STUDIES ON BIOLUMINESCENCE By E. Newton Harvey Published in: The Journal of General Physiology DOI: 10.1085/jgp.1.2.133 STUDIES ON BIOLUMINESCENCE. VII. REVERSIBILITYOF TI~E PHOTOGENICREACTIONIN CYPRIDINA. BY E. NEWTONHARVEY. (From the Department of Marine Biology, Carnegie Institution of Washington, Washington,and thePhysiologicalLaboratory,PrincetonUniversity,Princeton.) (Receivedfor publication, September 3, 1918.) In a previous paper of this series,1I have described two photogenic substances in the ostracod crustacean, Cypridina hilgendorfii, which I called photogenin and photophelein. Photogenin is destroyed below the boiling point, is non-dialyzable, and prepared by making a cold water extract of the luminous animal and allowing it to stand in the air until no more light appears on shaking. This indicates that one of the photogenic substances, photophelein, has disappeared, leaving the photogenin. Photophelein is not destroyed by short boiling and will dialyze. It is prepared by making a hot water extract of the lumi- nous animal. The hot water destroys the photogenin, leaving the photophelein. Whenever two such non-luminous solutions are mixed light aopears. On the grounds of method of preparation, relation to temperature, and dialysis, I regarded photogenin as comparable to luciferase and photophelein as comparable to luciferin, two photogenic substances described by Dubois~ in the beetle, Pyrophorus noctilucans, and in the mollusc, Pholas dactylus. Dubois believes that luciferase is an oxidizing enzyme which oxidizes luciferin, an oxidizable substance, with light production. Neither luciferase nor luciferin alone in solution can produce light but light appears if solutions of the two are mixed and continues as long as any luciferin remains unoxidized. Dubois has also been able to produce light by oxidizing luciferin (alone) with a small crystal of KMnO4, by H~O~(with or without blood containing 1Harvey, E. N., Am. J. Physiol., 1917,xlii, 318. 2Dubois, R., Compt.rend. Soc. biol., 1885,xxxvii, 559. 133 The Journal of General Physiology 134 PHOTOGENIC P,,EACTION IN CYPRIDINA hemoglobin), BaO2, PbO~, and other oxidizing agents. Through the kindness of Professor Dubois I have received some material of Pholas dactylus preserved in sugar and I can confirm his results on the effect of KMnO4 and other oxidizing agents in producing light with luciferin of Pholas. I have likewise repeated my own experiments with the photophelein of Cypridina using a whole series of oxidizing agents applied in the same way, as with the luciferin of Pholas and, as pre- viously, have failed to obtain any light with this substance.* The difference in our results is, therefore, not to be referred to a difference in method of experiment but to a difference in the substances them- selves. I found also that if one takes a concentrated solution of photogenin filtered through a porcelain or silicious filter candle to remove all granules and cell fragments and adds to it a little saponin powder, or amyl alcohol or NaC1 or other inorganic salt crystals or tissue ex- tracts of certain invertebrate non-luminous animals, that light would appear. Because NaC1 could not possibly be oxidized by photogenin (= luciferase)--or any other substance--and because of my inability to make photophelein (= luciferin) luminesce with oxidizing agents, I regarded the photogenin itself as the source of the light and the oxidizable body. I have compared photogenin to zymase and photophelein to the coenzyme of zymase, believing that we are dealing with a system similar to that of the enzyme-coenzyme system of yeast. Hence the name photophelein or body assisting in the pro- duction of light. 3The followingoxidizingagents (added ,where possible, in minute crystal or powder form) all gave light with Pholasluciferin, but no light with Cypridlna luciferin: KMnO,, K~Cr2OT,PbO2, Na~O2,BaO2, MnO2, KWe(CN)e, K2S20,, Na2BdOs,and H~O~. The followingoxidizingagents gave no light with either Phalas luciferin or Cypridina luciferin: K2CrOd, CrO3, KC103, KC1Od,FeCI,, KNO3, C1 or Br water, I in KI, Na hypochlorite, hypobromite, or hypoiodite, colloidalAg or Pt, benzoylperoxide,potato or turnip juice, or blood containing hemoglobinor hemocyanin. If H202 in addition to the oxidizingagent is added to Cypridinaluciferin,no light appears except a faint momentary flash with Na hypochloriteand hypobromite. As this faint flash also appears with thoroughly boiled extracts of Cypridina, lacking luciferin, it can have no significance. If H202 in addition to the oxidizingagent is addedto Pholasluciferinthe light is in some casesbrighter than with H202 alone. E. NEWTON HARVEY 135 I now believe that under the term photophelein I have previously included two separate substances. One of these is the thermostabile dialyzing substance extracted from Cypridina by hot water. A1- tho_ugh this substance cannot be oxidized with light production by oxidizing agents, it does oxidize spontaneously (also without light production) in the air and loses its power of producing light with photogenin. In the absence of air its solutions are stabile for months. Once oxidized it can again be reduced and will again give light if photogenin is added. It is therefore an oxidizable material and, I believe, similar to the luciferin of Pholas. I propose therefore to use Dubois' word luciferin for the thermostabile dialyzing substance of Cypridina in place of photophelein and luciferase for the thermolabile non-dialyzing substar~ce in place of photogenin. The source of the photogenic substances can be designated by prefixing the name of the animal as Cypridina luciferin, Pholas luciferin, etc. I suggest also that luciferin when oxidized be designated oxyluciferin. Luciferin is found only in luminous animals. In non-luminous ani- mals and probably also in luminous animals there is a second substance which I have formerly includecl in the term photophelein, and which may be properly so called, that acts in a manner similar to saponin, NaC1 crystals, etc., upon the extract of Cypridina which has stood until the light disappears. When we allow a Cypridina extract con- taining luciferin and luciferase to stand, the luciferin is not completely oxidized, even though the extract is thoroughly aerated, but some of it is bound (adsorbed or combined?) by other substances in the ex- tract. The saponin, NaC1 crystals, and extracts of non-luminous animals act by setting free the bound luciferin which is then oxidized and light appears. I suggest that the term photophelein be applied to these substances in tissue extracts. They are not destroyed by boil- ing. On standing some are stable while others are unstable. The best way to rid a luciferase solution of the bound luciferin is to shake it thoroughly with chloroform. Such a solution will give no light with extracts of non-luminous animals or saponin, NaC1 crystals, etc., but a brilliant light with Cypridina luciferin. An insight into the modus operandi of saponin, NaC1 crystals, or photophelein may be gained from the following experiments. Both luciferin and luciferase are adsorbed by many finely divided precipi- 136 PHOTOGENIC REACTION IN" CYPRIDINA tates and colloidal particles, such as bone-black, Fe(OH)~, kaolin, and others. If we take a colloidal Fe(OH)3 solution of the proper concentration (which can only be determined by experiment), add some dilute luciferase to it, and then after a minute luciferin, no light will appear. This is because the luciferase has been completely ad- sorbed by the colloidal Fe(OH)3, for if we now add some dilute luci- ferase to the above mixture light will appear but it will very quickly disappear because the new luciferase added is again very rapidly ad- sorbed, but not so rapidly adsorbed that we fail to get light at first. On adding more luciferase we may again get a momentary light but the additions cannot be made indefinitely because we finally reach a point where the colloidal Fe(OH)~ has become saturated with luci- ferase and then the mixture glows for a considerable time. It is ob- vious that for this experiment to succeed there must be more luciferin present than can be completely adsorbed by Fe(OH)3 and so little luciferase present that it is completely adsorbed by the Fe(OH)~. Suppose we have a mixture of Fe(OH)3, luciferase, and luciferin com- plying with the above conditions. Can we in any way r~move the luciferase from its adsorbed condition on the colloidal Fe(OH)3? This might theoretically be done in two ways and we actually find in practice that both methods are possible. Anything which precipi- tates the colloidal Fe(OH)3 will decrease the surface available for adsorption of luciferase and if the surface area is sufficiently de- creased some luciferase may be forced into solution again where it is able to oxidize the luciferin. If we add NaC1 crystals to the col- loidal Fe(OH)3 + dilute luciferase + more concentrated luciferin mixture, the Fe(OH)3 is precipitated and light appears. If in place of NaC1 crystals we add a trace of saponin, the colloidal Fe(OH)3 is not precipitated but light also appears. This is an example of the second method of removing luciferase from an adsorbed condition; namely, by using a material (saponin) which is more strongly ad- sorbed than the luciferase and which is able to replace it as adsorbed body. I believe these to be the explanations of the effect of NaC1 crystals, saponin, etc., in giving light with luciferase solutions, ex- cept that the luciferase is in excess and a small amount of adsorbed (or combined) luciferin is present which is liberated by NaC1 or saponin and gives light with luciferase. Photophelein probably acts in a manner analogous to the saponin. r.. ~wro~ ~v~Y 137 I have considered the thermostabile, dialyzing substance as simi~ lar to the luciferin of Pholas despite the fact that Dubois reports Pholas luciferin destroyed at 70C., whereas Cypridina luciferin is only destroyed by several minutes boiling in an open beaker. I find that this destruction of Cypridina luciferin on short boiling is due to the increased rate of oxidation at the boiling point and that no destruction of Cypridina luciferin will occur if boiled in an atmosphere of hydro- gen.4 Cypridina luciferin is truly thermostabile but is oxidized to oxyluciferin on boiling in the air. We may say that Pholas luciferin is similar but certainly not identical with Cypridina luciferin. If so, we should expect to obtain light on mixing Pholas luciferin and Cypridina luciferase, yet no light appears. Neither is there light on mixing Cypridina luciferin and Pholas luciferase, although the Pholas luciferase prepared from the material which Dubois sent me gave a rather faint light with Pholas luciferin3 We have, therefore, at least three substances concerned in light pro- duction: luciferin, luciferase, and photophelein. Luciferin is a body oxidizing with light production, dialyzable, and relatively resistant to heat. Luciferase is destroyed by boiling, non-dialyzable, and acceler- ates the oxidation of luciferin. While it may be used up in the re- action if mixed with a sufficient quantity of luciferin, luciferase has 41 have endeavored to repeat this experiment with the luciferin of Pholas sent me by Professor Dubois, but without success. Pholasluciferin boiled in a current of hydrogen for 15 minutes would give no light when a crystal of KMnO, was added. The hydrogen was produced in a Kipp generator and may have con- tained a little air. In my experience short (20 to 40 seconds) boiling of Pholas luciferin does not completely destroy its power of producing light when a crystal of KMnO4 is added. 51 believe the faint light obtained on mixing Cypridina luciferin and fire-fly or Noctilucaluciferase and viceversa,recorded in my former paper (Am.J. Phys- iol., 1917, xlii, 328) where luciferin is called photophelein and luciferase is called photogenin, is not due to the oxidation of luciferin by luciferase of the second species but is due to the presence of photophelein. I am led to this conclusion because the light is so faint, but cannot be sure until the cases are reinvestigated. The mixing of luciferin and luciferase of different species or genera of luminous ostracods, especially if the colors of their luminescence differed, would shed con- siderable light on this interesting question of specificity. A non-luminous Japa- nese species of Cypriclinadoes not contain either luciferin or luciferase but it does contain photophelein. 138 PHOTOGENIC REACTION IN CYPRIDINA many of the characteristics of an enzyme and certainly as much right to be called an enzyme as the peroxidases of plants which are also used up in the oxidation process. The Cypridina luciferase re- action appears to be specific to an extraordinary degree. Of many tried I have found no substances or plant or animal extracts which can take the place of luciferase6nor any substances 7or plant or ani- mal extracts s which can be oxidized with light production by luci- ferase. The light recorded with various extracts of luminous and non-luminous animals in my former paper is to be referred to the presence of photophelein, the third substance concerned in light production which probably acts by assisting the luciferin-luciferase reaction in the manner already suggested. Let us now turn to the oxidation product or products of luciferin. When luciferin is oxidized it must be converted into some substance or substances and I believe this change involves no fundamental destruction of the luciferin molecule, as it is a reversible process. I I have tried the blood or extracts of many species of animals or plants in- cluding those containing strong oxidizing enzymes both with and without H202 and have always failed to obtain light with Cypridina luciferin. Among others the juice of Indian pipe (Monotropa),potatoes, and turnips (containing strong oxi- dases and peroxidases), the blood of the ox and a worm (Arenicola) (containing hemoglobin), the blood of the squid (Loligo), Limulus and Sycotopus (containing hemocyanin), and extracts of Che~topterus(a luminous annelid), and the mollusc, Unio, (rich in manganese) were tried. Dubois reports that he has obtained light on mixing Pholas luciferin with the blood of divers molluscs and marine crus- taceans (Ann. Soc. Linn. Lyons, 1913, xl). I can confirm this statement for an extract of Unio,but obtained no light with Limulus blood, Sycotopus blood, squid (Loligo) blood, or turnip or potato juice and Pholas luciferin. Evidently Pholas luciferin is much more readily oxidized with light production than Cypridina luciferin. The following oxidizable substances have been tested: msculin, lophin, berga- mot oil, pyrogallol, gallic acid, aniline, adrenalin, phenol, a-napthol, p-phenylene- diamine, ortol, orcinol, hydrochinone, resorcinol, pyrocatechol, tannin, benzi- dine, gum guaiac, amidol, a-napthylamine, and the chromogen of the false in- digo plant (Baptisia). Luciferase, with or without H~O2,will not accelerate the oxidative color change in any of the above compounds. s I have regularly obtained a fair light on mixing luciferase well shaken with chloroform to set free any bound luciferin and boiled potato or turnip juice or boiled Limulus blood. The light is especially marked about the coagulum in the boiled Limulus blood. The significance of these results is not apparent. E. NEWTON HARVEY 139 shall speak of the principal if not the only product formed as oxy- luciferin. Most observers have considered a rather fundamental change to occur when the photogenic substance is oxidized. Thus the crystals of xanthine or some related substance in the reflecting layer of the fire-flyhave been regarded as the oxidation products of the luminous material thought to be nucleoprotein. Dubois 9 regards luciferin as a protein and states that it forms the same oxidation products as other proteins, amino-acids being mentioned as possible substances formed. It should be pointed out in this connection that the formation of amino-acids from proteins involves no oxidation but an hydrolysis. If we assume that the oxidation of luciferin changes the molecule but slightly, we at once think of comparing the change luciferin --* +.._ oxyluciferin with the change reduced hemoglobin ~ oxyhemoglobin. The condition is, however, not so simple as this, for oxyhemoglobin will again give up its oxygen providing the partial pressure of oxygen is sufficiently low, whereas oxyluciferin will not do this. We cannot reduce oxyluciferin solution by exhausting the oxygen with an air pump. There is another oxidation reduction system which can also be easily reversed, but not by merely removing the oxygen--that is the reduction of a dye such as methylene blue to its leuco base. I be- lieve the change which occurs when luciferin is oxidized is similar to that which occurs when the leuco base of methylene blue or sodium indigosulfonate is oxidized to the blue dye. My attempts to reduce the oxidation product of luciferin started from the observation that if one places a clear solution of luciferase in a tall test-tube, although it may give off no light at first when shaken, after standing a day or so.a very bright light would appear on shaking. This was especially true when the luciferase had be- come turbid and ill smelling from the growth of bacteria. Thinking that the bacteria produced a substance which could be oxidized by the luciferase, I tried growing bacteria and also yeast on appropriate culture media and after some days of growth mixing the culture media containing the products of bacterial or yeast growth with luciferase, 9Dubois, Ann. Soc. Linn. Lyons, 1914, Ixi, 169. 140 PHOTOGENIC REACTION IN" CYPRIDIN'A expecting to obtain light. But no light appeared. However, if a little crude luciferase solution was added to the bacterial or yeast cultures and then allowed to stand for some hours, light appeared whenever they were shaken. Indeed such cultures behaved much as a suspension of luminous bacteria which has used up all the oxygen in the culture fluid and will only luminesce when, by shaking, more oxygen dissolves in the culture medium. Realizing that in bacterial cultures in test-tubes anaerobic conditions soon appear, and also the strong reducing action of bacteria upon many substances (for instance nitrates or methylene blue) under anaerobic conditions, it occurred to me that the bacteria might be utilizing the oxygen of the oxidation product of luciferin, reducing it to luciferin again. We must remem- ber that since crude luciferase solution is a cold water extract of a luminous animal allowed to stand until all the luciferin has been oxi- dized, it must contain oxyluciferin as well as luciferase and will give light if the oxyluciferin is again reduced and oxygen admitted. This appears to be the correct explanation of the above experiments. Not only bacteria but also tissue extracts have a strong reducing action in absence of oxygen. Thus, muscle tissue stained in methylene blue will very quickly decolorize (reduce) the methylene blue if oxy- gen (air) is kept away, but the blue color immediately returns if air is admitted. Oxyluciferin, i.e. a solution of luciferin which has been completely oxidized by boiling or standing in air until it no longer gives light with luciferase, if mixed with a suspension of ground frog muscle and kept in a well filled and stoppered test-tube for some hours, is reduced to luciferin and gives a bright light if now poured into luciferase solution. Frog muscle suspension alone or oxyluci- ferin alone gives no light with luciferase, nor will a mixture of frog muscle suspension and oxyluciferin, if shaken with air for several hours. Only if this last mixture is kept under anaerobic conditions is the oxyluciferin reduced. The reducing action of tissues is said to be due to a reducing enzyme (reducase or reductase), itself composed of a perhydridase and some easily oxidized body such as an aldehyde.1 In the presence of the 10Bach, A., Biochem. Z., 1911,xxxi,443; xxxiii,282; 1912,xxxviii, 154; 1913, lii, 412. E. NEWTON" HARVEY 141 perhydridase the oxygen of water oxidizes the aldehyde and the hy- drogen set free reduces any easily reducible substance which may be present. There is a perhydridase in fresh milk, spoken of as Schard- inger's enzyme,n which is destroyed by boiling. If some aldehyde is added, fresh milk will reduce methylene blue to its teuco base or nitrates to nitrites, upon standing a short time. If shaken with air, the blue color returns. There is no reduction unless an aldehyde is added or unless some boiled extract of a tissue such as liver is added. The boiled liver extract has no reducing action of its own but supplies a substance similar to the aldehyde which has been spoken of as a coenzyme. Milk will reduce methylene blue without aldehyde if bacteria are present in large numbers. There is also no reduction if the milk, methylene blue, and aldehyde are agitated with air. The temperature optimum is rather high, 60-70C. I find that milk is a favorable and convenient medium for the re- duction of oxyluciferin and that it acts without the addition of an aldehyde or the presence of bacteria. There is probably a substance acting as the aldehyde in the luciferase-oxyluciferin solution. No light appears if milk is added to a luciferase-oxyluciferin solution, but if the mixture is allowed to stand in absence of oxygen, light will appear when air is admitted. The air can be conveniently kept out by tilling small test-tubes completely with the solution and closing them with rubber stoppers. Oxyluciferin may also be readily reduced by the use of the blood of the horseshoe crab (Limulus) allowed to stand until bacteria develop.~ This experiment is of special interest because the blood contains hemo- cyanin which is colorless in the reduced condition and blue in the oxy- condition. The color change thus serves as an indicator of the oxygen concentration in the Mood. A sample of foul smelling Limulus blood full of bacteria will become colorless on standing in a test-tube for 10 or 15 minutes but the blue color quickly returns if shaken with air. Such a blood has the power of reducing oxyluciferin through the activity of the bacteria which it contains. Fresh blood has very little if any reducing action. 11Schardinger, F., Chem. Ztg., 1904,xxviii, 704. 12Alsberg, C. L., J. Biol. Chem., 1915,xxiii,495. 142 PHOTOGENIC REACTION IN CYPRIDINA As almost all animal tissues contain reductases, it is not surprising to find that a freshly prepared and filtered extract of Cypridina con- taining oxyluciferin .and luciferase which gives no light on shaking, will, on standing in a stoppered tube for 24 hours at room tempera- ture, give light when air is admittedY While this may be due to the development of bacteria with a reducing action it does not seem likely, as under the same conditions methylene blue is not reduced in 24 hours and there is no turbidity or smell of decomposition in the tube. In 48 hours bacteria do appear and methylene blue is also reduced. If we add chloroform, toluene, or thymol to the tubes of Cypridina extract to prevent the growth of bacteria, and allow them to stand 48 hours, upon admitting air the tube with chloroform gives no light but the tubes with toluene and thymol do give light although it is not so bright as if they were absent. I believe that these sub- stances have a destructive action on the reductases, most complete in the case of chloroform. I have not been able to demonstrate that a Cypridina extract will reduce methylene blue or nitrates to nitrites either with or without the addition of acetaldehyde. This may be due to the fact that oxy- luciferin, which is also present, may be reduced more readily than either nitrates or methylene blue and so is reduced first. Dubois14 has described in Pholas a precursor of luciferin, which he calls proluciferin, which is converted into luciferin by another en- zyme, coluciferase. The proluciferin is not destroyed by boiling and the coluciferase will withstand a higher temperature than luciferase and may be freed of luciferase in this manner. He cites an experi- ment15to prove the existence of proluciferin and coluciferase in Pholas but I have been unable to repeat this with Cypridina. One might suppose that on allowing an extract of Cypridina (luciferase) to stand in absence of oxygen, some proluciferin, assuming this to be present, would be converted into luciferin which would give light if air was admitted. But we can allow a boiled extract of Cypridina (contain- ing no coluciferase) to stand with milk or muscle tissue suspensions 13This experiment may also be performedwith Pholasluciferasewith a simi- lar result. 14Dubois, Compt.rend.Soc.biol.,1907,lxii, 850; 1917,lxxx,964. 15Dubois, Compt.rend.Soc.biol.,1917,lxxx,964. E. IN-EWTOI~ HARVEY 143 in absence of oxygen and, upon admitting air and adding luciferase, obtain light. As luciferase is only found in luminous animals it does not seem likely that a coluciferase would be widespread but we do know that a reducing enzyme occurs in milk and tissue extracts--is, in fact, widespread. It seems more logical to interpret the above experiments as due to the reduction of an oxyluciferin to luciferin rather than the conversion of a proluciferin to luciferin. Indeed, we can reduce oxyluciferin by means which do not involve the use of animal extracts and consequently are free from the ob- jection that "coluciferase" may be responsible for the result, but which, nevertheless, are perfectly well known reducing methods. Perhaps the best of these is reduction by palladium black and sodium hypophosphite. The latter is oxidized in presence of palladium and nascent hydrogen is set free.16 The nascent hydrogen reduces any easily reducible substance which may be present, such as methylene blue or oxyluciferin. Oxyluciferin is not reduced by palladium alone or hypophosphite alone but methylene blue is reduced by palla- dium black alone. If hydrogen sulfide is passed through a solution of methylene blue the dye is very quickly reduced and becomes colorless. If the H2S is driven off by boiling the colorless methylene blue solution, the blue color again returns on cooling. Oxyluciferin can also be reduced to a certain extent by H2S. Sulfur dioxide or oxides of nitrogen (prepared by the action of HNO3 on Cu) had no reducing action on either methylene blue or oxyluciferin. Dilute acid favors the reduction of oxyluciferin. If one saturates an oxyluciferin solution with COs or adds a little dilute acetic acid and allows the solution to stand for 24 hours, a certain amount of reduction will occur. No reduction occurs if the solution is saturated with pure hydrogen and allowed to stand 24 hours. If one adds some Mg powder to oxyluciferin and then dilute acetic acid in successive additions as the acetic acid is used up in formation of Mg acetate, the oxyluciferin will be reduced relatively quickly. Nascent hydro- gen is produced in the reaction and is no doubt the active reducing agent, while the acid accelerates the reduction. Soured milk also has ,eBach, Ber. chem. Ges., 1909,xlii, 4463. 144 PHOTOGENIC I~EACTION IN CYPRIDINA quite a marked reducing action. Acid thus favors reduction and hinders oxidation, while alkali favors oxidation and hinders reduction of the oxyluciferin. While I have not studied the properties of oxyluciferin so fully as those of luciferin, as far as I can judge both substances give the same general reactions and possess identical properties. If we make a con- centrated hot water extract of Cypridina,it will contain all the sub- stances of the animal soluble in hot water and not coagulated by heat and may be spoken of as crude luciferin solution. If air is bubbled through this solution for some time, all the luciferin is oxidized and it may then be spoken of as crude oxyluciferin solution. Both crude luciferin and crude oxyluciferin solution are yellow in color, but I do not believe that either luciferin or oxyluciferin is yellow in color be- cause an ether or benzene extract of Cypridinais also yellow although luciferase, luciferin, and oxyluciferin are all insoluble in ether and benzene. The yellow pigment which can be observed to make up part of the luminous gland of Cypridinais not luciferin or luciferase. It may be a pigment related to urochrome. When tests are applied and precipitating reagents are added to crude luciferin and crude oxyluciferin solution they give identical results in each case. Thus if crude luciferin is saturated with (NH,),S04 or MgSO,, a flocculent precipitate forms which may be demonstrated to contain most of the luciferin. Oxyluciferin solution also gives flocculent precipitates on saturation with (NH4)~SO4 and MgSO4 and these contain most of the oxyluciferin. To demonstrate this the precipitates, after washing, are dissolved in a small amount of water mixed with fresh milk (or frog muscle suspension) and allowed to stand in a stoppered tube for 24 hours. If any oxyluciferin is pres- ent it will be reduced to luciferin and give light when luciferase is added. One-half saturation with (NH,),SO, or MgSO4 or saturation with NaC1 salts out no material from either crude luciferin or oxyluci- ferin solution. Picric acid gives no precipitate but only an opales- cence in both cases. In a similar manner it may be shown that most of the oxyluciferin is precipitated by phosphotungstic acid but not by acetic acid or CO~, in this respect also agreeing with the behavior of luciferin. Like luciferin the oxyluciferin will pass porcelain fil- ters, dialyze through parchment or collodion membranes, is soluble E. NEWTON HARVEY 145 in absolute alcohol but not in ether or benzene, and is undigested by salivary diastase, pepsin HC1, Merck's pancreatin in neutral solution, and erepsin. The salivary diastase and the pancreatin (containing amylopsin, trypsin, and lipase) were allowed to digest for 4 days at 38C. without showing any evidence of digestive action. Oxyluci- ferin is partially but not completely precipitated by basic lead acetate and tannic acid. As luciferin is so easily oxidizable a substance, we should expect to find that it will reduce just as glucose will reduce. However, a con- centrated solution of luciferin has no reducing action on Fehling's (alkaline Cu), Barfoed's (acid Cu), Nylander's (alkaline Bi), or Knapp's (alkaline Hg) reagent. Glucose will reduce methylene blue in alkaline (not in neutral) solution but luciferin will not reduce methylene blue in alkaline or neutral solution. It would seem, then, that luciferin must contain no aldehyde group. If so, we should ex- pect to obtain reduction of some of the above reagents. Just what group is concerned in the oxidation is unknown at the present time and speculation regarding it in the absence of more experimental data can be of little value.