View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

THE USE OF THE CHORIOALLANTOIC MEMBRANE OF THE DEVELOPING CHICK FOR CULTURE OF DERMATOPHYTES—A MODIFIED TECHNIC A PRELIMINARY REPORT UPON ITS USE FOR SERIAL PASSAGE* BETTY M. PARTRIDGE, M.A.

The chorioallantoic membrane of the develop-downy and granular strains in culture is T. men- ing chick embryo provides a suitable living sub-tagrophytes. A conversion of these downy forms strate for the cultivation of pathogenic fungi.into granular forms, during the course of serial This was first demonstrated by Moore (1) inpassage in guinea-pigs, was observed by Georg a histo-pathological study using a variety of(4). As the growth and production of lesions on fungi, including four dermatophytes: Micro-the chorioallantoic membrane by dermato- sporum canis, Epidermophyton floccosum (E.phytes was known, from earlier workers, to be inguinale), Trichophyton mentagrophytes(T. rapid compared with other laboratory animals, gypseum) and T. schoenleini (Achorion Schoen-it was decided to investigate the use of this leini). More recently, Showalter (2) studiedtechnic as a method for passage of cultures. By the morphological changes of the first threethe serial passage of T. rubrum from one chick dermatophytes, listed above, when in contactembryo membrane to another, it was hoped to with the living chick membrane. determine whether there was correlation between The object of the investigation, describedstrains and any alteration in their morphology in the present paper, was primarily to developand pathogenicity after passage. an improved in vivo technic for cultivating dermatophytes in the chorioallantois, and to METHODS AND MATERIALS compare the results so obtained with those of The cultivation technic employed by Moore previous workers. The dermatophyte chosenwas based on that described by Goodpasture (5) for investigation was Trichophyton rubrum, anand his collaborators in their work on viruses. organism which was not grown on the chori-Briefly this involved preparing 10 to 14 day old oallantoic membrane by either Moore or Sho-fertile hens eggs by cutting a 'window' in the waiter. shell, 1cm. square, with a carborundum disc. Trichophyton rubrum is of particular interestThe underlying shell membrane was also cut because of the frequency with which it causesand the shell and its membrane carefully re- long-standing ringworm in man. A study of themoved to expose the chorioallantoic membrane, clinical manifestations of infections caused bywhich was then inoculated with a saline suspen- this fungus, and the morphology of the strainssion of the fungous culture. The 'window' was isolated, has shown little direct relationship. Aclosed by surrounding it with a layer of vase- similar conclusion was reached by Silva, Kestenline-paraffin and applying a cover slip. The and Benham (3) in studies of their T. rubrumeggs were incubated, 'window' uppermost, at isolates. Cultures of T. rubrum vary consider-34°C. ably and range from white, downy forms to Burnet (6) described a modification of Good- granular and occasionally dysgonic forms; thepasture's technic for chorioallantoic inoculation, rate of pigment production also varies with thein which an artificial air-space was created above strain. Both granular and downy forms of thisthe chorioallantois to give a larger area for in- organism may occasionally produce hair infec-oculation. This tchnic has been widely utilized tion, usually merely a follicular invasion; simi-in the study of viruses and rickettsiae (Bever- larly for glabrous skin and nail infections. idge and Burnet, (7)), and was used by Sho- Another dermatophyte which produces bothwalter in his dermatophyte studies. He em- ployed Moore's technic of cutting a 'window' in *Fromthe Department of Medical Mycology,the shell and shell membrane to form a large Institute of Dermatology, London, W.C.2., Eng- land. opening into the artificial air-space. The inocu- Received for publication September 8, 1958. lum was applied by means of an inoculating 605 606 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY needle to the exposed surface of the choriollan-lying chorioallantoic membrane. Using a standard toic membrane, and the 'window' sealed withrounded steel burr, a circular 'window', 3 mm. a cover slip and vaseline-paraffin as before. diameter, was carefully drilled through the shell to expose the shell membrane. By this procedure, I. General Technic accidental damage to the underlying membranes was minimised, compared with previous tech- The technic employed in the present studyniques of cutting and picking away a larger area also embodies the use of the artificial air-spaceof the shell. A smaller aperture into the air-space above the chorioallantois, with modificationswas made by drilling through the shell and shell by Alexander (8). The principle feature of thismembrane with a finer burr, or puncturing them method is the small amount of manipulationwith a needle. (Diagram I, A.) of the chorioallantoic membrane required dur- Separation of membranes—A drop of sterile ing inoculation. This decreases the developmentphysiological saline was placed over the circular of non-specific lesions due to trauma whicharea of exposed shell membrane which was then carefully slit, using a sterile needle and pressing may cause difficulty in the interpretation ofdown and backwards at right angles to the fibres. any fungous lesions produced. Little trouble isAir was withdrawn from the air-space by gentle caused by air-borne contaminants, probablysuction with a rubber teat placed over the aper- because of the natural resistance of the eggture, and the chorioallantoic membrane separated tissues. However, as this work had to be carriedand dropped away from the shell membrane to out in a routine bacteriology and virus labo-produce an artificial air-space with the chorio- ratory, any risk of contamination and dissemi-allantoic membrane as its 'floor'. The saline was nation of fungus spores, due to cover-slip dis-utilised as a fluid wedge during this stage and its placement, for example, had to be minimized.disappearance acted as an indicator that the mem- The following method has been used and provedbranes had separated. Prior to inoculation, the egg was again candled to determine the extent and effective. position of the artificial air-space, and to ensure that the original air-space had been obliterated. Inoculation The area of chorioallantoic membrane available Pre-incubation—Fertilised eggs were incubatedfor inoculation was approximately 2 to 2.5 cm. in in a standard electric, thermostat-controlled eggdiameter, with the overlying shell membrane incubator, with a temperature of 37°C, forced air-about 5 mm. above its centre. (Diagram I, B.) circulation and relative humidity of 60 per cent. It was important to drop the chorioallantoic The eggs were automatically turned twice daily.membrane and carry out inoculation as soon as Ten days incubation was allowed—the minimumpossible after drilling, otherwise it tended to ad- period fnr develnpment nf the chorioallantois. here to the shell membrane, making it difficult to Candling end drilling—The fertile eggs wereavoid puncture of the chorioallantois and causing selected by candling and the limits of the air-minor trauma in the area corresponding to the space marked nn the shell. The pnsitinn of theopening in the shell. embryo was noted by its spontaneous movements, Inoculum—The unexposed chorioallantoic mem- and a point marked above it on the shell, carebrane was inoculated by introducing a saline sus- being taken to avoid blood vessels in the under-pension of the fungus into the artificial air-space.

.1

-AIR SPACE

HEM ERA NE S B

DIAGRAM 1. Arrangement of membranes and cavities in a 10-day old chick embryo, (A) before, and (B) after preparation for inoculation, as seen in longitudinal section. Points of drilling (arrowed). CHORTOALLANTOIC MEMBRANE FOR CULTURE OF DERMATOPHYTES 607

For this study, fungous suspensions were made byThese were stained by the periodic-acid-schiff vigorous shaking of a loopful of mycelium andmethod in order to trace the invasiveness of the spores from a fresh Sabouraud's agar slope culturefungus, and with haematoxylin and eosin to de- in 2 to 3 ml. of sterile physiological saline. Largertect any cellular reaction. The other half was fragments of mycelium were allowed to settle out,ground up with a few drops of sterile physiological and the cloudy supernatant fluid containing sporessaline in a Griffith's tube, and the emulsion used and hyphal fragments was used for inoculation.for culture. For large-scale work a uniform inoculum is recom- mended. Using a 1 ml. sterile syringe, the fungous II. Technic for Serial Passage suspension was introduced by inscrting the tip of Materials Used—For this prelimiuary survey, the needle through the slit in the shell membrane,3 forms of recently isolated Trichophyton rubrurn and by holding the syringe vertically, slowly ejecting the inoculum on to the chorioallantoicstrains were chosen. Single-spore cultures were membrane without touching either it or the shellobtained by growing spore suspensions on corn- membrane. Extensive trauma of the chorioal-meal agar and isolating single germinating lantois must have been caused by Showalter'sspores (Georg, (9). A typical single-spore cul- technic of inoculating the exposed membrane di-ture of each strain was selected. The type of rectly with an inoculating needle, hut such damagecolony it produced on a Sabouraud's agar plate, was practically eliminated by the above pro-and its origin were as follows: cedure. In this pilot study, 0.1 and 0.2 ml. fungous Strain (a), No. 4244—granular, well-pigment- suspension was used as an inoculum. ing colony, isolated from a scalp kerion with hair Incubation—After inoculation, the two aper- tures in the shell were sealed with triangles ofinvolvement. "Sellotape". The eggs were placed on slatted wire Strain(5), No. 4497—semi-downy, well- trays, with the circular aperture uppermost, andpigmenting colony, isolatcd from a groin in- incubated at 34°C. fection. Throughout this technic, care was taken to Strain (c), No. 4974—white, downy colony, avoid shaking the live embryo or allowing the eggwith slow pigment production, also isolated to cool down. Once the circular aperture was cut,from a groin and axilla infection. the egg was kept with this uppermost to avoidAll 3 strains produced abundant, typical micro- disturbing the position of the artificial air-space. conidia, and red pigment on cornmeal dextrose agar. A few macroconidia were present in Harvesting strain (a). The eggs were usually harvested after 4 and 7 Inoculation and Harvesting—Using the tech- days incubation, when the chick embryos were 14nic described, two or more eggs (depending on and 17 days old respectively. A longer incubationavailability) were inoculated with each strain, period was found to be unwise, as the period of in- cubation before the chick hatches is approxi-and the membranes harvested after 4 and 7 days mately 21 days. incubation at 34°C. On harvesting, the egg shell was cut in half Culture and Passage—Two sets of cultures horizontally, using fine scissors, and the upperwere made from the emulsion prepared from half, containing the chorioallantoic membrane, re-each infected membrane. Each set consisted of moved and upturned in a petri dish for examina-one plate and two slants of Sabouraud's agar, tion of the artificial air-space. With a fine scalpel,and two malt tellurite agar slants (for any bac- a cut was made, leaving acentimetremarginterial inhibition). These were then incubated, around the periphery of its 'floor', and the chorio-one set at 25°C. (our standard temperature for allantoic membrane gently stripped away and placed in a petri dish in sterile physiologicaldermatophyte growth) and one at 30°C. to saline. Stained smears were made of any fluidboost the growth rate. The cultures were ex- present within the artificial air-space to detect theamined weekly and subcultures made when presence of germinating or non-germinatingnecessary. As the dermatophyte used in this fungus spores, mycelial fragments and any con-survey was a relatively slow grower, no cultures taminating organisms. The chorioallantoic mem-were discarded under a month. Of the cultures brane was examined against a dark backgroundrecovered from the infected membranes, a and the presence of opacities and foci of infectionsingle clean isolate from a 7 day infection of noted. It was then divided into two. One half waseach strain was selected and used for inoculating placed in formol-saline and used for sections.a new batch of 10 day old fertile eggs. By this 608 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY method, 6 serial passages, referred to as pas-haps the inoculum was most concentrated, al- sage I, II, etc., were made throughout the coursethough there was always the possibility that of approximately one year. part of the tissue reaction in this area was non- Of the isolates recovered after each serial pas-specific, resulting from slight trauma associ- sage, a subculture was kept and a comparisonated with inoculation. As a result of separation, of the morphological changes was made afterthe area of membrane forming the 'floor' of the each passage, and at the end of the series.artificial air-space was always slightly thicker Plates of Sabouraud's agar (25 ml. per plate)compared with the surrounding chorioallantoic were inoculated singly with each series of iso-membrane in contact with the shell membrane. lates, and the plates incubated at 30°C. Macro-Slight hemorrhage always occurred with sepa- scopic and microscopic examination of eachration of the two membranes, giving rise to colony was made. slight opacity or granulation around the pe- riphery of the 'floor'. Hemorrhage was clearly RESULTS visible as pigmentation on occasions where this I. General Observations operation had been too hasty or clumsy. The general appearance of the infection was Macroscopic appearance of focal lesions the same for all 3 strains of T. rubrum. Four By using the technic described for inocula-days after inoculation, the shortest period be- tion of the chorioallantoic membrane, the focifore harvesting, the lesions were manifest as of infection were well distributed, with a slightopaque, almost white nodules, varying in size tendency to accumulate around the peripheryfrom 1 to 3 mm. in diameter. With a longer and also to develop in the centre, where per-incubation period, i.e. 7 or 8 days after inocu-

FIG. 1. Chorioallantoic membrane showing foci of infection (arrowed) by Trichophyton rubrum. (7 day infection, strain (a). Stained Periodic-acid-schiff.) X2 CHORIOALLANTOIC MEMBRANE FOR CULTURE OF DERMATOPHYTES 609

p

FIG. 2. Development of parasitic phase (arthrospores) on contact with ectoderm. (8 day infection, strain (a), passage I. Stained P.A.S.) X600 lation, the general response was more pro-itself is a thin sheet of tissue composed of an nounced (Fig. 1.) The membrane had becomeouter cctoderm and inner entoderm, each con- even more thickened and nodular, the lesionssisting of a single layer of flattened cells. Be- tending to become confluent and deepening totween these lies the mcsodcrm, which acts as a cream in color. Inoculation of one batch ofsupporting tissue to the network of blood ves- eggs (passage II) was tried after only 7 dayssels and capillaries. incubation instead of the standard period of Trichophyton rubrum grew readily on these 10 days, and the membranes harvested after atissues, and their response to fungous invasion further 7 days incubation. Although the chori-followed the same pattern for all 3 strains. The oallantois was not fully developed at this earlydevelopment of this organism from the sapro- stage, a mild infection was visible, comparablephytic phase, i.e. spores and hyphac present in with a 4 day infection in the previous batchthe inoculum, into a parasitic phase and its (passage I.). adaptation to a parasitic mode of life was rapid. Not infrequently, in later passages, a whiteSections of the membrane harvested 4 days aerial fungous mycelium was visible, growingafter inoculation showed that this had already up from the nodules. On occasions, and withtaken place. The spores and hyphac of the inocu- strain (c) in particular, fungous invasion ap-lum had germinated to form mycelium which peared to have spread into the original air-spacewas frequently seen as a loose aerial web of where it produced a white aerial growth on thehyphac over parts of the membrane. Penetra- shell membrane. This spread may have beention occurred where the hyphac were in con- caused by the close position of the artificial andtact with the cctoderm. The initial reaction of original air-spaces or to clumsy handling of thethe membranes was hypcrplasia with prolifer- egg after inoculation, but it was possibly due toation of the cctoderm where fungal hyphac had the increase in parasitism of the fungus, es-infiltrated. Little or no reaction was shown by pecially as this phenomenon tended to occurthe underlying mcsodcrm in these areas. Chains in the later passages. That this fungous growthof arthrosporcs, the parasitic spore form, were was not a contaminant but T. rubrum wasusually produced, either on the surface where proved by culture. Growth of dcrmatophytesthey appeared large and rounded, or within the on the shell membrane has also been demon-proliferating ectodcrm, where they germinated strated by Neuhauser (10). She utilized the(Figs. 2 and 6). This picture of the germinating shell membrane 'floor' of the air-space in non-hyphac producing chains of arthrospores cor- fertile eggs as a means of isolating the 'common'responds with Showalter's observations on other dermatophytcs from pathological material. dcrmatophytcs. Increased parasitic activity of the fungus was Histopatholoqy of the lesions shown by the growth of invasive branching The chorioallantois is primarily the respira-hyphae into the rapidly increasing ectodcrm. tory organ of the embryo and the membraneBeneath this hyperplastic cctodcrm, the mcso- 610 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY

A

FIG. 3. Section through infected membrane, showing hyperplasia and proliferation of ectoderm (7 day infection, strain (a), passage II. Stained P.A.S.) X26 derm showed a corresponding reaction, becom-ment and foci of invasion by the fungus, which ing thickened by oedema and cellular prolifer-acted as a stimulus (Fig. 8). ation, and by the migration of ectodermal cells II. Results of Serial Passage in the form of islets or whorls which became sub- merged in the inflammatory tissue arising fromVariation of Pathogenicity it (Fig. 3). Following response of these tissues, Evidence that serial passage had brought fungous elements were often found deep in theabout an increase in the pathogenicity of the mesoderm, always within whorls of proliferat-fungous strains used was insufficient, patho- ing ectodermal cells (Fig. 4). genicity being assessed by the speed with which Advance of the infection was shown by thethe organism was able to invade the ehori- loss of architecture and degeneration of theoallantoic tissues and bring about death of the proliferating ectoderm in the area below theembryo. However, even in this series, eonsider invading fungus (Figs. 5 and 6), together withing the comparatively small number of eggs a development of fibrous tissue in the form ofinoculated and allowing for variation between fibroblasts and chronic inflammatory infiltra-the batches of eggs used, a difference in the rate tion in the superficial mesoderm lying belowof invasion by the 3 strains of T. rubrurn was oh. the region of infection (Fig. 7). The underlyingserved. There was no evidence that this was entoderm in these areas showed a correspondingdue to the size of the inoculum. tendency to proliferation with the formation Rate of invasion of elongated papillae. The overall reaction of the membrane tissues Strain (a) showed a more pronounced patho. was a response towards the parasitic develop-genicity during serial passage compared with CHORIOALLANTOIC MEMBRANE FOR CULTURE OF DERMATOPHYTES 611

a

FIG.4. A. Hyperplastie eetoderm and fungous elements within whorl of ectoderm proliferating ioto mesoderm. X420 the other 2 strains. In passage I, white nodularsitic activity. Sections through a nodule from a lesions were clearly visible on the membrane7 day infected membrane (inoculated after only 4 days after inoculation. Sections of membrane7 days incubation) showed a high degree of from an S day infection showed fungal infiltra-response to the invading fungus, with hyper- tion and hyperplastic response of the ectoderm,plasia and proliferation of the ectoderm, whorls with a slight proliferation of cells into the meso-of cells containing fungous elements being derm. The fungus, recovered by culture, wasfound in the actively increasing mesoderm passaged through 7 and 10 day old eggs (pas-(Fig. 3). The chick embryos did not survive a sage II) and showed a marked increase in para-third passage with this strain. However, the 612 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY

C, 0

I

FIG. 5. B. Mycelial invasion and degeneration of ectoderm. X610 membranes recovered from these eggs, 6 daysfirst 3 passages, appeared to be less rapid than after inoculation on this occasion (passage III),with the previous strain, and there was a tend- showed evidence of fungous infection, and sec-ency of the fungus to develop a covering aerial tions through one of them showed an intensiveweb of mycelium over the ectoderm. reaction of all the tissues to the invading fungus, With strain(b), nodules on the membrane with proliferation and degeneration of bothwere visible from passage I onwards. Sections eetoderm and mesoderm. of an S day infection (passage I) showed ecto- The reaction of the chorioallantoic tissues todermal proliferation and hyperplasia where the invasion by strains (b) and (c), as seen in thefungus penetrated, with little reaction in the CHORTOALLANTOIC MEMBRANE FOR CULTURE OF DERMATOPHYTES 613

FIG. 6. C. Germinating arthrospore of invading mycelium within degenerating ectoderm. X1200 mesoderm, compared with a 7 day infectionbranes from passages V and VI were generally from passage VI, where sections showed avery thick and nodular. Sections through a marked increase in parasitic activity of the7 day infection from passage VI showed a thick fungus, with a corresponding reaction of thecover of aerial myeelium with an acute reac- tissues including the entoderm. tion of the underlying membrane tissues, compa- With strain(c), there was no obvious para-rable with that seen with strain(a), (Fig. 7). sitic development of the fungus until after theThe fungus was found deep in the mesoderm third passage, when nodules were visible on thewithin whorls of proliferating eetoderm. The membrane 6 days after inoculation. The mem-mesoderm exhibited a marked inflammatory 614 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

II. - • P0 — — a.a —a— si

S. 'I

Fin. 7. Section through infected membrane, showing development of fibroblasts and chronic cellular infiltration of mesoderm around fungus-containing ectoderm whorls, below degenerating invaded ecto- derm. Note aerial web of mycelium over ectoderm. (7 day infection, strain (c), passage VI. Stained P.A.S.) X350 response, with development of fibroblasts and vested on the fourth and seventh day after in- chronic cellular infiltration below the degener-oculation, the approximate time of deaths ating invaded ectoderm. occurring during the incubation periods was estimated by the development and condition Survival of Embryos of the embryo. The viability of the embryos during the course With strain (a), the chick embryos did not of serial passage varied considerably. As the survive a third passage, even 4 days after eggs were examined and the membranes har- inoculation. That the fungus had produced an CHORIOALLANTOIC MEMBRANE FOR CULTURE OF DERMATOPHYTES 615

— FIG.S. Section through nodule showing response of eetoderm, mesoderm and entoderm to stimulus of fungous invasion. (7 day infection, passage V. Stained P.A.S.) X13 infection, in this and subsequent passages, wasExamination revealed a very slight increase in shown by the presence of small nodules andthe granular appearance of the colonies obtained. thickening of the ehorioallantoic membrane,There was no apparent difference between the even though autolysis of the egg contents hadisolates of the same strain and passage recovered frequently commenced. The fungus was re-after 4 or 7 days growth on the membranes. It covered from these membranes, usually on thewas found during the course of this work that malt/tellurite agar. Death of the embryo inthe higher temperature (30°C) was more favor- these later passages may have been associatedable for the growth of T. rubrum isolates, and with a particularly virulent fungous strain. Inwas adopted for all subeulturing. This temper- the eggs inoculated with the other two strainsature was also used by Silva, Kesten and Ben- of T. rubrum, 50 per cent of the embryos sur-ham (3). vived with strain (b) and a greater percentage On completion of the series, sets of Sabou- with strain (c). Further study on a large-scaleraud's agar plates (25 ml. per plate) were investigation is required before any definiteinoculated singly with each series of isolates conclusions can be drawn. obtained from each passage, and incubated for 4 weeks. Morphological examination of each Variation of Culture Strains series revealed the following features: The 3 strains of T. rubrum were recovered in Strain (a)—Changes from the original granular culture from the ehorioallantoie membranestype (Fig. 10) took place in this series, some after each serial passage; subcultures were madecolonies developing white, downy sectors. Micro- on to Sabouraud's agar from the isolates ob-scopic examination revealed the presence of tained. As the series progressed, there appearedmacroeonidia in the original colony type and in to be no radical change in the general macro-the granular sectors of the downy colonies: only scopic appearance in tube cultures of the organ-mieroconidia were present in the downy por- isms. Halfway through the series, sets of Sabou-tions of these colonies. The development of raud's agar plates were made from the isolates.downy sectors in some colonies appeared spas- 616 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY modic, especially when compared with the tubedowny type of colony showed a tendency to cultures, where the granular form was main-develop a slightly more granular form, notably tained throughout the series. by the sixth passage. There was also an increase Strain (b)—The isolates obtained by serialin the rate of pigment production. Again, macro- passage showed an apparent morphologicalconidia not found in the original and early transition from the original white semi-downydowny forms were observed in the later more colony form with slight radial folds, to thegranular colonies. flatter, pink-tinted, finely granular form re- covered after the fifth and sixth passage (Fig. JH5CU55JON 10). Microscopically, macroconidia were not The results of this study showed that the found in the original and early isolates, butchorioallantoic membrane of the developing were plentiful in the granular sectors of thechick provided a suitable medium for the in later forms. vim culture of Trichophyton rubrum. Growth Strain (c)—IDuring serial passage, the originalof this organism compared favorably with re-

•"W* S41/

Fm. 9. Section through human skin, showing invasion of epidermis by T. rubrum. Stained by P.A.S. X900 (by courtesy of Dr. C. Calnan, Royal Free Hospital & St. John's Hospital) CHORIOALLANTOIC MEMBRANE FOR CULTURE OF DERMATOPHYTES 617

FIG.10. Cultures of T. rubrum on Sabouraud's agar, (4 weeks at 30°C.) A. Original granular type colony, strain (a). B. Original semi-downy type colony, strain (b). C. Granular type colony, strain (b), produced after sixth passage through chick membranes. suits obtained by Moore with other dermato-phyte infection, the difference being one of phyte species. degree. Both tissues originate from the same The mode of invasion and overall pattern ofgerminal layers. In man, the epidermis arises tissue response to T. rubrum was the same forfrom the ectoderm, and the dermis with its all 3 strains studied. The difference betweenvascular supply, is mesodermal in origin. In- them lay in their rate of adaptation to a para-fection of the chorioallantoic membrane tended sitic mode of life. Reversion to this phase wasto cause chronic granuloma and development of indicated by the germinating spores and hyphaefibrous tissue, resulting in nodular lesions. of the saprophytie phase (the inoculum) pro-Acantholysis of the hyperplastic ectoderm in ducing arthrospores and invasive hyphae oncontact with the invading fungus was usual at contact with animal tissue, i.e. ectoderm. this stage. The fungous mycelium, although The reaction of the membrane to parasiticpenetrating deep into the inflammatory meso- invasion by T. rubrum was comparable withderm, was nearly always found within whorls the response of the human skin to dermato-of proliferating ectodermal cells unless degener- 618 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY ation had taken place. The response to invasionwould therefore appear from the previous tech- of the human skin by T. rubrum produces a lessnic employed that 'mechanical trauma' tender marked cellular reaction, the most noticeableto mask the true picture of the type of fungous feature in the epidermis being hyperkeratosis.infection produced. Fungous growth is limited to the epidermis Although the 3 strains of T. rubum produced a (Fig. 9). similar response in the chorioallantoic tissues This procedure for in vivo culture proved boththere appeared to be a difference in their inva- simple and effective. Although it was not possi-siveness and the production of lesions. This was ble to observe the daily growth of the fungusmost obvious with strain (a) which quickly through the 'glass window' as in the previousreverted to its parasitic phase and caused technics employed, the condition of the em-infection. It was apparently a virulent strain bryo and the artificial air-space created aboveproducing a corresponding picture in human the chorioallantois could always be checkedtissues, where it caused a kerionic infection in a daily by candling. child's scalp. The development of infections by The advantage of this modified method laythe other two strains was less marked, though not so much in the increased area of the unex-indications of an increase in parasitism became posed chorioallantoic membrane available formore emphasised during serial passage. Strain inoculation, but in the marked decrease in(c) showed a slower reversion to the parasitic traumatic damage to the delicate membranephase than strain (b). Even when this did occur during its preparation and inoculation. Theand an infection was well established, the sapro- technique employed by Moore involved a cer-phytic phase was still evident as aerial growth tain amount of mechanical trauma of the limitedover the lesions. region of underlying chorioallantois available The morphology of the isolates recovered after for inoculation. The 'window' was first cut inmembrane inoculation showed that during the the shell, and the shell membrane penetratedcourse of 6 serial passages, there was a tendency with a 'spear-point needle and cut along thefor the semi-downy strain (b) and the downy shell incision', both being removed to exposestrain (c) to produce more granular colonies the underlying chorioallantoic membrane whichsimilar to those of strain (a). This observation was inoculated direct with the fungus. would seem to correspond to a similar conversion The results so obtained showed a violentfrom downy to granular forms, recorded by George inflammatory reaction of the tissues. With M.in the case of T. mentagrophytes. Future world cartis, E. floccosum and T. schoenleini, though toincludes a study of this organism. a lesser extent with T. mentagrophytes, Moore With the development of granular type recorded a loss of continuity in the ectoderm,colonies, a corresponding increase in the pro- and its replacement with necrotic inflammatoryduction of macroconidia by strains (b) and (c) was tissue, arising from the mesoderm. The mesodermalso recorded. Benham (11) observed that are itself showed marked oedema with perivascularartificial medium, blood agar base, stimulated infiltration and an increased number of distendedthe production of macroconidia in culture, and capillaries. The overall tissue response to invasionsuggested that tryptose was the ingredient by all 4 dermatophytes simulated the histologicalresponsible. In this present study, the production appearance of 'traumatic ulcers'. of macroconidia in cultures obtained on artificial By use of a non-mechanical procedure formedia (e.g. Sabouraud's agar) would appear to separating the shell membrane and underlyinghave been stimulated by animal passage, i.e. by chorioallantoic membrane, no breaks in thethe chorioallantoic membrane. Macroconidia are ectoderm were observed, and there was no notice-only occasionally seen in primary isolates froir able change in the vascular supply, although theman. Conidia are a means of dispersing the fungus mesoderm showed an inflammatory response. Noin the saprophytic phase, and the thicker-walled lesions resembling 'traumatic ulcers' weremulticellular macrocondium constitutes a more observed using this technic. Although a differentresistant spore form than the single-celled micro- organism was used in this study, it seems unlikelyconidium. Their production in culture therefore that it would produce less tissue reaction than theis a response of the fungus to environmental other dermatophytes, especially strain (a). Itconditions. The development of macroconidia in CHORTOALLANTOIC MFMBFANE FOR CULTURE OF DERMATOFHYTES 619 the cultures grown from emulsions of the infectedratories, Colindale, London, by courtesy of the membranes harvested during serial passage,late director, Col. H. J. Bensted and Dr. C. H. P. could thus be attributed to a reaction of theBradstreet of the Standards Laboratory. My fungus to previous parasitic stimulus within thethanks are due to Dr. J. J. O'Donnell, St. John's egg. This hypothesis could equally well be appliedHospital for Diseases of the Skin, for assistance to isolates from human infections, the type ofwith the histological interpretation, and to Mr. colony produced being an indication of theR. Lunnon, Institute of Dermatology, for the parasitic condition of the fungus itself within thephotography. tissues, not necessarily shown by the clinical REFERENCES appearance. Carrying this hypothesis further, it might account for the slight variation in isolates 1. Moonx, M.: The chorioallantoie membrane of the developing chick as a medium for the of the same strain obtained from different areas cultivation and histopathologic study of the of the body. pathogenic fungi. Science, 89: 514, 1939. Am. J. Path., 17: 103, 1941. SUMMARY 2. SHOWALTER, W. V.: Morphological studies of dermatophytcs in chick embryo membranes. The ehorioallantoic membrane of the develop- Tr. Kansas Acad. Sc., 57: 149, 1954. ing chick was us9d for the in vivo culture of 3. Siiv, M., KESTEN, B. M. ANn BENHAM, R.W.: Trichophytonrubrum infections: a clinical, Trichophyton rubrum. The technic employed was mycologic and experimental study. J. based on the creation of an artificial air-space Invest. Dermat., 25: 311, 1955. 4. CEORG, L. K.:Therelationship between above the membrane, which was inoculated with downy and granular forms of Trichophyton a fungous suspension through a small aperture. mentagrophytes. J. Invest. Dermat., 23: 123 1954. By this method, the chorioallantois remained 5. GOODPASTURE, E. W. AND BUDDINGH, G. J.: unexposed and relatively undamaged during The preparation of antismallpox vaccine preparation. by culture of the virus in the chorioallantoic Results of histopathologie studies revealed membrane of chick embryos and its use in human immunisation. Am. J. Hyg., 21:319, that the fungous lesions produced were not 1935. masked by mechanical trauma compared with 6. BUENET, F. M.: A virus disease of the canary of the fowl-pox group. J. Path. & Bact., 37: previous technics. 107,1933. Three strains of T. rubrum were studied. The 7.BEYERJDGE, W. I.B. ANDBURNET, F.M.: The mode of invasion and overall pattern of tissue cultivationof viruses and Rickettsiae in the chick embryo. H.M.S.O. Special Report response were the same for all 3, although the Series, Med. Res. Council, London, No. 256, rate of infection varied. 1946. (Reprinted 1955). Preliminary observations on serial passage 8. ALExANnER, R. A.: Studies on the neurotropic virus of horse sickness. VI. Propagation in (6 were carried out), using the modified technic the developing chick embryo. Onderstepoort described, indicated a slight increase in parasitic J. Vet. Sci., 11:9, 1938. adaptation by two strains, accompanied by a 9. GEORG, L. K.: A simple and rapid method for obtaining monospore cultures of fungi. gradual conversion from downy to granular type Mycologia, 39: 368, 1947. colonies together with an increase in production 10. NEUHAUSER, I.: Avian egg-shell membrane as of maeroconidia. a culture medium for dermatophytcs. Arch. Dermat. & Syph., 75: 401, 1957. ACKNOWLEDGEMENTS 11.BENHAM,R. W.: Effect of nutrition on growth andmorphology of the dermatophytes. I. The experimental work in this study was Developmentof macroeonidia in Tricho- carried out at the Central Public Health J4abo- phyton rubrum. Mycologia, 40:232, 1948.