Parasitic Protozoa

55113

Preface to the Second Edition

Julius P. Kreier; John R. Baker

The second edition of Parasitic Protozoa follows the first edition by approximately 14 years. During this time new information about the parasitic protozoa has accumulated. This edition attempts to accommodate the new information without missing the goal of the first edition, which was to present a balanced review of the status of parasitic protozoa with solid information not likely to become quickly outdated.

All of the chapters have been completely rewritten, some by the original authors. In some cases new authors have been chosen because previous authors and dear friends have died, among whom are R. H. Whittaker, A. Zuckerman, and Earl H. Fife, Jr. In other cases, the original authors were not available for a variety of reasons: some have retired, some changed fields, some no longer wished the task, and regrettably we have simply lost track of some.

Some changes have been made in coverage. There has been some expansion in the coverage of the protozoa affecting animals in the aquatic environment, and the reviews of the rickettsial organisms in the Anaplasmataceae, Bartonellaceae, and Ehrlichieae are no longer included. The introductory chapters on broad classification and taxonomy are very different from those in the first edition. A new chapter entitled The Nature of Protozoa has been added. The chapter on broad classification is based on cladistics and takes a very different view of the biological system from the corresponding chapter in the first edition. The chapter on sys- tematics of parasitic protozoa has also been much changed and reflects the state of flux in protozoan taxonomy that exists today. In many respects a better grasp of the areas of taxonomy and systematics can be gained by a comparative reading of the chapters in the first and second editions than by just reading the new chapters in the second edition.

We wish to thank the staff of Academic Press for their valuable aid in preparation of these volumes, and we wish to give special thanks to Edna Chandler who faithfully transformed much editorial scratching into clear, correct, and legible transcript.

Preface to the First Edition

Julius P. Kreier

The parasitic protozoa are a large and diverse group. Many are of interest to physicians and veterinarians because they produce disease in man and his livestock. Others, which seldom produce disease, should be familiar to the practitioner of medicine and to the research scientist because they are present in the animal body and thus must be recognized to avoid a misdiagnosis, while still others, such as the intestinal and rumen protozoa, perform a useful function in the animal’s economy, and their presence is an indication of health rather than disease.

I have included in these volumes protozoa parasitic in animals, such as fish and insects, which are not usually included in books on pathogenic protozoa. I did this because I believe veterinary medicine should concern itself with all species of animals, excepting man, whose care falls to the physician. From a more practical standpoint, I feel the inclusion of parasites of diverse species is appropriate in a book on protozoa of veterinary and medical interest because no matter how we set ourselves off from nature we remain a part of it, and thus we inevitably share parasites with the other species with which we live.

Because of the wide range of parasites and the volume of material available, no single author could hope to be qualified to write on all of them; thus I have chosen to have each chapter written by someone qualified in that area. This course of action, while it avoids the problems of the limitations of a single author, has problems of its own, the most serious being the variability in the authors’ styles and attitudes which produces unevenness in the treatment of the contributions. For this I accept responsibility as editor. For all that is good and useful in these volumes I thank the authors of the chapters and the staff of Academic Press who have aided in the production of these volumes. I also wish to thank the Army Malaria Project, whose support of my research has made it possible for me to continue my interest in protozoology.

Chapter 1

Trypanosoma rangeli

Antonio D’Alessandro-Bacigalupo; Nancy Gore Saravia

I Introduction

Trypanosoma rangeli is a parasite of man, domestic and wild animals, and triato- mine insects in the New World. In contrast to the pathogenic Trypanosoma cruzi, it is harmless to the mammalian host but damaging to the vector. The distribution of T. cruzi and T. rangeli often overlaps and they infect the same vertebrate and invertebrate hosts; therefore, their differentiation becomes of real, practical importance. Biologically, T. rangeli is unique in several respects. During development in the insect vector T. rangeli penetrates from the intestine into the hemocoel and then enters the salivary glands. Inoculative transmission to vertebrate hosts occurs during probing for a capillary during biting. The question of whether posterior station transmission occurs has been controversial but we know now that it may occur.

Taxonomically, T. rangeli was included by Hoare (1972) in the subgenus Herpetosoma, section Stercoraria. It is considered to be a possible phylogenetic link between the Stercoraria and the Salivaria. Recent observations support this position (D’Alessandro, 1976 and unpublished observations; see also Section IIIA).

II Morphology in Blood and Culture, Life Cycles in Vertebrate and Invertebrate Hosts; Transmission

The blood trypomastigote of T. rangeli in the vertebrate has the general characteristics of T. lewisi and other members of the subgenus Herpetosoma. It is a large, slender trypanosome, averaging from 26 to 34 μm in length including the flagellum. The undulating membrane is well developed and the nucleus is located in the anterior half of the body; the kinetoplast is small, round, and subterminal (Figure 1.1). The measurements presented here were drawn from various reports by D’Alessandro (1976). The mean total length has been reported to range from 26.4 to 33.8 μm (range 25–37 μm). The mean distance from the posterior end to the center of the nucleus is from 12.9 to 14.8 μm (range 10–17 μm). The mean distance from the center of the nucleus to the anterior end of the body is from 6.9 to 8.9 μm (range 5–12 μm). The mean nuclear index is between 1.6 and 2.0 (range 1.1–2.8). The mean diameter of the kinetoplast is 0.7 μm; the mean distance from it to the posterior end of the body is between 3.4 and 4.4 μm (range 1.8–7 μm). The mean distance from the posterior end to the nucleus is between 9.5 and 9.7 μm (range 8.2–10 μm); and the mean length of the free flagellum is from 8.1 to 9.5 μm (range 5–11 μm) long.

Figure 1.1 Trypanosoma rangeli in mouse blood smear (day 6 after infective bite); (x2000).

A LIFE CYCLE IN THE VERTEBRATE HOST

Little is known about the life cycle of T. rangeli in the vertebrate host, although thousands of human infections and many natural and experimental infections in several species of domestic and wild mammals have been observed (Table 1.1). Unlike most other members of the subgenus Herpetosoma, T. rangeli lacks host specificity.

Table 1.1

Wild Animals Naturally Infected with Biologically Proven Τ. rangeli

Note: Unnamed or ill-defined T. rangeli have been mentioned from several other mammalian species but details for conclusive diagnosis were absent (see Table 12, in D’Alessandro, 1976).

a References: D’Alessandro et al., 1984; D’Alessandro and Barreto, 1985; Christensen and Herrer, 1979; Zeledon et al., 1975; Sousa and Dawson, 1976; Sousa et al., 1974; Miles et al., 1983b; D’Alessandro et al., 1986; Deane et al., 1972.

b 15/70 (21%) with T. rangeli and 1/70(2%) T. cruzi.

c In Colombia, where more than 2000 animals were carefully examined and 109 isolates well evaluated, 64 were T. cruzi, 14 T. rangeli, and 31 had both infections. The overall infection rate was 95% T. cruzi versus 45% T. rangeli.

d Of 122 animals collected in the Pacific slopes of Panama Canal zone 95% had T. rangeli and 40% T. cruzi.

Some early observations suggested that T. rangeli was able to produce clinical symptoms. However, it is now generally agreed that the parasite is not pathogenic for the mammalian host but rather to the insect vector, this being one important characteristic of the parasite. As a consequence, the prepatent period in man has been reported in only a few experimentally infected persons in whom it ranged between 22 and 108 days (D’Alessandro, 1976). This diversity was probably due to the low parasitemia, perhaps due to a dilution effect, a consistent feature particularly in a large host such as man. In experimental animals infected by insect bite, metatrypomastigotes from the salivary glands of the insect directly enter the blood because triatomines are capillary feeders. This feature of the infection can be demonstrated by feeding infected and uninfected bugs simultaneously on a mouse. When this is done, the uninfected bugs become infected (D’Alessandro, 1961). As had been observed when African trypanosomes are transmitted by the bite of tsetse flies, when a triatomine is probing the skin of a mammal for a capillary, it may deposit a large number of metatrypomastigotes and rounded forms, some dividing, in the tissues. The number of trypomastigotes in the host’s blood increases with time after the bite. The early blood forms morphologically resemble the trypomastigotes in the salivary glands. As the infection progresses, they become progressively larger, and the fully developed blood form is present by the third day. The round forms may also develop into trypomastigotes. The process of transformation into blood forms ended by the first day in baby mice and by day 3 in adult mice, at about the time of the peak of parasitemia (Aeñz, 1981).

The duration of parasitemia in man may be as long as 18 months. In mice and rats parasitemia may last from 7 to 12 months, and may last from 16 months to 3 years in larger animals but is usually of short duration (2–3 weeks) (D’Aless- andro, 1961). The level of parasitemia is usually low, from 5 to 7 trypanosomes or fewer in 5 mm³ of blood.

Parasitemias of 5000/5 mm³ of mouse blood have been reported, which are higher than what would result from the original number of culture forms of the Venezuelan strain of T. rangeli used to produce the infection, if no reproduction occurred. This observation merits confirmation (Urdaneta-Morales and Tejero, 1986; Tejero et al., 1988). Parasitemias in two baby mice produced by the inoculation of massive numbers of infective salivary gland forms or by the bite of infective vectors are the highest reported. They reached 150,000 and 20,000/5 mm³, respectively. Despite the high levels reached, parasitemia, nevertheless, lasted for only about 2 weeks (Grewal, 1969; Aeñz, 1981). In our experience, parasitemias of 900/5 mm³ were obtained in mice bitten by Rhodnius prolixus, which had been inoculated intracoelomically with a cyclically maintained stock (D’Alessandro and Hincapie, 1986).

The reproductive phase of T. rangeli in the vertebrate host is not known. Most searches for tissue forms have been unsuccessful. There are three recent studies that provide data to the contrary; however, they need to be confirmed using cloned strains of T. rangeli to eliminate any possibility of contamination with T. cruzi or by use of hybridization in situ with species-specific DNA probes to demonstrate clearly that the intracellular forms are T. rangeli. The illustrations in two of the papers show pseudocysts and amastigotes indistinguishable from those of T. cruzi (Grogl and Kuhn, 1984; Scorza et al., 1986; Urdaneta-Morales and Tejero, 1986).

Blood trypomastigotes that appear to be in division have been observed in man and rodents. These observations are most frequently made on individuals with recent infections induced by the bite of infective vectors and may be forms that were dividing in the vector (D’Alessandro, 1976). This case is supported by the observation of dividing metatrypomastigotes in the saliva and salivary glands of vectors from where they could readily be injected (D’Alessandro, 1972; Cuba Cuba, 1974a; Aeñz, 1981).

It seems unlikely that T. rangeli merely survives without multiplying in the vertebrate host in the light of the existence of long-lasting infections of up to 3 years duration. Based on the observation that culture forms of T. rangeli added to monolayers of He La and sarcoma cells at 37 °C developed into blood forms, some with signs of division, it was suggested that T. rangeli could also reproduce in the mammalian. host (Molyneux, 1973). However, Molyneux (1973) stated that other members of the subgenus Herpetosoma divide either as amastigotes or epimastigotes. The observation of dividing round forms, with or without a flagellum, in viscera of animals inoculated with T. rangeli suggested that these may be the replicating forms of the parasite. However, these forms may have originated in the salivary glands of the vector and been introduced during a blood meal (Deane, 1969; Aeñz, 1981).

B LIFE CYCLE IN THE INVERTEBRATE HOST

Among the unique features of the life cycle of T. rangeli in the insect are invasion and multiplication in the hemocoel and invasion and reproduction in the salivary glands. Infective metatrypomastigotes are formed in the salivary glands (Fig. 1.2). In addition to transmission by bite, there may be transmission by the fecal route. Flagellates are excreted with the feces and transmission by inoculation of flagellate-bearing feces has been reported to occur; however, this mechanism is controversial and not accepted by all investigators. To make matters more complicated, T. diasi T.myrmecophagae, T. cebus, T. saimiri, T. barnolai, and other T.rangeli-like organisms have been observed in the intestine, but not in the hemolymph or salivary glands of various triatomine insects. Therefore, the occurrence of trypanosomes in the insect’s gut is not proof that it is a vector. Before a given species can be accepted as a vector of the parasite, transmission must actually be demonstrated. Unfortunately, necessary experiments to demonstrate experimental transmission have seldom been carried out and therefore, while a considerable number of genera and species of triatomines have been found to have T. rangeli-like flagellates in their intestines, which of them are vectors is unknown (D’Alessandro, 1976). The reports of infection of insects and of insect transmission of infection are summarized in Tables 1.2, 1.3, and 1.4. In most instances the species tabulated in Table 1.2 were foreign to the strain of T. rangeli used and, therefore, their capacity as vectors for sympatric parasites cannot be excluded.

Figure 1.2 Trypanosoma rangeli (Giemsa-stained; x2000) in salivary glands of (A) Rhodnius prolixus, (B) Triatoma infestans, (C) R. neglectus, (D) T. protracta, (E,F) T. patagonica. C and D show dividing metatrypomastigotes (also observed in the other species of bugs). F shows mass of meta- trypomastigotes recognized by the evident round kinetoplast well separated from the nucleus. (Reprinted from D’Alessandro, 1972 by permission of the editors, Journal of Medical Entomology.)

Table 1.2

Records of T. rangeli and T. rangeli-like Flagellates in the Intestine of Triatomines with Uncertain Vector Capacitya

a Updated from D’Alessandro, 1976. Only D’Alessandro et al., 1981, which was not cited in the original report, is cited here.

b Salivary glands and/or hemolymph reported free of infection.

Table 1.3

Records of T. rangeli in Hemolymph but Not in the Salivary Glands of Triatomines and Cimex with Uncertain Vector Capacity a

a Updated from D’Alessandro, 1976. Only the references not cited in the original report, viz.

b Cubâ Cuba, 1975a.

c Mfinter-Goedbloed and Oliveira, 1976.

d Marsden et al., 1979; D'Alessandro and Hincapie, 1986, are cited here.

e Also natural infection.

Table 1.4

Records of T. rsmgeli in Salivary Glands of Triatomine Vectorsa

a Updated from D’AIessandro (1976). Only references not cited in the original report, viz.

b Cuba Cuba, 1974a, 1974b, 1975b.

c D’AIessandro et al., 1981; D’AIessandro and Barreto, 1985.

d Carcavallo et al., 1975; Miles et al., 1983b.

e Carcavallo et al., 1975; Miles et al., 1983b; Aeñz and East, 1984.

f D’Alessandro and Hincapie, 1986.

g Tovar and Urdaneta-Moraies, 1989, are cited here.

Table 1.4 lists the natural and experimental vectors of T.rangeli, that have been shown to actually transmit the infection by bite. Rhodnius prolixus was the first recognized vector and has the most extensive distribution. In Panama and Peru, and probably in Ecuador, it is replaced as the common vector by R. pal- lescens and R. ecuadoriensis, respectively. The specimens of R. dalessandroi (previously considered to be R. brethesi) were collected from a Colombian palm tree. They transmitted T. rangeli by bite. The T.dimidiata (− T. dimidiata capitata) studied was collected within human dwellings in a Colombian village and was reported to have T.rangeli in its salivary glands. This is the only report of a Triatoma species with a natural infection. The R. robustus and R. pictipes were collected from palm trees in Venezuela and Brazil and were found with T. rangeli in their salivary glands. The R. neglectus, T.protracta, T. patagonica, T. vitticeps, and T. infestans studied were proven, the latter by three investigators, to be anterior station experimental vectors of foreign stocks of T. rangeli (see Figure 1.4).

Trypanosoma rangeli develops cyclically in insects in any nymphal stage and in insects of both sexes. The infection is not lost during molting.

The behavior of T. rangeli in the invertebrate host varies with the strain and age of the stock, and with the way the parasite has been maintained; e.g., culture tube to culture tube, mouse-bug-mouse, bug-mouse-culture-bug, and so forth.

In general, the flagellates ingested with a blood meal begin to multiply after they reach the midgut. Some reach the rectum and are excreted with the feces. They do not colonize the rectal wall as does T. cruzi (D’Alessandro, 1961; Zeledon and Blanco, 1965). Some ingested flagellates may enter the hemolymph. From there the parasites may invade the salivary glands where they reproduce and differentiate into metatrypomastigotes infective to the mammalian host. A similar direct short-circuit route from gut to salivary gland has been reported to occur in tsetse flies infected with African trypanosomes (Otieno et al., 1976; Minter, 1989).

It has been reported that small flagellate or aflagellate forms, round or oval, from about 5 to 7 μm in diameter occur in the lumen of the wide and slender midgut. There are, however, many more epimastigotes and trypomastigotes with slender posterior ends, small, round kinetoplasts, and a poorly developed undulating membrane in that location. These latter forms are also present in the feces. The epimastigotes have been reported to range from 28 to 67 μm in length and the trypomastigotes from 23 to 49 μm (D’Alessandro, 1976; Vallejo et al., 1988). The dividing forms are mainly epimastigotes. Short trypomastigotes have been seen in feces, although their morphological features are quite different from those of the metatrypomastigotes in the salivary glands (Tobie, 1964).

The flagellates in the hemolymph are rather similar to those in the intestine, being epi- and trypomastigotes. When the insect is first infected, there are few parasites in the hemolymph, but later huge numbers of long epimastigotes with particularly elongated posterior ends develop there. Many of these undergo binary division. In some instances the large numbers of organisms give the hemolymph a whitish color. There may also be forms in multiple division creating huge masses from which epimastigotes emerge. Metatrypomastigotes, some dividing, are also present in the hemolymph. In addition to extracellular forms, there are flagellates inside hemocytes. These may be either coiled, long epi- or trypomastigotes, dividing amastigotes, or spheromastigotes. These forms have been considered to be developing stages by most workers, but some consider them to be phagocytosed parasites, about to be digested by the hemocytes. Probably in some instances digestion by hemocytes can occur (D’Alessandro, 1976).

Detailed studies of the development of various strains of T.rangeli in R. prolixus, R. ecuadoriensis, and R. robustus have been published (Tobie, 1970; Cuba Cuba, 1974a,b, 1975a,b; Aeñz, 1981,1983a,b). Minter (1989) has used these descriptions to make a drawing from which the reader may obtain a comprehensive vision of the various developmental pathways T. rangeli may take in the triatomine vector (Figure 1.3).

Figure 1.3 Developmental pathways of T. rangeli in Rhodnius spp. The nonmultiplicative bloodstream trypomastigotes (1) are drawn into the wide midgut lumen (2) where multiplication occurs; epimastigotes and trypomastigotes (long and short) are the predominant morphological types. Similar forms occur in the rectum (3) and are discharged with the feces, in which short trypanosomes may also be found; inoculation of feces sometimes leads to: (4) posterior-station transmission. No colonies of attached epimastigotes have been found in the rectum. Parasites escape from the gut (2) into the hemolymph. Some parasites enter hemolymph and undergo intracellular development (from 7 to 9), in which dividing amastigotes and spheromastigotes are the predominant forms, but trypomastigotes arise from the unrolling of vacuolated spheromastigotes and are liberated when the hemocyte ruptures (8). The fate of the trypomastigotes is not known (9). Most parasites in the hemolymph are extracellular; epimastigotes and trypomastigotes again predominate. Binary division of epimastigotes may give way to multiple-division forms (5). Epimastigotes finally form a palisade along the outer membranes of the pyriform salivary glands, where giant multinucleate forms may also occur, before penetrating (6) the gland cells flagellum foremost to complete development to the infective metatrypanosomes, which are inoculated into a new vertebrate host in the saliva. (Modified and reproduced with permission of Dr. Donald Minter and Baillièr e Tindall (Publishers). From Appendix I Medica l Protozoology, p. 1,293. In Manson’s Tropical Diseases, 9th ed. P.E.C. Manson-Bahrand D.R. Bell editors. Baillière Tindall, 1989.)

The penetration into the hemocoel by T. rangeli was studied with the aid of an electron microscope. This study revealed that penetration usually takes place through the slender midgut epithelial cells. Penetration of T. rangeli into the hemocoel may also take place by passage between the slender midgut cells when the infection is not massive (Watkins, 1971a). One to many parasites were observed in parasitophorous vacuoles in undamaged cells. When the parasites left the cells, they did not always leave the anterior end first. While penetrating the basal lamina, T. rangeli were contained in a vacuole. The pores in the cells and basal lamina produced by penetration were repaired. Large numbers of long and slender epimastigotes were present in the hemocoel. These were in or between muscle and tracheal cells and in hemocytes. In this study the hemolymph of 2-5% of bugs was invaded, but the salivary glands were invaded in every one of these bugs examined. The salivary glands of all bugs whose hemocoels were inoculated with flagellates became infected (Hecker et al., 1990).

In R. prolixus and other vectors hemolymph invasion occurred within the first 50 days after the infective blood meal. The invasion times ranged between 15 and 183 days (Groot et al., 1951,1953; Groot, 1954; Grewal, 1957; D’AIessandro, 1972; Cuba Cuba, 1975a). Invasion has been reported to have occurred as early as 24 hours after the blood meal in some bugs (Aeñz, 1979).

About a week after the invasion of the hemolymph, epimastigotes, some undergoing binary and multiple fission, are present in the salivary glands. There are also long trypomastigotes present at this time. By the tenth day metatrypomas- tigotes start to appear. These, within a few days, constitute most of the forms present. There are also some very long epimastigotes (up to 140 μm). Other trypomastigotes may be long or of medium size (13–30 μm or more). The metatrypomastigotes are short (8–13 μm). They have a centrally located nucleus and a round subterminal kinetoplast, sufficiently large to cause the body of the flagellate to bulge. The undulating membrane of the short form is simple and the flagellum is short. The short metatrypomastigotes of T. rangeli look like the short infective forms of T. cruzi (D’Alessandro, 1976; Aeñz, 1981) (Figure 1.2).

In bugs infected by ingestion of flagellates, the time interval between ingestion and production of the metacyclic trypanosomes, that is, forms infective when injected with saliva during probing and biting, is usually 10–17 days or longer (range 10–180 days) (Groot, 1953a, 1953b; D’Alessandro, 1972; Cuba Cuba, 1975b). Directly injecting T. rangeli into the hemocoel does not speed development and the time to infectivity is similar (about 2 weeks) (Tobie, 1961; D’Alessandro and Hincapie, 1986).

Electron microscopic studies show that the parasites accumulate around the salivary gland capsule. The inner layers of the capsule are disrupted and the parasites pass between the muscle cells to reach the basement membrane of the glandular cells. The basement membrane is penetrated flagellum first. The plas- malemma of the cell is invaginated and as a result a vacuole is created in which the trypanosomes cross the glandular cells and reach the lumens of the salivary gland ducts. The outer membranes of the glands enclose many multinucleated giant flagellates of unknown significance (Ellis et al., 1980, 1982). Flagellates of this type were also seen by Watkins (1971b) with the aid of a light microscope.

In the salivary glands tissue various intracellular forms have been seen. These include amastigote-like forms, coiled and uncoiled epimastigotes, and metacyclic trypomastigotes, some in division (D’Alessandro, 1976; Aeñz, 1980,1981). According to Hecker et al. (1990), the parasites divide in the lumen of the gland where transformation to metatrypomastigotes also occurs. The metacyclic trypomastigotes produced remain free in the saliva.

In some infected insects the large numbers of parasites may change the color of the salivary gland from pink to whitish. The numbers of flagellates in the gland have been estimated by various authors to range from 100,000 to 10,000,000/ml (D’Alessandro, 1976; Aeñz, 1983b).

It should be understood that in our experience, and in that of other workers, not all exposed vectors become infected. Infectivity of T. rangeli is probably affected by the way the flagellates have been maintained and for how long. Not only do some vectors fail to become infected in the gut after ingesting T. rangeli but, of those whose guts become infected, only a relatively low proportion develop parasites in the hemolymph; and even in those with hemolymph infection, salivary gland infection may not occur (D’Alessandro, 1976). The distribution of T. rangeli and T. cruzi in the slender midgut, hindgut, and salivary glands of triatomines, infected experimentally by feeding on parasitemic mice and those naturally infected, has been studied by dissection (D’Alessandro, 1972). Trypanosoma rangeli occurred more frequently than T. cruzi in the slender midgut (56-83% versus 53%), but in the rectum the converse was true (16-52% versus 97%). The salivary gland infection rate of T. rangeli was 30% in R. prolixus and 63% in R. neglectus. Twelve and 24%, respectively, of these infections were found in the salivary glands but not in the digestive tract. With the passage of time infections were lost from both species of insects: from gut, 32 and 36%; from hemolymph, 55 and 25%; and from salivary glands, 8 and 20%, respectively (D’Alessandro, 1972).

There is some evidence that insects other than triatomines and Cimex (Table 1.3) may become infected with T. rangeli. In one instance T. rangeli was isolated from two wild sandflies captured in Brazil (Miles et al., 1983b). In Panama sandflies fed on two-toed sloths infected with a parasite considered to be T. rangeli and one of 292 developed epimastigotes in its gut (Christensen and Herrer, 1979). The role of sandflies in T. rangeli transmission appears, however, to be negligible.

Triatomines may be infected by ingestion of trypomastigotes from a vertebrate host, including metatrypomastigotes just injected by another bug. Triatomines may also be infected by feeding on each other. In a laboratory small nymphs have been infected by ingesting parasites from blood in the gut or hemolymph of another bug. This mechanism is known as infection by cannibalism or haemo- toklepty, but it is not known how important it may be in maintaining the parasite in nature (D’Alessandro, 1976; Aeñz, 1982a).

C TRANSMISSION TO VERTEBRATE

Anterior station or inoculative transmission to vertebrates is the indisputable route of infection. The vector injects infective metatrypomastigotes with the saliva during probing or biting. This method of transmission is more efficient than the posterior station or contaminative method used by T. cruzi. The efficient nature of transmission is one of the explanations for the maintenance of transmission in the presence of low levels of salivary gland infection found in bugs. Table 1.4 lists the species proven to be inoculative vectors of the parasite in nature or under experimental conditions.

As mentioned earlier, only in a small proportion of bugs exposed to T. rangeli are the trypanosomes able to complete the cycle of transmission under experimental conditions. Many stocks, particularly those that have been maintained by serial culture, rarely infect the bug’s salivary glands after ingestion. To increase rates of infection parasites can be inoculated directly into the hemocoel. The salivary gland infection rate, when hemocoel inoculation is used rather than feeding of the same stock through a membrane, increases from 2.5 to 70% (D’Alessandro et al., 1986). Ingestion of bugs with infective salivary glands by mice did not produce infection, but the inoculation of forms obtained by dissection of salivary glands did. Infection resulted from inoculation by a variety of routes including the subcutaneous (7/8 mice), intraperitoneal (2/8), and intracardial (2/3) routes, and by instillation into the stomach by a gastric tube (1/17) (D’Alessandro, 1976).

Posterior station transmission is controversial. At one time D’Alessandro, the senior author of this chapter, after evaluating the few successful attempts reported, concluded that very probably they were spurious. D’Alessandro concluded that infections were probably caused by contamination of the feces with metatrypomastigotes from the salivary glands and that the contamination may have resulted from metatrypanosomes sucked into the digestive tract from the salivary glands as the bugs were feeding (D’Alessandro, 1961). At present D’Alessandro is no longer sure that fecal transmission does not occur.

The parasite remains viable and infective to mice for up to 4 days in the wide midgut but not in the slender midgut or in the contents of the rectum. However, if wide midgut contents are mixed with feces, the mixture is infective (D’Alessandro, 1976). Since 1976 we have been able to infect mice with T. rangeli by inoculation of feces of/?, prolixus. We have been particularly successful with bugs that had been recently brought to the laboratory from the field and with bugs infected with a cyclically maintained stock of T. rangeli. In these studies fecal inoculation was made by syringe, and the resulting infections were proven to be caused by T. rangeli on the basis of the morphology and behavior of the parasite when it was isolated from the infected mice and grown in hemoculture (D’Alessandro, unpublished observations; D’Alessandro et al., 1986; Miles et al., 1983b). The flagellate forms responsible for transmission in these studies have not been determined. Based on these experiences, we now believe that transmission by fecal contamination does occur. As both types of transmission occur, T. rangeli can be regarded as a link between stercorarian and salivarian flagellates.

III

Definition, Taxonomy, and Nomenclature

The term "T. rangeli-Uke" does not merely suggest that the parasite is a species of the subgenus Herpetosoma (type species T. lewisi), but rather that it can develop in triatomine insects. Trypanosoma rangeli has a wide host range and, except in those cases when development in the bug is proven, a parasite should not be identified as T. rangeli-Mke. Various mammalian trypanosomes of doubtful taxonomic status develop in the intestines of triatomines and flagellate stages indistinguishable from those of T. rangeli develop there. To confirm that a trypanosome is T. rangeli, efforts should be made to demonstrate typical development in the salivary glands and transmission to mammals by bite as well as by inoculation of feces. The organism must be isolated from the animal and established in hemoculture, if one is to confirm that it is T. rangeli. Because T. rangeli only infrequently crosses the intestinal wall and enters the hemocoel following normal feeding, it is desirable to inoculate the parasite directly into the hemocoel when evaluating the ability of the organism to grow there. As stated elsewhere in this chapter, T. rangeli is more likely to be transmitted by bite, if infection is induced in the insect by inoculation into the hemocoel rather than by feeding (71% versus 2.5%) (D’Alessandro et al.,1986).

A TAXONOMY

The key issues in the taxonomy of T. rangeli are the controversial transmission of trypanosoma by triatomine feces and the fact that the short trypomastigotes present in the feces are morphologically different from those observed in the salivary glands. Hoare (1972) stated that T. rangeli displayed characteristics of both Stercoraria and Salivaria. That is, Hoare believed that T. rangeli had the essential features possessed by Stercoraria and in addition had become adapted to development in the anterior station in the salivary glands. Hoare considered T. rangeli to be a member of the subgenus Herpetosoma but one providing a clue to the transition between the Stercoraria and Salivaria. D’Alessandro (1976) agreed with this view, despite the fact that he could demonstrate only anterior station transmission in the strains with which he worked. D’Alessandro interpreted the posterior station transmissions observed by other authors as possibly being spurious infections caused by contamination of the feces with parasites in blood or in the cultures used to infect the insects involved in the experiments. He reported that half of the 85 Herpetosoma isolates from nonhuman South American primates he studied infected triatomines, which yielded feces infective to mice by inoculation. He also reported that the infections did not pass from the guts of the insect to the salivary glands. D’Alessandro classified the trypanosomes as trypanosome I, infection of the gut alone, or trypanosome II, infection of the gut and hemolymph. D’Alessandro suggested that the type of contaminative transmission observed in the strains derived from nonhuman primates was so different from that of T. rangeli isolated from man, triatomines, and animals other than primates that these strains may represent a different trypanosome species. He further suggested that these flagellates that infect nonhuman primates may constitute an evolutionary link between Stercoraria and Salivaria: T. cruzi being at one end (strictly contaminative transmission) and T. rangeli at the other end (strictly