Utilization and Digestion of Carbohydrates by the Adult Blowfly

The first part of the investigation consisted of a series of feeding experiments with different sugars and allied substances and, in extent, was necessarily limited by the extreme rarity and scarcity of many of the substances concerned, the quantities available being in some cases very small (1-2 g.). A preliminary account of the feeding experiments has already been given (Fraenkel, 1936).

In all the experiments the flies were fed from the time of emergence on a diet which consisted solely of water and the substance to be tested. The flies were kept in 7 lb. jam jars, which had a filter paper overlying a layer of sand at the bottom to keep conditions as dry and clean as possible. Twenty flies were normally used in each experiment, but when the amount of available diet was limited the number was reduced to fifteen or ten. The temperature was kept constant at 27° C. throughout. Most substances were fed dry and in quantities which lasted for only a few days so as to minimize contamination, water being supplied by means of soaked cottonwool. In all cases where the length of life was not appreciably increased by the special diet when given in dry form the experiment was repeated with the substance dissolved in water, the soaked cotton-wool being then omitted. This was to exclude the possibility that the special diet might be tasteless, or even possess an unpleasant taste, and thus be refused by the flies, which would be forced to take it, however, when thirsty, in the repeat experiment. Such solutions had to be changed daily to obviate contamination, so that much larger quantities of material were required than was the case when the diet was fed dry. For this reason some of the rarer substances could not be fed in liquid form. Fortunately, in almost every case wherein a substance proved to be of great nutritional value it was eaten in the dry state as readily as cane sugar itself, and the only substances, namely, lactose, starch and glycogen, on which the flies subsisted for an appreciably longer period when fed in aqueous solution, were available in quantity.

A fly deprived of water and all nutrient lives for about $212$ days, while that given water alone may live up to days. Every increase in length of life above about 4 days therefore can be ascribed to the positive effect of the added nutrient. In most experiments it was found that one or two flies will outlive the rest for a considerable period, as will be seen from the following experiment, which is typical.

Twenty flies on a diet of α-methyl-glucoside

It seems, therefore, that the number of days which the longest living fly in any one experiment lives does not give a good indication for the nutritional value of a particular diet concerned, and a much more significant average figure for survival was considered to be the number of days before half the total number of flies were dead. This figure is also much less liable to be influenced by the number of flies used in the experiment.

The results of all the feeding experiments are given in Table I. They show that the carbohydrates used can be divided roughly into three groups: (1) substances on which flies survive well, half the insects being dead in about 18-35 days, (2) substances on which survival is no better than on water alone, i.e. about 3-4 days, and (3) substances on which the flies survive for a few days longer than those insects receiving water alone.

Group 1. The first group consists of the hexoses glucose, fructose, galactose and mannose ; the disaccharides sucrose, maltose, trehalose, melibiose ; the trisaccharides raffinose, melezitose; the polysaccharide dextrin; the glycosides α-methyl-glucoside, a-methyl-galactoside ; and the sugar alcohols mannitol and sorbitol, for the flies live on these substances quite as well as on sucrose. The difference in the number of days which the flies lived on these substances may have some significance, but it is not regarded as a valid one in these experiments, for one always finds variations in the viability of flies which are reared at different times and in different batches. The viability may be influenced, for instance, by the amount of food eaten by the larva, the temperature at which the larvae were reared and also the particular conditions of the experiment, such as humidity and infection with micro-organisms. To obtain differences which could be regarded as significant would require more standardized conditions of breeding, the use of pure strains, the employment of a much greater number of flies and several replicates of each experiment. In most cases the amount of material available as diet was too limited for such a purpose. The longevity of the flies on sorbitol, however, proved to be so much superior to that on all the other substances tested that we can safely regard this difference as being significant.

Group 2. The sugars which had no nutritional value at all were the pentoses arabinose and rhamnose; the hexose sorbose; the disaccharide cellobiose; the polysaccharide inulin; the glycosides β-methyl-glucoside, β-methyl-galactoside, a-methyl-mannoside, β-methyl-fructopyranoside ; the phenol-glucosides helicin, salicin, arbutin; and the alcohols dulcitol, erythritol and inositol. In all these cases the flies did not seem to touch the dry substance but appeared to drink freely of an aqueous solution of it; hence it seems certain that the substance can have no nutritional value for these insects.

Group 3. Of the third group of substances, i.e. those on which the length of life was somewhat increased, xylose was eagerly eaten, but not utilized to a great extent. Lactose was not touched in the dry state, but was taken readily in solution and was utilized to only a limited extent, for the walls of the cages soon became splashed with white stains of apparently unaltered lactose, which had passed the gut and had been excreted as faeces. These particular feeding experiments were repeated twice, using 100 flies in all so that the results are more reliable than most of the others. It is suggested that the limited degree of utilization of lactose by the fly is due to bacterial action in the gut, and not to digestion by enzymes produced in the gut.

The two polysaccharides starch and glycogen were obviously not touched as dry substances, but did maintain life to a limited extent when fed in aqueous solution. The gut of the adult fly contains amylase (Wigglesworth, 1929), but it seems likely from the feeding experiments that this enzyme is present in too weak a concentration to produce glucose in sufficient quantity for maintaining the life of the fly. It was obvious in the experiments with starch and glycogen that the majority of flies died at an early stage, and only a small percentage were able to live for a considerable period. In one particular experiment, using starch dissolved in water, the survival of the flies was as follows :

Tests were also made to find out to what extent sugars can be replaced by other substances. It was assumed that some of these might be converted in the fly organism to sugars and used as such, which would be manifested by their effect on the survival. In vertebrates, for instance, proteins are known to be convertible to sugars, while there are other substances which are known to be intermediary products of carbohydrate metabolism. The choice of substances for the experiments with flies was thus largely determined by their convertibility into glycogen after injection into the blood stream of vertebrates.

With regard to the possible utilization of proteins, it has been shown previously (Glaser, 1923; Evans, 1935) that meat, which is the natural food for the adult blowfly, keeps the insects alive for no longer than water alone, and I myself have found that on a diet, of meat and water flies will survive for only 3 days, while on peptone, which probably contains all the essential elements of meat protein, they may live for $412$ days. Another natural source of protein is milk ; on milk alone half the flies were dead after $412$ days, and the last fly died after 17 days. The survival here was no more impressive than on lactose dissolved in water. It can therefore be concluded that proteins are not converted into sugars in the fly. This is not due to the deficiency of the gut of the adult fly in proteinases, as has been claimed by Wigglesworth (1929), for the fact that feeding of proteins is indispensable for the formation of eggs shows that the gut can digest these substances. Furthermore, the presence of the proteinases “trypsin” and “erepsin” can easily be demonstrated with the alkaline casein method (Cole, 1933, Exp. 273) and the biuret test. It would appear, therefore, that the adult fly either cannot produce intermediary carbohydrates from protein, or does so at such a slow rate that it cannot maintain the ordinary muscle metabolism.

Of the substances which are known partly as intermediary products in the carbohydrate metabolism, partly as good glycogen formers in the liver of vertebrates, the following have been tested: glycol, dihydroxyacetone, alanine, glycine, butyric acid, isobutyl alcohol, lactic acid, pyruvic acid, succinic acid, malic acid, citric acid. For the acids the sodium salt as well as the free acids were tested. The results of all these feeding experiments were negative. It is possible that this was due to the fact that some of the substances were not eaten by the flies at all, or only in very small quantities ; that they acted as poisons ; that they were not absorbed in the gut, or that they acted adversely by their high osmotic pressure. No particular precautions had been taken, however, to obviate certain of these effects, so the negative results obtained ought not to be taken as final. But none of these adverse factors seems to be applicable to the experiments with glycine or alanine and the non-utilization of alanine is particularly interesting in view of the fact that this amino-acid, which is closely related to pyruvic acid, is utilized as a glycogen former in mammals (Neubauer, 1928). This observation seems to indicate that the course of carbohydrate metabolism is not identical in insects and vertebrates.

The only non-carbohydrate which was utilized, although only to a slight extent, was glycerol. The following are the protocols of two different feeding experiments, in which an approximately 2 % solution of glycerol was supplied :

There can be no doubt, therefore, that glycerol does prolong the life of the fly and it seems most likely that this is due to the conversion of glycerol into a carbohydrate or one of the intermediary products which would normally arise in carbohydrate metabolism. It is thus possible that after the difficulties of presentation have been overcome some of the other non-carbohydrate substances, which have so far given negative results, may prove to be convertible into and utilizable as carbohydrates by the fly.

From what we know of the absorption of sugars in the gut of other animals, it is very likely that pentoses, hexoses and sugar alcohols are immediately absorbed from the blowfly gut without any previous change in their structure, but that di-, tri- and polysaccharides and glycosides are only absorbed after they have been broken down to monosaccharides. It is to be expected, therefore, that all the di-, tri- and polysaccharides and glycosides which are utilized by the fly will also undergo hydrolysis in the gut, which must, therefore, presumably, produce all the enzymes which are necessary for this purpose, although the possibility that the breakdown is due to bacterial action must not be overlooked.

To test for the presence or absence of carbohydrases in the gut, digestion experiments were carried out in vitro, with di-, tri- and polysaccharides and glycosides as substrates. The enzyme preparations were made from the guts of newly emerged and unfed flies, because in this state the presence of large numbers of bacteria in the gut seemed unlikely. A number of guts were dissected out from the flies, ground up in water with sand and the resulting suspension of gut tissues distributed into small specimen glasses in., so that each tube contained approximately the substance of one-half of a gut. About 0.3 ml. of a 2% solution of the substrate was added, together with 5 drops of a buffer solution and 1 drop of toluene to prevent bacterial growth. These tubes were incubated for 24-48 hr. at 36° C. and subsequently tested for the presence of glucose, fructose or galactose by the osazone method. Since only 6 mg. of the substrate were needed for each individual test, it was possible, even when the sugars were available only in very small amount, to test over a wide range of pH values, the necessary controls being made with gut suspensions in which the enzymes had been destroyed by boiling. An incubation period of 24 hr. proved to be sufficient in all the tests that gave a positive result. The only instance in which the osazone test failed was in the case of β-methyl-fructopyrano-side; here glucosazone was formed abundantly not only in the controls with the boiled gut suspensions, but also with the pure substance alone.

The results of these digestion experiments are shown in Table II. The system of registering the intensity of an enzyme action as negative, uncertain, weak, medium or strong by use of the symbols—, ±, +, + + or + + + is self-explanatory, but the gradation, of course, is purely arbitrary and is based only on a subjective judgement of a result and not on any quantitative determination, which was impossible with the small amount of material available. Comparison of Table II with Table I shows that all the di- and trisaccharides and glycosides which are of great nutritional value for the fly are converted by an enzyme present in the gut into monosaccharides. On the other hand, those substances which have no nutritional value are all unattacked. The only exception to this appeared to be lactose, which proved to be of some slight nutritional value, but for which the presence of an enzyme (lactase) could not be demonstrated. The reason for this may be either that the enzyme is present only in very weak concentration, or that, as has already been suggested, the attack in the gut is due to micro-organisms and not to an enzyme produced by the gut itself.

In most cases the enzymatic action of the gut suspensions was very vigorous, as seen by the very abundant formation of glucosazone crystals. Under the conditions of these tests there was no visible difference in the enzyme actions over a wide range of pH from 3.6 to 7.0, except in the case of a-methyl-galactoside, which was only hydrolysed at pH 5.4. An optimum enzyme action at pH 5.4 was also shown in some of the experiments with melibiose and raffinose and since these three substances are a-galactosides, it seems that the range of pH for enzyme action is much narrower for a-galactosidases than for a-glucosidases.

Simultaneously with the writer’s preliminary note on the utilization of sugars by the adult blowfly appeared a paper by Haslinger (1936) dealing with almost the same problem. Haslinger’s method of testing differed from mine in so far as each fly was fed by him individually and the various diets contained equivalent amounts of sugars. In order to induce the flies to feed, a constant amount of sucrose was added to each test-solution, which was sufficient to render it sweet for the flies, but insufficient to prolong their life. I have fed all substances in surplus quantity, and induced feeding, where it was necessary, by dissolving them in water. Table III shows that in spite of the difference in method and the small number of flies used in each experiment there is a remarkable agreement between the results of the two independent sets of experiments. The only other work with insects, in which the nutritional value of sugars has been determined by the survival method, was undertaken by Phillips (1927) and Vogel (1931) with adult bees. The results of Vogel’s and Haslinger’s experiments (Table III), taken from the latter’s paper, are expressed in terms of the nutritional value of sucrose, which is taken as unity. The adult bee resembles the adult fly inasmuch as in both insects life can be adequately maintained on sucrose alone, but, with regard to the majority of the other substances, although there is qualitatively a good agreement in their nutritional value for both insects, there are also considerable differences. Mannose not only fails to serve as a diet for the bee, but has, strangely enough, a poisonous effect. Cellobiose and arabinose, which were of no nutritional value for flies, have a considerable value for bees. Xylose has a much greater nutritional value for bees than for flies. On the other hand, galactose, lactose, melibiose and raffinose, all sugars which contain galactose, are of much greater nutritional value for flies than for bees. Melibiose, in fact, has no value at all for bees. It is interesting to note that in two of the feeding trials sorbitol proved to be of considerably higher nutritional value than all the other sugars tested, including sucrose. This phenomenon is so far unexplained, but it may possibly be due to the easier conversion of sorbitol into one of the intermediary products of carbohydrate metabolism after absorption into the blood.

The snail Helix pomatia seems to be the only animal in which the presence or absence of carbohydrases in the gut has been tested hitherto for a wide range of substances (Bierry, 1911, 1912; Oppenheimer, 1925, 1935). Here the gut juice has been shown to hydrolyse not only every di- and trisaccharide and glycoside tested, but also polysaccharides of such a resistant nature as cellulose, inulin and chitin.

Almost nothing is known at present about the nutritional value of a wide range of sugars and allied substances for any representative of the vertebrates, nor does a systematic enzyme survey similar to that with Helix by Bierry and with the fly by the writer appear to have been undertaken. Such information as is available so far is incomplete and often contradictory, but the data given in the last two columns of Table III, compiled mainly from Isaac & Siegel (1928), Fischer & Niebel (1896) and Oppenheimer (1925, 1935), represent what seems to be the most probable position at the present time. From this it seems that the range of substances which are digested in the vertebrate gut is much smaller than in the invertebrates so far tested, but it should be remembered that most of the present information on the utilization of pentoses, hexoses and sugar alcohols by vertebrates is limited to the part which these substances play in glycogen formation after injection into the blood ; it is not known whether the substances are utilized to a similar extent when fed.

A review of feeding experiments with sugar alcohols brings out some strange contrasts. Sorbitol, which on oxidation yields glucose and fructose, is fully utilized by bees and flies, as also is mannitol, which is oxidized to mannose. From this one would expect that dulcitol, which is oxidized to galactose, could be utilized as well, but this has not been borne out by experiment. In the case of vertebrates the position is still more confusing. Sorbitol is fully utilized when injected into the blood, but, strangely enough, not at all when fed (Waters, 1938); mannitol, on the other hand, is not even utilized when injected, nor is dulcitol (Fromherz, 1928).

The present study with blowflies seems to be the only one to date in which feeding experiments have been correlated with ad hoc investigations (in vitro) of the enzymes concerned ; flies are, indeed, extremely favourable material for such a research, since they are small enough for the former purpose yet large enough for the latter. Throughout the investigation the opinion has been expressed that the observed utilization and digestion of those substances which do not pass readily through the gut wall is due to the action of enzymes produced by the gut wall and not by micro-organisms living in the gut lumen. To establish this beyond doubt it would have been necessary to carry through the feeding experiments with sterile flies, and although it is possible to produce aseptic flies by rearing the larvae under aseptic conditions, it seemed impossible to keep them so for periods of weeks while changing the food daily. The main reason for assuming that micro-organisms do not play an essential part in the effects observed is the close agreement between the results of the feeding experiments and the enzyme tests in vitro, in which sterile conditions were obtained. The action of living bacteria is therefore excluded in the tests.

There remains only the possibility that the gut harbours a large amount of bacteria and that the enzyme action observed in the tests was due to enzymes extracted from them and not from the gut wall. To test this, a count of the bacteria present in the guts of newly emerged flies was undertaken with the aid of Dr S. E. Jacobs, to whom I wish to express my sincere thanks. From ten guts a suspension (io ml.) was prepared by the usual method, during which, admittedly, conditions could not be kept strictly sterile, and tests were subsequently made by incubation on nutritive agar. No bacterial growth took place in those which contained 1 ml. of this suspension. If, however, the latter had been previously diluted tenfold with water i ml. produced thirteen bacterium colonies. A hundredfold dilution under similar conditions gave only two colonies, while more dilute suspensions gave none at all. The gut suspension contained, therefore, 130-200 bacteria per gut. This small number is almost certainly derived from infestation during the course of the preparation and not from the interior of the gut. The fact that no bacteria developed in the tests with the undiluted suspension suggests that an inhibitor for bacterial growth was present in the gut. It can therefore be said with certainty that the gut suspensions used for the enzyme tests contained so few bacteria that the strong enzymatic action observed in vitro must be attributed to enzymes extracted from the gut wall.

Table II shows that of all the di- and trisaccharides and the glycosides tested, at least eight are hydrolysed by enzymes secreted by the gut of the fly. In keeping with older views as to the specificity of enzymes one would have to assume the presence in the gut of eight different and specific enzymes, namely, sucrase, maltase, trehalase, melibiase, raffinase, melezitase, α-glucosidase and αa-galactosidase. This assumption would appear somewhat far-fetched in view of the fact that only two of the sugars concerned, namely, sucrose and maltose (the latter as a breakdown product of starch and glycogen) would occur commonly in the natural food of the flies; and, indeed, certain of the compounds hydrolysed are not met with at all among natural products. On the other hand, two of them that are utilized only slightly, if at all, i. e. lactose and cellobiose, are available to flies, the one in milk and the other in decomposing wood. In considering these observations, the recent views of Weidenhagen become of interest. According to Weidenhagen (1931), di- and trisaccharides or glycosides are hydrolysed, not by an enzyme specific for each compound, but by glycosidases whose action is determined by the identity of the sugar whose potentially aldehydic carbon atom enters the glycoside link, and by whether combination occurs through the α- or β-position. This theory postulates the existence of only five enzymes, capable of hydrolysing all known di- and trisaccharides and glycosides based upon glucose, galactose and fructose, namely: α-glucosidase, which splits all α-glucosides (sucrose, maltose, turanose, trehalose, melezitose); β-glucosidase, which splits all β-glucosides (gentiobiose, cellobiose, phenylglucosides); α-galactosidase, which splits all α-galactosides (melibiose,’ raffinose); β-galactosidase, which splits all β-galactosides (lactose); and α-h- fructosidase, which splits all β-fructosides (sucrose, raffinose, gentianose). Applying this scheme to our results for the feeding experiments with flies, we find the closest possible agreement. All the α-glucosides (sucrose, maltose, trehalose, melezitose and a-methyl-glucoside) and αa-galactosides (melibiose, raffinose, a-methyl-galac-toside) tested are digested, but none of the β-glucosides (cellobiose, β-methylglucoside, phenolglucosides) or β-galactosides (lactose, β-methyl-galactoside). Whether β-h-fructosidase is present in the fly cannot be decided from these experiments, since of the fructosides tried sucrose is also split by α-glucosidase and raffinose by α-galactosidase. The fact that β-methyl-fructopyranoside has no nutritional value for the flies is irrelevant, since fructose in sucrose and raffinose is of the furanose type. Whether fructosidase is involved in the breaking up of sucrose and raffinose can only be decided by testing with α- and β-methyl-fructofuranoside, which I was unable to obtain. All the findings as to the utilization and digestion of di-, and trisaccharides and glycosides may, therefore, be fully explained by assuming the presence of only two enzymes, α-glucosidase and α-galactosidase, in the gut of the fly, and they provide an illuminating commentary on the validity of Weiden-hagen’s views.

1. A series of different sugars and sugar alcohols has been fed to adult flies and their nutritional value determined by their effect on longevity.

2. All the di- and trisaccharides and glycosides which are of great nutritional value for the flies are split in the gut of the flies by enzymes the presence of which could be demonstrated in vitro. No enzymes could be found by the same method which would split any of the substances which had been shown to have no nutritive value.

3. Weidenhagen’s theory of the specificity of carbohydrases offers a convincing explanation of the results of the feeding and digestion experiments. The presence of only two enzymes, α-glucosidase and α-galactosidase in the gut of the fly would account for the different action of various di- and trisaccharides and glycosides.

I am much indebted to Prof. D. M. S. Watson of University College, London, in whose Department part of this work was carried out. I have to thank also the Academic Assistance Council for a grant which enabled me to start this investigation, and Prof. J. W. Munro who obtained further financial support for it through the Agricultural Research Council. Finally, I should like to thank those whose names are mentioned in Table I for their kindness in providing me with samples of rare substances. Prof. A. C. Chibnall and Dr H. W. Buston were also good enough to read and criticize the manuscript.