able to estimate the total protein more accurately than by the biuret method. Alcoholic bromphenol blue is totally unsatisfactory and in this respect, all alcoholic solutionsshould be avoided. Any contact with alcohol must be followed by a thorough rinse in distilled water to prevent irreversible curling. Staining is accomplished by laying the strips on the surface of the stain and allowing the stain to soak into the strip from below. This allows the air to be displaced and all areas of the strips should be devoid of air before the 20 strip is plunged beneath the surface of the buffer. This is the same procedure as used for impregnating the strips with buffer solution. After staining, the excess dye is removed from the strips by repeated rinses in 5% acetic acid. While the strips are still wet, they are laid between sheets of filter paper and pressed overnight. Cellulose acetate strips may be stained for glycoproteins and lipoproteins. Evaluating ElectrOphoretically Separated Serum Proteins Some investigators believe that the quantification of serum and/or plasma protein fractions is unnecessary. These investigators believe that only drastic changes in the rela- tive quantities of serum and/or protein fractions are of interest. In example, Fujiya (pp. 313.) stained the protein fractions with bromphenol blue, removed the excess dye, developed the maximum color intensity with ammonia vapor, and then impregnated the strips with paraffin. After compen- sating for the optical density of the paper strip and the paraffin, the optical density of the dye was recorded at 2 mm intervals along the strip. From these data, the electro- phoretic patterns were determined. This method obviously precludes the evaluation of subtle changes in plasma proteins. Another method of evaluation is the elution of the stain from the individual fractions. The eluted dye in a standard volume is then evaluated in a colorimeter or 21 spectrophotometer and compared to a standard curve to obtain the grams of dye present. Thus the per cent of total protein in each fraction can be quantified for statistical analysis. Two major difficulties are encountered with the elution- evaluation method. First, many volumetric centrifuge tubes are necessary to accommodate the fractions from several strips. Accompanying this difficulty is the increased number of man-hours required to evaluate each fraction of the several strips. A second difficulty experienced in the elution of paper strips is that paper fibers contribute to the optical density of the solution. These fibers are very difficult to remove even with prolonged centrifuging. Cellulose acetate yields no fibers and, therefore, does not present this dif- ficulty. Strips stained with Ponceau S may be eluted in 0.1 N. NaOH followed by an equal amount of acetic acid. Strips stained with bromphenol blue are eluted in 0.5% sodium carbonate in water. The solution of the previous problems is found in the use of a scanner which both draws the electrophoretic pattern and accurately records the area under each curve. This eval- uation method was used throughout this study by utilizing the Beckman Analytrol scanner. To scan the bromphenol blue stained paper strips, the following procedure is used: (1) At a point three-quarters of the distance between the anode end of the strip and the first protein band, make a thin pencil mark parallel to the 22 bands. This will serve as a reference point. Creasing the paper is equally effective. (2) At the time the strips are placed in the ammonia bath, the power and lamp switches of the scanner are turned to "on." (3) While the maximum color intensity is developing in the ammonia bath, select the slit width, remove any dust from the interference filters, place the recording pens in position on the cable riders and fill the pens with ink. (4) After the strips have been exposed to the ammonia for 15 minutes, a strip is placed in the scanning channel and the pen switch is turned "on." The strip Should be placed so that the light beam is between the anode end of the strip and the pencil mark. (5) A recording chart is placed in position and the traveling pen is adjusted to zero. This pen is then calibrated with the use of a 0.9 neutral density filter. (6) Turn the motor switch to "on" and scan the strip. Turn the pen switch and motor switch to "off" before the paper strip passes out of the light path. Each strip was individually calibrated and recorded using the above procedure. ,The analytrol scanner requires Specific adaptation and adjustments if gly00protein and lipoprotein stained strips are to be scanned. The reader is referred to the operators' manual for these specific instructions. An error in these instructions is that two, not one 0.9 neutral-density filters are required for scanning gly00protein stained strips. With- out using two such filters, the potential differences between the photoelectric cells cannot be balanced. 23 To scan cellulose acetate strips stained with Ponceau S stain, the 0.5 mm Slit width is used. However, the slit height must be reduced to approximately 0.5 cm. The previous interference filters (500 mu) may be used, but a 0.9 neutral density filter must be placed in front of the rear photo— -electric cell. Clearing of the strips with immersion oil, or other reagents appears to be unnecessary. Cellulose acetate strips must be mounted on some media such as filter paper strips, or on plastic strips which have one side adhesive, Otherwise, the drive wheel will tear the cellulose acetate. Also, it is best to place the paper strip between the acetate and the photoelectric cell to prevent the pressure foot from tearing the acetate strip. Results obtained on cellulose acetate stained with Ponceau S differ slightly from the results obtained from paper strips stained with bromphenol blue. This is to be expected and is the product of reduced chromatographic capture and sharper resolution. In this study, cellulose acetate was used to separate the plasma samples obtained from invertebrates. It also was used to provide a check on the quality of the separation obtained on filter paper. Evaluation of the Electrophoretic Method It was of primary importance to the study as a whole to ascertain the limitations of the electrophoretic method used. In particular, it was necessary to know if the 24 results obtained from a single 10 cmm portion of serum accurately characterized the total serum in that sample. Serum samples obtained from yellow bullheads were pooled to obtain approximately 2 ml of hemoglobin—free serum. This was mixed and centrifuged to remove any protein which might have been precipitated by antigen reactions. Sixteen paper strips were moistened with commercial vernol buffer (pH 8.6, ionic strength 0.075) in the previously described manner. A 10 microliter portion of the pooled serum was applied to each strip. A constant current of 0.104 MA/cm width of strip was applied for 14 hours. At the end of the separation, the protein on the strips was de-natured by placing the strips for 30 minutes in an oven heated to 110°C. These strips were then stained using the alcoholic bromphenol blue method and evaluated in the analytrol scanner as pre- viously described. The fractions of protein on these strips separated poorly and the commercial buffer was considered responsible for this difficulty. Nevertheless, it was decided to analyze these data. It was believed that these data would be useful since it would indicate the reproducibility of data obtained under adverse conditions. Due to the poor separation of the fractions, it was necessary to group the data into the four most prominent fractions. By doing this, certain fractions were evaluated as if they were only one fraction. Thus, for the purposes of this evaluation, this serum will be considered to consist of only four protein fractions. 25 A statistical analysis of the distribution of protein in the four fractions of pooled yellow bullhead serum is sum- marized in Table 1. It can be seen that the indicated quantity within each fraction varies slightly from portion to portion of the serum. In spite of this variability, the standard error of the mean of each fraction was less than 0.5%. This variability is obvious when the coefficients of variation are compared. It can readily be seen that varia- bility is inversely related to the quantity of protein present within a given fraction. Thus the data obtained from fractions having low protein concentrations are likely to be less accurate than the data from fractions having higher protein concentrations. Several factors contribute to the above relationship. Probably the most important factor and the most difficult to control is the human error involved in the delimination of a fraction's boundaries. This problem becomes particu— larly acute when the fractions are not resolved into distinct, well separated zones. To test the influence of the analytrol scanner on the data, one strip of rainbow trout (Salmo gairdnerii) serum was scanned 14 consecutive times. The results of the statistical analysis are summarized in Table 2. Variability again de- creased with increasing protein concentration. Although the width of the range for each fraction was 2% or less, in general plus or minus 0.5% constituted the range from the mean. The largest standard error of the mean of any fraction 26 TABLE 1. Summary of data obtained from electro- phoretic analyses of sixteen portions of pooled yellow bullhead serum. Fraction Fraction Fraction Fraction IIIII III-III IIIIIII IIIVII Mean 26.8 33.9 35.7 3.6 Variance 3.044 3.114 2.927 0.417 Standard Deviation 1.754 1.765 1.711 0.646 Standard Error of the Mean 0.436 0.441 0.428 0.162 Coefficient of Variation 6.54 5.21 4.79 17.94 Range 23.7—30.0 30.8—37.0 32.5-38.8 2.2-4.6 Width of Range 6.3% 6.2% 6.3% 2.3% 27 TABLE 2. Summary of data obtained from fourteen repeated scannings of the same strip containing rainbow trout serum protein. Fraction Fraction Fraction Fraction Fraction III" "II" "III" "IV" "V" Mean (x) 5.6% 14.2% 16.1% 22.5% 41.1% Vari nce (s ) 0.159 0.215 0.118 0.318 0.584 Standard Deviation(s) 0.399 0.464 0.344 0.564 0.764 Standard Error 0.106 0.124 0.092 0.150 0.764 Coefficient of Variation 7.1 3.27 2.14 2.50 1.86 Range 5.0—6.0 13.7-15.5 15.4—16.4 21.8—23.2 4.00-42. Width of Range 1.0% 1.8% 1.1% 1.4% 2.1% 95% Confidence Limits of X" 5.42% 13.93% 15.86% 22.22% 40.62% 5.88%- 14.45%- 16.26%— 22.86%— 41.51%- 28 observed was 0.204%. Again, human errors in delimiting the protein fraction's area were probably the main source of variability. In this aspect of the electrophoretic analysis, experience is most important if variability is to be decreased. The reliability of the Analytrol unit to record the area under a given curve using different slit widths was also tested. Matched observations on 16 different strips using 2.0 and 1.0 mm Slit width were performed. The mean of the differences obtained were not significant at the 95% level. However, the appearance of the curves obtained using different slit widths is different. Curves drawn using the 2.0 mm slit width are considerably smoother than curves drawn using the 1.0 slit width. However, the latter may be of greater use since it is more sensitive to variations. Since the Analytrol scanner draws the curves for each fraction and gives the area under each curve by integration, it was decided to attempt to use the area of the total curves to estimate the total protein concentration in terms of grams per 100 m1 of serum. Samples of bovine serum albumin, alpha, beta, and gamma globulin were applied at various known concen- trations to the strips, migrated a short distance and then fixed, stained, and evaluated. The relationship of area under the curve to protein concentration for a given electrophoretic run was found to be linear. This is depicted in Figure 2 for bovine beta globulin. Two balancing cams were used to evaluate the curves and the B-5 cam yields data less variable than the B-2 cam. The B-5 3500 -—1 30.0__ 25.0 a 20.0- CM 15.0 a 10.0'— 5.0 - Figure 2. Uptake of bromphenol blue by bovine betaglobulin as evaluated by scanning in a Spinco Analytrol. [fiji‘fiITrllTjiflfi l 2 3 4 5 6 7 8 9 10 Grams of Betaghlbulin per 100 ml ’ (6 Lambda sample applied) 29 30 cam yields curve area linear to the amount of colored material on the strip, and is corrected for Beer's law. The B-2 cam yields curve areas linear to optical density and it is.also corrected for Beer's law. On different electrophoretizruns, the values obtained for curve areas were found to be proportional to the amount of protein on the strip. .However, the "Y" intercept value was found to change considerably. The slope of the regression line appeared to be so similar for each electrophoretic run that it was not tested statistically. Therefore, it was con— cluded that the area under the curve could not be used to estimate protein concentration when staining with bromphenol blue. However, this relationship does allow the determination of the per cent of the total protein constituted by a single curve (fraction). If one wishes to convert these values to grams per 100 m1 of serum, the total protein concentration must be determined by a method such as the biuret method. The differences observed for samples evaluated on dif- ferent electrophoretic runs may be due to variation of the sample volume, differences in the temperature of fixation, differences in protein concentration, or other factors. Sample volume is strongly suspected to be the main contributor although it is known that fixation temperature is extremely important. The latter was difficult to control in the oven used for this study. 31 Methods for the Collection of Blood Methods for the collection of blood samples from fish were explored and cardiac puncture was found to be the best method. Direct cardiac puncture consisted of exposing the heart by an incision and then inserting a hypodermic needle connected to a 1 m1 tuberculin syringe directly into the bulbus arteriosus. Indirect cardiac puncture consisted of probing for the heart after one has ascertained its position. The former method destroys the specimen but the latter method generally incurs only low mortality. A #22 guage hypodermic needle is preferred over a #26 gauge, since the former do not plug easily. Coating the needle with mineral oil helps to reduce clots from forming in the needles. The blood should not be expelled through a hypodermic needle as this tends to induce hemolysis. Attempts to draw blood from the dorsal aorta on Centrarchids, Ictalurids, and Percids were unsuccessful, probably due to the small Size of the specimens, but worked well on Salmonids and Carp. Collection of blood by severing the caudal peduncle offered no advantage since it bled poorly, clotted rapidly, destroyed the specimen, and had the risk of contaminating the blood with mucous and body fluids. Collection of blood from crayfish (Orconectes pr0pinquus) was accomplished by inclining the hypodermic needle toward the pericardial cavity after entrance at the dorsal soft— membrane between the cephalothorax and the first abdominal segment. 32 Collection of blood from larval insect forms such as those of the dobson fly (Corydalus cornutus), burrowing mayfly (Hexagenia limbata), caddis fly (Hydropsy_che 32.), and midge larvae was accomplished by drawing a melting-point determin— ation capillary tube into a fine point, rinsing it with anti- coagulant, and inserting it under the skin of the abdomen. After the initial filling by capillary action and hydrostatic pressure, it may be necessary to apply gentle pressure to the general body surface of small larvae to obtain an adequate blood sample volume. The decision of whether to investigate plasma or serum must be decided by the investigator in view of the aspects which are to be investigated. In cases where the cellular components of the blood are of no interest to the investi— gation, serum will suffice. However, the use of anticoagulants to provide the investigator with plasma allows the determin- ation of both cellular and non-cellular components. Since fish and invertebrate blood clots very rapidly at summer temperatures, it is often useful to prolong or prevent the coagulation of such blood samples. In the case of inverte- brate blood it is always necessary to use an anticoagulant to obtain plasma since their blood apparently will not under- go clot retraction to express the serum. 7 The selection of an anticoagulant is again a matter of choice. Larsen and Snieszko (1961) have compared various anticoagulants and consider the practice of coating syringes and capillaries with 10% heparin to be the best method. 33 However, it must be recognized that heparin is a polysaccharide and its use may preclude the evaluation of glycoprotein in the plasma. In this study both 10% heparin and 10% citrate were used and citrate was selected for most of the work. Citrate does not need refrigeration and remains potent indefinitely. However, heparin is the choice for coating microhematocrit capillaries. Although only 0.3 ml of whole blood are needed for analysis of microhematocrit, total non-cellular protein con— centration, hemoglobin concentration, red blood cell counts, white blood cell counts, and electrophoretic analysis of the plasma protein, as much blood as possible was always obtained. After the blood sample has been collected, the needle should be detached from the syringe. Samples for microhema- tocrit, hemoglobin concentration, and both erythrocyte and leucocyte counts are taken from the syringe. Smears for visual inspection of the blood also may be taken at this time. The remaining blood is gently expelled into a narrow centri- fuge tube for subsequent separation of the cellular and non— cellular protein by centrifuging. Both the microhematocrit tube and the centrifuge tube should be centrifuged as soon as possible since whole blood tends to interact with the glass and results in some hemolysis and/or clotting. However, it is not necessary to read the hematocrit tube or separate the supernate immediately. Coating the glass centrifuge tube with neutral paraffin or use of plastic centrifuge tubes will eliminate this problem. 34 Adjustment of the total non—cellular protein concen- tration of invertebrate blood appears to be necessary for some species. Reduction of protein concentration to the range of 3—5 g/100 ml can be accomplished after the total non-cellular protein concentration has been determined. This adjustment assures linearity between protein concentration and the dye uptake. Although no attempt was made to evaluate the differ- ences which may exist between fish anesthetized with l:60,000 tricane methyl sulfonate (MS 222) and those not so anesthetized, there is no evidence that significant differences exist. All fish used in these experiments were so anesthetized. 0n the basis of taking blood samples from many species, two relationships seem apparent: (1) larger fish yield larger quantities of blood by cardiac puncture; (2) if species differences alone are considered and size differences are relatively constant, one can form a rank-order of apparent, relative blood volume based on the ease and quantity of blood obtainable. This is depicted in Figure 3. This rank-order is based on subjective, empirical observations and may not be related to total blood volume. Peripheral dialation, changes in blood pressure, and other systemic factors could strongly alter the quantity of blood obtainable. Such influences are presumed to have been at a minimum since the specimens were anesthetized at the time of blood sampling. 35 .oESHo> oooap Scotsman ow oocmgoHOp COHQSHHOS HSOHSHQES mo dacmcofipmaom .m ossmfim 36 ooosaaam soHHo» omosaaam sooam pmonaasm czopm anemone Hostage anaeoaoo LoxoSmnsno season Seacoaceao caches: oocampno moESHo> owned coeds- soflos Smfimcsm :oEEoo :mHMCSm comma exam csososoz Sofiaoaceao scones; to tea; oosawcoo mosaso> eased: Hopxofim mmmsw HHHwSSHm oaoomso xowam mmmm EudoEHHme mmmm EpsoEowsmq Seacoaceao sea: casewooo mossHo> HHmSm \ \ \ . Pollution Intolerant Pollution Tolerant 37 It readily can be seen that the "pollution tolerant" species yield the largest blood sample volumes. For example, bullheads of size 15—20 cm consistently yield samples of at least 1.0 ml of blood. By contrast, largemouth and small- mouth bass of Size 20—30 cm yield only 0.5 m1 of blood, or less. Shell (1962) reported a similar observation for small- mouth bass.l If this relationship is indicative of blood volume, then bullheads would have a much larger antibody pool than the largemouth bass. Prosser and Brown (pp, 913,) point out that smaller blood volumes are more efficient in transporting oxygen, et cetera, but that a larger blood volume would be useful if it also served as a source of antibodies and phagocytic cells. ELECTROPHORETIC PATTERNS OF FISH AND INVERTEBRATES A survey of the electrophoretic patterns of twenty- two Species of freshwater fish and two species of inverte- brates was conducted. Specimens used for these determin— ations are presumed to be representative of the Species. In most cases, the Specimens were taken from their natural environment and acclimated to the laboratory for several days. Some Specimens were sacrificed at the collection site to obtain blood. The electrophoretic patterns were deter— mined using the previously listed conditions for paper electrophoresis. These patterns were then traced and are shown in Figures 4 through 29. While comparing the following electrophoretic patterns, no attempt should be made to compare the various curves on the basis of curve height alone. The area of a given curve is indicative of the relative quantity of protein therein. Therefore, it is permissible to compare the areas of curves (fractions) on a Single electrophoretic pattern, but it is not permissible to compare curve areas on different electro- phoretic patterns. After curve areas have been converted to relative values (per cent of total protein) or absolute values (grams per 100 ml), they are readily comparable. Also, there is no evidence that proteins of similar electro-chemical characteristics have similar physiological or biochemical characteristics. Figure 4. Electrophoretic pattern of human plasma proteins. (For comparison to fish plasma and/or serum proteins.) 39 I I I I I I fibrinogen gamma beta alpha globulins albumin globulin globulin I Denotes the application point of the sample on the paper strip. Electrical field is positive to the right and negative to the left of this point. 40 Figure 5. Electrophoretic pattern of the Bowfin (Amia calva). I II III IV V Fractions Per Cent of Total Serum Protein (three specimens) 1 II 111 1V V 11.6% 32.5% 30.2% 16.2% 9.3%1 11.1% 31.4% 31.4% 18.5% 7 4%1 14.4% 29.1% 35.5% 20.0% 6.0%2 lThese fish were acclimated to the laboratory. 2This fish was collected and sampled during winter-kill conditions. 41 Figure 6. Electrophoretic pattern of the plasma proteins of the Rainbow Trout (Salmo gairdnerii gairdnerii). I II III IV V VI Fractions Per Cent of Total Plasma Protein (five Specimens)l 1 11 111 1V V VI 6.5% 15.7% 10.5% 14 4% 40.7% 11.8% 4.3% 11.5% 9.3% 20.1% 46.0% 8.6% 5.0% 13.0% 11.0% 18.0% 40.0% 13.0% 6.1% 14.2% 12.2% 19.3% 38.7% 9.1% 5.1% 11.0% 13.2% 18.3% 41.1% 11.0% lThese specimens were acclimated to the laboratory. 42 Figure 7. Electrophoretic pattern of the plasma proteins of the Kamloops Trout (Salmo gairdnerii kamloops). 1 II III 1v \7 VI VII Fractions Per Cent of Total Plasma Protein (one Specimen)l II III IV V VI VII 2.8% 3.8% 16.1% 34.2% 18.0% 20.9% 3.8% lThis Specimen was acclimated to laboratory conditions. 43 Figure 8. Electrophoretic pattern of the serum proteins of the Brown Trout (Salmo trutta). /\ I II III IV V VI Fractions Per Cent of Total Serum Protein (four specimens)l I II III IV V VI 4.1% 9.4% 29.5% 17.7% 27.2% 11.8% 4.3% 13.6% 34.7% 19.8% 18.6% 8.6% 2.9% 13.7% 29-995 18«6% 26.9% 7.7% 3.1% 11.1% 24.4% 19.6% 29.7% 11.7% lThese specimens were collected and sampled at the side of their hatchery raceway. 44 Figure 9. Electrophoretic pattern of the serum proteins of the Brook Trout (Salvelinus fontinalis). I II III IV V VI Fractions VII Per Cent of Total Serum Proteins (four specimens) 1 11 111 1v V VI VII 6.5% 13.0% 25.2% 15.2% 20.2% 16.6% 2.8% 5.2% 9.7% 28.4% 16.5% 20.3% 15.7% 3.7% 3.7% 8.3% 28.7% 13.6% 21.2% 21.2% 3.0% 3.5% 7.7% 28.9% 11.9% 25.1% 19.1% 2.9% These specimens were collected and sampled at the side of their hatchery raceway. 45 Figure 10. Electr0phoretic pattern of the serum proteins of the Western Grass Pickrel (Esox americanus vermiculatus). I I II III IV V VI ‘ Fractions Per Cent of Total Serum Proteins (five Specimens) I 11 111 IV V VI 4.4% 23 0% 15.0% 35.3% 15 0% 7.0%1 2.0% 20.0% 17.0% 39 0% 15.0% 7.0%1 1.9% 23.1% 26.4% 29.8% 13 9% 4.6%1 4.1% 20.8% 12 5% 43.3% 13.3% 5.8% 2.7% 15.4% 15.4% 45.4% 7.2% 4.5 lThese fish were acclimated to the laboratory. Figure 11. Electrophoretic pattern of the Northern Pike (Esox lucius). I II III IV V Fractions 1 Per Cent of Total Serum Protein (six specimens) I II III IV v 6.4% 17.6% 38.2% 31.7% 5.8% (winter) 7.6% 25.1% 38.4% 24.4% 4.1% (winter) 8.7% 16.2% 12.5% 8.7% 53.7% (summer) 7.5% 18.9% 42.4% 26.5% 4.5% (spring) 6.1% 18.4% 39.2% 30.0% 6.1% (spring) 8.7% 17.5% 38.5% 29.8% 5.2% (Spring) 1All specimens were sampled at the collection site. Figure 12. Electrophoretic pattern of Goldfish serum (Carassius auratus). I II III IV Fractions Per Cent of Total Serum Protein (nine specimens)1 I II III IV 18.6% 16.6% 10.2% 54.4% 8.5% 8.5% 25.6% 57.3% 9.4% 9.4% 24.3% 56.0% 7.0% 5.0% 25.0% 62.0% 6.7% 8.3% 16.8% 68.0% 7.0% 9.5% 25.0% 58.0% 12.0% 7.3% 18.3% 62.0% 12.0% 10.3% 14.4% 65.0% 9.8% 8.8% 18.8% 63.0% lThese specimens were acclimated to the laboratory. 48 Figure 13. Electrophoretic pattern of the plasma proteins of the Carp (Cyprinus carpio). I II III IV V VI Fractions Per Cent of the Total Plasma Protein (two specimens)1 I II III Iv V VI 16.2% 12.6% 10.2% 13 2% 10.1% 37.6% 24.2% 14.6% 12.1% 9.7% 9.7% 29.1% lThese specimens were acclimated to the laboratory. 49 Figure 14. Electrophoretic pattern of the serum proteins of the Lake Chubsucker (Erimyzon sucetta). I II III IV V VI Fractions Per Cent of Total Serum Protein (one specimen)l I II III IV V VI 12.2% 23.0% 16.0% 10.8% 25.0% 12.8% lThis specimen was collected and sampled during winter-kill conditions. 50 Figure 15. Electrophoretic pattern of the plasma proteins of the Northern Hog Sucker (Hypentelium nigricans). * /\ I II III IV V VI VII VIII Fractions Per Cent of Total Plasma Proteins (two specimens)1 I II III IV V VI VII VIII 13.2% 12.4% 18.6% 8.2% 18 6% 7.4% 18.2% 3.3% 12.9% 17.9% 14.6% 5.4% 17.1% 6.6% 20.9% 4.1% lThese Specimens were collected by electro-shocking and blood was collected at the sample Site, June, 1962. 51 Figure 16. Electrophoretic pattern of the serum proteins of the Black Bullhead (Ictalurus melas). I II III IV V VI Fractions Per Cent of Total Serum Protein (two specimens)1 I II III IV v VI 13.9% 10.8% 10.8% 8.5% 9.1% 56.8% 6.7% 5.7% 7.6% 7.6% 4.8% 67.3% lThese specimens were acclimated to the laboratory. 52 Figure 17. Electrophoretic serum pattern of the Yellow Bullhead (Ictalurus natalus). I II III IV V VI VII Fractionsl Means of Data Obtained from Seven Specimensl I II III -+ IV V VI VII 18.29% 9.10% 28.91% 14.92% 25.67% 2.74% 1These data obtained from the control fish (Appendix I). Figure 18. Electrophoretic pattern of the serum proteins of the Brown Bullhead (Ictalurus nebulosus). I II III IV V Fractions Per Cent of Total Serum Proteins (one Specimen)l I II III IV V 21.5% 24.4% 14.7% 38.6% 4.5% 53 lThis Specimen was collected and sampled during winter—kill conditions. Figure 19. Electr0phoretic pattern of Channel Catfish serum (Ictaluris punctatus). 1 I II III IV V VI Fractions Per Cent of Total Serum Protein (one specimen)l 54 #— I II III IV V VI 14.5% 11.2% 9.6% 12.9% 21.2% 30.1% 1 This specimen was acclimated to the laboratory. 55 Figure 20. Electr0phoretic pattern of the plasma proteins of the Rock Bass (Ambloplites rupestris). I II III IV V VI VII VIII Fractions Per Cent of Total Plasma Protein (six Specimen)l II III IV V VI VII VIII I...) IU-IZ‘ID .4% 5.8% 14.9% .2% 11.0% 14.7% .5% 6.6% 14.5% .2% 6.0% 16.2% .2% 9.0% 16.9% .7% 9.0% 20.8% .4% 18.4% 23.3% 16.4% 12.9% .1% 11.0% 29.4% 15.9% 10.4% .1% 16.5% 26.9% 17.0% 11.3% .1% 26.2% 24.5% 12.2% 6.1% .2% 15.1% 21.8% 17.5% 7.8% .8% 14.9% 22.6% 17.1% 6.7% Ul'\IO\UlO\U'I 1These specimens were collected by electro-shocking and their blood was drawn at the collection Site with the use of heparin. Collection was made June 27, 1962. 56 Figure 21. Electrophoretic pattern of the serum proteins of the Green Sunfish (Lepomis cyanellus). l W I II III IV V VI Fractions Per Cent of Total Serum Protein (two Specimens)l I II III IV V VI 24.0% 14.8% 20.3% 18.5% 11.1% 11.1% 11.8% 14.0% 7.4% 29.6% 22.2% 14.8% lThese Specimens were collected during winter-kill conditions, and may not be typical of the species due to this stress. 57 Figure 22. Electrophoretic serum pattern of the Pumpkinseed (Lepomis gibbosus). I II III IV V VI VII VIII Fractions Per Cent of Total Serum Protein (one Specimen)l I II III IV V VI VII VIII 2.6% 5.3% 17.4% 10.7% 26.8% 12.2% 8.7% 16.1% lThis Specimen was acclimated to the winter conditions in a lake and sampled there. 58 Figure 23. Electrophoretic pattern of the serum proteins of the Bluegill (Lepomis macrochirus). I II III IV V Fractions For the per cent of total serum proteins see Appendix I, Table 14, control specimens. 59 Figure 24. Electr0phoretic pattern of the serum proteins of the Smallmouth Bass (Micropterus dolomieui). I II III IV v VI'VB: VIII Ix Fractions Per Cent of Total Serum Protein (three specimens)l II III IV V VI VII VIII IX 1.9% 1.8% 4.5% 10.7% 13.7% 9.8% 16.6% 7.8% 10.7% 11.7% 16.6% 12.8% 13 7% 8.2% 16.5% 9.1% 9.1% 11.9% 16.5% 10.0% 14.5% 8.2% 15.5% 4.5% 9.0% 12.7% 16.3% lThese specimens were acclimated to the laboratory. 60 Figure 25. Electrophoretic pattern of the serum proteins of the Largemouth Bass (Micropterus salmoides). I II III IV V VI Fractions Per cent of total serum proteins see control specimens Appendix I. 61 Figure 26. Electrophoretic pattern of the serum proteins of the Black Crappie (Pomoxis nigromaculatus). I II III IV V VI Fractions Per Cent of Total Serum Protein (two specimens)1 I II III IV V VI 5.4% 19.0% 25.0% 14.5% 13.6% 10.9% 6.4% 18.4% 35.2% 16.0% 12.8% 11.2% lThese Specimens were acclimated to the laboratory. 62 Figure 27. Electrophoretic pattern of the plasma proteins of the Yellow Perch (Perca flavescens). I I II III Fractions Per Cent of Total Plasma Protein (nine specimens)1 :7 I II III 42.6% 48.2% 9.0% 35.5% 49.4% 15.0% 34.3% 54.6% 11.0% 40.7% 45.1% 14.2% 40.5% 48.4% 11.1% 41.9% 43.0% 15.1% ‘ 41.3% 40.7% .‘ 18.0% 40.9% 46.8% 12.4% 31.2% 54.7% 14.1% lThese Specimens were acclimated to the laboratory. Figure 28. ElectrOphoretic pattern of the plasma proteins of the freshwater crayfish (Orconectes prppinquus). . I II III Fractions Per Cent of Total Plasma (one Specimen) I II III 9.3% 11.8% 78.8% Figure 29. Electrophoretic pattern of plasma proteins of the dobson fly larvae (Corydalus cornutus). I II III IV V Fractions Per Cent of Total Plasma (one Specimen)1 64 I _ . II III IV v 2.0% 36.0% 47.2% 14.3% 0.6% lAcclimated to the laboratory. 65 The migration of each fraction is not a true function of time and because the electr0phoretic separation conditions vary slightly from separation to separation, the position of each fraction for a single species will vary slightly from one electrophoretic migration to the next. Also, there is some variation from specimen to specimen. In spite of these variations, the number and position of the fractions allows species identification. imilar results were obtained by - Lillevik and Schloemer (1961) who used the free proteins of skeletal muscle to show species di ferences. An attempt was made to determine the morphological differences which distinguished rainbow trout (Sglmg_ gairdneriigairdnerii)from kamloops trout (Salmo gairdnerii kamloops), based on a taxonimic key by Schultz (1938). This failed to distinguish between specimens of kamloops trout purchased from a licensed fish breeder, from rainbow trout obtained from the Michigan Conservation Department hatchery at Wolf Lake. The serum proteins of these specimens were analyzed electrophoretically by the methods previously listed. Tracings of the electrophoretic patterns of rainbow trout and kamloops trout are shown in Figures 6 and 7. It can readily be seen that their electr0phoretic patterns differentiate these sub- species. In addition to having a different number of frac— tions, these fractions contained considerably different pro- portions of protein, gly00protein, and lipoprotein. 66 The electrophoretic patterns of one species of immature aquatic insect and one species of crayfish were determined, and these appear to be species specific also. Van Sande and Karcher (1961) found species differences in the hemolymph protein of insects. Ball and Clark (1953) reported species differences in the amino acids of Culex mosquitoes. This procedure may be useful to aquatic entomologists in constructing taxonomic keys to the immature insect larvae, by associating the electr0phoretic patterns of immatures with adults. Protein fraction I when separated using the described conditions is relatively immobile due to its iso-electric character. In this respect it conforms to the description of human gamma-globulin known for its antibody power. The means of fraction I for the various species (derived from Figures 5 through 27) were arranged in rank-order. This rank—order is presented in Figure 30 and proceeds from lowest to highest quantities of fraction 1. When the species are compared by this rank-order, it can be seen that empirically-determined pollution—intolerant fish have low quantities of fraction I. By contrast, empir— ically—determined pollution—tolerant fish have large quantities of this fraction. Ictalurids contain large quantities of this fraction and are well known for their pollution tolerance. Centrarchids and salmonids usually have less than half of the quantity of fraction I normally present in bullhead serum. 67 one .Sopmz 08mm one CH mCH>HH p02 mcoEHooamH H 0p msflpmooom mofloogm Sham mo Sopsouxsmm .Apowmsp Hospo>v w.w mg pm Essen So\pcm mammao Sacha CH Campopg oHSpooHouomH wo pcoo Sod cmoE .om onsmam I I I l R R bi R C) wx <3 U\ cl #1 H utaqoad oIJqoatg—osl JO queo Jag Increasing "Empirical" Pollution Tolerance = Serum ** Plasma *= 68 **DBGQIIHE UMOJH *daeo **peeqttng MOII9X **HSTJURS UGGJD **HSIJQEO IGUUBQO *JaxonsSoH **UTJM08 **JSXOUSQUHO 9X91 *PPGQIIUS MOPIS **H9TJPIOD **IIIBGNIS **sseg uqnomeSJeq **BMT& UJGQQJON **9Tdd9JO x0212 *qnoal moquteg 4*QUOJL XOOJH **lnoam umoag **I3JXGTa SSBJD *qnoag sdootwex *SSSBH WQNOMIIBWS 9*paasutxdmn3 *sseg xoog *UOJad 69 .Yellow perch have no apparent iso-electric protein at pH 8.6 and may present an exception to this generality. This relationship of the quantity of iso-electric pro- tein to empirical pollution tolerance is based on small sample Sizes. .However, this relationship appears to occur with regularity too great to have occurred by chance alone. If this protein fraction is related to the "hardiness" of the species it may be due to the environment or to genetic dif- ferences. Stress tends to increase the quantities of the iso-electric protein in fish and the above listed relation- ship may be due to the stresses of the environment. Thus, bullheads could be comparable to other fish living in the same water conditions. A comparison of the common warm—water fish indicates that it is not the gross condition of the water which is responsible for high levels of this iso-electric . protein. Fish of several species were held in clean tap-water for two months, with feeding. At the end of this period, they had not changed the relative quantity of this fraction. This relationship of pollution tolerance to the quantity of iso-electric protein (in vernol buffer at pH 8.6) may be of value in classifying species of fish in regard to their relative hardiness. Several reasons support this hypothesis .as follows: (1) this material is found in highest relative quantities in ”pollution" tolerant species; (2) since "pollu- tion" tolerant species appear to have larger blood volumes, and larger relative quantities of this fraction, the product would produce a larger absolute quantity of this fraction per 70 gram of body weight; (3) the abundance of this material ap- pears to be related to the tolerance potential of the species; and (4) this material tends to increase during "acute stress" periods. Although further information concerning the relation— ship of iso-electric protein plasma and/or serum to hardiness will be required, it appears possible to use this criterion as an aid in ranking hardiness in fish species. One possible use of this relationship would be in the selection of breeding stock. If one stock was found to exhibit higher quantities of this material than other stocks, providing this increase was not due to stress, then one could reason that this stock had a higher genetic—antibody-potential than the other stocks in question. By in-breeding fish which exhibit high quanti- ties of iso-electric plasma and/or serum, one might produce hardier offspring. These offspring, due to their "genetic antibody—potential" might be more able to withstand the dif- ficulties of hatchery or stream life. Mertz (gp;_git.) indicates that the optimum pH for enzyme activity is near its iso-electric pH. At the iso- electric pH the protein molecule exists in the di—pole (zwitter ion) form and retains its maximum number of reactive Sites. This suggests that the blood of pollution tolerant fish con— tains a greater quantity of proteins capable of maximum activity at or near physiologic pH. ARTIFICIAL STREAM SYSTEM An artificial stream system was constructed which would allow the investigation of the influence of a diurnal oxygen pulse on the serum protein fractions of fish. This system was so constructed that only the oxygen content of the ex- perimental section would be varied, and all other aspects of water quality would be approximately equal between the con- trol and experimental section. This condition was achieved by utilizing a re-circulating water system which was both drained into and supplied from a single reservoir. Figure 31 presents a schematic representation of this stream system. The control and experimental sections (A and B) were constructed of one~fourth inch thick acrylic plastic. Their measurements were; 30 cm deep, 60 cm wide, and 150 cm long. Inlets were subsurface and outlets were located at both surface and subsurface levels. In operation each stream contained approximately 225 liters of water. Water drained by gravity flow from the artificial streams through a nylon organdy filter and into the regener- ation reservoir (C). This reservoir was constructed of three—quarter inch marine plywood which was coated with plastic resin. Its dimensions are: 16 inches deep, 30 inches wide, and 96 inches long. Beginning with the entrance end, the reservoir contained four aeration compartments, which were 72 .Uowuomopoga pew «manpmgomEop wpseudo... cowzxopo>aommflmcmcfiHHOSpmoo Sow Empmmm EmoSpm Hmaoamflpmm wcflpmHSoSHolom .Hm ogsmflm 73 m Emogpm Hmpcoeflhogxm m Emogum Homecov 74 constructed to form water-falls. In each of these compart— ments, compressed air was forced through six air—stones (D). The result is much mixing and subsequent near saturation of the water with oxygen. The fifth and final compartment contained a flat, block-tin refrigeration coil (E) and a bronze-stainless steel sump-pump (F). The continuous flow of water passes over the refrigeration coil and the sensitivity of the thermostat VNMS such that the temperature of the water in both units remained within i 1° C. of 20° C. The flow of water from the pump (F) is divided between the control (A),the experimental (B) sections, and the re— generation reservoir (C). Before the water passes into the experimental section, it must pass through two oxygen- stripping columns (G), one of which is not shown. These columns were constructed of one-quarter inch thick, 4—inch inside-diameter acrylic plastic tubes, which measured four feet high. Each column was filled with unglazed ceramic raschig rings (H) to a height of approximately 42 inches. A constant head of pressure is maintained by overflow out— lets (1). During degassing, nitrogen gas is introduced into the bottom of these columns and allowed to bubble to the top. The rings help to control the size of the nitrogen bubbles and smaller rings produce smaller bubbles with pr0portion- ately greater surface area. 75 The removal of the oxygen from the water by nitrogen gas bubbles is effected by oxygen‘s greater solubility in nitrogen gas than its solubility in water. Brown (1957) points out that this process will not produce "nitrogen gas bubble disease" in fish. Each section was illuminated by two 40 watt fluorescent tubes (J) which were suspended approximately four inches above the lids of the control and experimental sections. An automatic timer switch controlled these lights, turning them on at 6:00 P.M. and off at 6:00 A.M. Black polyethylene sheets were used to shield exterior light from the stream system. It was found that two 40 watt tubes were too bright and caused the fish to be rather excitable. To correct this, a single layer of brown wrapping paper was laid on top of the section lids. The flow of water into each section was approximately 5 liters per minute. This flow was rapidly dissipated as the water entered the stream and the fish appeared to be in a resting condition at all times. Temperature, pH, alkalinity, total carbonate alkalinity, and dissolved oxygen were determined for the control and experimental sections during non-degassing periods, and the mean of the differences between the streams was not found significant at the 20% level. The same results were obtained during degassing with the exception of the dissolved oxygen content. 76 A float valve inlet (K) connected to the reservoir was installed in the fifth compartment to resupply water lost by evaporation and that water used in chemical tests. The Influence of a Diurnal Oxygen-Pilseon Bluegills The bluegills used in this experiment were collected by seining pools of the Lake Lansing drain, Ingham County, Michigan during May, 1961. These fish were held in the laboratory in aquaria for three weeks, during which time they were fed commercial fish-food pellets. Twenty-four specimens were placed in the artificial stream by non- random assignment in the following manner: pairs of nearly equal size fish were formed, one of each pair being placed in either the control or the experimental section. Acclim- ation continued for another two weeks before degassing began. Feeding continued throughout the experiment. Degassing consisted of reducing the oxygen content of the experimental section from an average of 8.3 ppm to an average of 3.5 ppm, eight hours per day, for nine days. Water samples were collected non-randomly during the non- degassing and degassing periods to provide a check on water quality. The means of the chemical data determined during degassing periods are listed in Table 3. On the tenth day specimens were sacrificed and blood samples were collected by indirect cardiac-puncture. Although citrate was used to coat syringes and capillary tubes, it failed to prevent the rapid coagulation of the 77 blood taken from the experimental specimens. The collection of serum then became the primary objective. TABLE 3. Summary of the water chemistry during the degassing periods for bluegills. Dissolved Total Temp. °C. Oxygen ppm pH Alkalinity ppm Control Mean 20.25 8.3 8.56 203 Range 19.0—21.o 8.0—8.6 8.4—8.8 192 — 216 Experimental Mean 20.25 3.5 8.52 201 Range 19.0-21.o 2.5—4.1 8.4—8.7 188 - 21o The samples were allowed to clot at room temperature for four hours. During this time, much of the serum was ex- pressed from the clots before centrifuging. Serum samples were then collected and stored in individual three-inch glass testitubes. A 10 microliter portion of each serum sample was electrophoretically separated using commercial vernol buffer and a fourteen hour separation period. The unused serum was stored at -5° C. until other portions could be separated and analyzed for glycoproteins. This storage precluded lipoprotein analysis. Five clearly distinguishable serum protein fractions were consistently separated. A sixth fraction was obtained 78 occasionally and consisted of separating fraction II into two separate fractions. This difficulty may have been due to genetic differences, but it is believed that the commercially prepared buffer powder was responsible for the poor separation. This buffer is prepared for the separation of human serum and such buffer is here considered to be totally unacceptable for the separation of fish serum. The positional relationships of the protein fractions, glycoprotein fractions and lipoprotein fractions of this species is depicted in Figure 32. The width of the frac— tions is not necessarily indicative of the amount of material contained therein. Rather, this illustrates the location of proteins which are conjugated to polysaccharides (glyco- protein) and/or lipids (lipoprotein). Three glycoprotein fractions and ikuu° lipoprotein fractions were separated. It should be noted that protein fraction I contained lipoprotein and did not appear to contain glycoprotein. Protein fraction II contained both glycoprotein and lipoprotein, as did protein fraction III and IV. Protein fraction V contained glycoprotein but did not appear to contain lipoprotein. The above results are important since they clearly demonstrate fundamental differences between bluegill serum and mammalian serum. Weil (pp, 223,) points out that in human serum the majority of the lipoproteins are carried in the beta fraction and that the glycoproteins are contained mainly in the alpha fraction. PPPPPPPP Glycoproteins 3 W8; point 79 80 Since male specimens predominated both the experimental and control groups, the serum collected from females was not statistically analyzed. Therefore, differences due to sexual dimorphism were not ascertained. Samples collected from bluegills during the winter of 1961-1962 indicate that such a condition exists. The statistical analyses of the influence of this diurnal oxygen-pluse onthe distribution of serum proteins in the control and experimental bluegill specimens are shown in Table 4. Protein fraction I most closely resembles gamma globulin in its electro-chemical characteristics. In general, the specimens of the experimental lot appeared to increase the relative quantity of this fraction. This increase was not statistically significant, but it is believed that a larger sample would have demonstrated statistical significance. The mean of protein fraction II was also higher in the experimental Specimens, but like fraction I this was not statistically significant. The variance of fraction II in the experimental fish was significantly different at the 95% level from the variance of the controls. The mean of fraction III in the controllot was sig- nificantly greater at the 95% level than the mean of frac- tion III in the experimental lot. The variance of this fraction was also significantly greater at the 95% level in the control lot than in the experimental lot. 81 .Hm>mH mo.mm the em memopm mapcmoamacwam* .H xHUCmqa< CH wmeH Rams.o gmmm.sz *Roo.m *essm.mm mmas.a moceeee> H.0H :.mm m.w H.0H m.mH coaseanw> no ecoaeanemoo momm.o Ramm.: emma.a Roam.m emsw.o new: was no pophm osmGCMpm am.m-a.s am.mm-o.mm am.mm-m.wa Rm.wm-m.:m Rm.aa-m.w mmcmm amm.w eoH.sm emo.am emm.mm mam.m cmm: HmpcmEHmexm amm:.o momQ.HH ammo.ma mom.: emmm.m eecmahe> m.m m.mH s.ma m.m m.ma eoapeatm> no enmaoaeemoo emm:.o awsm.m memm.m Roam.fi meza.a new: we» no poyhm Uhmpcmpm mm.m-m.s mo.mm-m.om ea.om-s.ma Rm.mm-:.om Rm.aa-m.s emcee aew.e awe.mm _*aee.em gee.mm aaa.a eemz Homecoo > >H HHH HH H coapomhm :oapomhm coapompm coHpompm coapowpm H.mHHme5Hn came mo 839mm one CH mcampOLQ mo soapsnflppmao one so whoamucmwmxo HprSHo m go moCmsHMCa mnp mo mmeESm .: mqmgB 82 The mean of fraction IV showed a general increase in the experimental lot and a general decrease in fraction V in the experimental lot was noted. Variance between the control and experimental lot for these two fractions was not statis- tically significant. Correlation analyses were performed to determine the relationship of these protein fractions to body length and weight. These correlation coefficients are listed in Table 5, and show the influence of low dissolved—oxygen. Fractions I and II have a significant negative correl- ation at the 95% level to body weight in the control lot. Fractions I and II were not significantly correlated to body length. Other fractions of bluegill serum protein were not significantly correlated to either body weight or body length. In general, the highest correlation coefficients were found in the relationship of the fractions to body weight. Previous mention has been made of the electro-chemical similarity between ”gamma globulin" and bluegill serum protein fraction "I." In the control lot this fraction is negatively correlated with both body length and body weight. While a "larger" specimen would be expected to have lower quantities of this fraction, the larger ones of the experimental lot tended to have more of this fraction. The stress from low-oxygen apparently reverses this relationship and the 'size" of the specimen appears to have 83 some influence on the amount of fraction "I" produced. This may be related to the lower metabolic rate of the larger fish. TABLE 5. Coefficients of correlation for the dis- tribution of serum proteins to total body length and body weight of bluegills.’ W m Fraction Fraction Fraction Fraction Fraction I II III IV .V Weight Control -.90%* -.87* +0.59 +0.40? -O.44 Experimental +0.14 -O.38 -O.35 +0.03 -O.63 Total Length Control -O.79 —O.79 +0.49 +0.08 -O.44 Experimental +0.19 +0.56 -0.l0 +0.08 -0.26 *Significant at the 95% level. It should be noted that the correlation coefficients of the control lot werg generally higher than those of the experimental lot. The latter were lower than those of the control lot in nine of the ten analyses. It also should be noted that a general correlation sign difference exists be- tween the control and experimental lots. However, in cases where this correlation sign-difference was absent, the cor- relation coefficient of the experimental lot was usually much lower than that of the control lot. Therefore, not only a reduced correlation, but also an opposite relationship con- trasted the control from the experimental lot. As such, this 84 suggests both quantitative and qualitative changes in serum protein fractions of bluegills subjected to a low dissolved- oxygen pulse. Three fractions of glycoprotein were separated in the serum of bluegills. In general, glycoprotein fraction I (Figure 32) was contained in protein fraction II, glyco- protein fraction II was contained in protein fraction IV, and glycoprotein fraction III was contained in protein frac- tion V. The distribution of glycoprotein in the serum fractions of the control and experimental lot was tested statistically, and the results are summarized in Table 6. The mean of fraction I waslfih8%ix1the control lot and 10.3% in the experimental lot. The mean of fraction II was 26.2% in the control lot and 35.0% in the experimental lot. The differ- ence between the means of the control and experimental group of fraction II was statistically different at the 95% level. The mean of fraction III was 70.8% in the control lot and 54.8% in the experimental lot. The difference between these means was also statistically significant at the 95% level. As previously noted, the protein in the glycoprotein fractions were not significantly different between the control and experimental lots. Thus the glycorprotein content of these fractions had changed in the experimental lot, without significantly altering the levels of the protein. To demon- strate this further, Pearsonian correlation analyses were performed on the protein to glycoprotein levels in each F) . o>m \. m: em so>mohm R: 1 moH H mHm w o + R e . ** .Ho>oH Rmm mzp pm hopmohm RprmoHMHumHmR Rm.mm Ro.:HH Rm.eH Rm.mmH RR.wq RR oH mecmHtm> a.mH m.om m.mm a.mH o.mm m.om (OHemHtm> wo pZmHOHMchOO Rnaa.m RRHH.R RHmm.m R::m.m 1+r.e RHmH.m new: one on wopsm Usmwcm ow RmmlH: Rom-Hm RmH-w RHm-o.mm Ro.mm-o.mH Rc.aH-o.© ememm Rm.am **Ro.mm Rm.oH *Rm. on Rm.mm Rm.oH new: HHH HH H HHH HH H .: QoHpomhm :oHpomhm COHpowhm COHHompm CoHpompm CoHpomgm EopH HmpcmEHhodxm Homecoo .mHHHmmzHQ onE mo Ezhom ogp CH mQHmposmooRH m Ro :oHpsxwmeHw mflp co omHSQICmmeo Hmntho m R0 cocosHmzH one mo RHmEEdm .m mqm