Chemical trails and the parasites that follow them

Parasitic relationships constitute an important ecological association, responsible in large part for controlling the population of all living things. Parasitism spans the entire breadth and depth of the tree of life from viruses to 40-foot-long tapeworms living in the intestinal lumen of 40-foot-long whales. Infectious agents keep evolutionary pace with their hosts by natural selection, becoming increasingly more adapted to restricted (i.e., host species-specific) associations. Well known examples include medically important pathogens, such as the human malarias, down to our innocuous pinworm, Enterobius vermicularis. Although parasitism is commonplace, the precise mechanisms that permit a pathogen to successfully infect its host vary widely. Humans serve as the host for >350 different protozoan and helminth species (1). Although there are common features of their biology that unite some into related groups regarding their means of gaining entrance into our body (fecal–oral, skin penetration, or vector-borne, for example), the details of their lives as pathogens remain unique to each organism. That specific parasitic relationships arose independently many times throughout the last 3.5 billion years of Earth's history now is taken as fact. Hence, what we learn from one association at the molecular level may not apply even to closely related organisms that infect different hosts (e.g., human vs. dog). A number of them have been adapted for study in the laboratory, and a few medically relevant ones and their mechanisms of pathogenicity have been the subjects of intensive investigation for some time. This research activity has increased over the last few years as complete DNA sequences of some parasites have led to the study of their proteins (2). The first obvious step in any infection is that of contact between the two organisms. Just how parasites find their way in the world, both outside and inside the host, is largely unknown (3). Work presented in this issue of PNAS by Safer et al. (4), showing that urocanic acid is an important host molecule detected and responded to by Strongyloides stercoralis, is an important and clear example of the kind of information that is needed to begin to unravel these complex biological interactions. The ultimate reward for doing so presumably will be better-designed control measures that take advantage of these new findings.

For the vast majority of human-specific microbial infectious agents, the organism is brought into contact by our essential behaviors: breathing, eating, drinking, or sexual activity. Arthropod vectors also play an important role in their transmission. In those infection schemes, the pathogen has to do nothing more than be present in the environment. We do the rest. In contrast, several unrelated groups of multicellular organisms (e.g., some larval nematodes and trematodes) seek out their hosts by actively crawling or swimming toward them. Molecules specific to the target organism serve as environmental cues that function to guide them. To accomplish this, these pathogens use well developed nervous systems that are designed to detect, then follow, concentration gradients of heat, CO₂, skin lipids, etc. until they arrive at the source. It is often the case that, during their journey, they use up glycogen stores and die. However, least we take solace in this fact, their sheer numbers ensure that at least some will encounter us and eventually complete their life cycle. For example, the hookworms and S. stercoralis infect some 500 million people each year, and the schistosomes infect some 250 million (1).

Although it is known that these two groups of parasites have the capacity to find us in aquatic environments, the precise way(s) in which they do so remain undefined at the molecular level. The way (i.e., behavior) parasites find their hosts is presumed to encompass a complex set of behaviors all under the control of parasite neurons dedicated to receive a wide spectrum of external sensory inputs. The aquatic cercarial stage of schistosomes finds its host employing several coordinated strategies. That of Schistosoma haematobium can detect and respond to a concentration gradient of l-arginine in human skin (5), a host-seeking response similar in kind (i.e., chemotaxis) to the infective larva of S. stercoralis (Fig. 1). Cercariae, in addition, also use a combination of positive phototropism and negative geotropism behaviors. Speculation favors two different sets of neurons that are simultaneously operational, one set for chemoreception and the other for sensing its physical environment. Coordination of these two sensory inputs, resulting in a successful hunt for a host, is an absolute requirement. The first set combines to bring the cercaria to the surface of its aquatic environment, whereas the second set allows it to hone in on its intended victim. As far as is known, nematode parasites do not have the capability to respond to light sources or the pull of gravity, so these additional host-seeking activities are probably not an option for Strongyloides.

Fig. 1.

Strongyloides-infective larva.

Once in contact with a potential host, another array of recognition signals must be received and processed for the relationship to begin. Each proceeding step in the infection process becomes more and more selective for a specific pair of organisms. Agents that find their hosts by simply following trails of CO₂, body heat (mostly arthropods of various sorts), or methane and body movement, as is the case for tsetse flies and their ruminant targets, tend to be more nonspecific in their choice of vertebrate host than those that have evolved further mechanisms to hone in on molecules that identify them specifically (human vs. cow, for example).

Deciphering the chemical basis of each step has proven to be a major undertaking. That is why so many of them await discovery. The high interest among the scientific and public health communities in this field and the advent of modern biological methods for investigating chemical ecology (6) have yet to be brought to bear in a concerted effort to identify the precise molecules that function to limit the host range for the great majority of parasites that cause human disease. This lack of progress is probably due to the fact that working with many of them would require using human subjects because of their inability to mimic the infection process in other host species. Therefore, identifying even one host molecule responsible for attracting a given parasite to us is cause for celebration. Safer et al. (4) chose to work on S. stercoralis, an important cause of diarrheal disease worldwide, because it infects not only humans but other primates and dogs as well. These mammals all produce ample amounts of urocanic acid. In addition, this parasite can be raised in vitro as a free-living nematode, so continuous production of infective larvae ensures an adequate supply of biological material, making S. stercoralis an ideal laboratory subject for study.

The results presented here advance the field in several respects, from both a medical and public health perspective, because the shared goal of these two broadly defined disciplines is to limit, and in a very few instances eradicate, as many parasites from the human condition as possible. Safer et al. (4) identified a specific host molecule that S. stercoralis uses as its guide to our skin. They then discovered that divalent cations interfere with the parasite's ability to detect it. As so often is the case, an initial discovery of one kind suggests the next step in the overall process. By removing heavy metal ions from the urocanic acid preparation, Safer et al. (4) found that, contrary to prior studies, there was a negative association between the presence of calcium, manganese, or magnesium and the activity of the organic acid in attracting larvae. This finding led to the idea that perhaps a topical application containing harmless amounts of any of these metals could serve as a “smoke screen” for the larva of S. stercoralis, inhibiting its ability to detect its host.

When urocanic acid leads Strongyloides to contact the skin of its host, and perhaps even to penetrate it, the host-seeking phase of its life cycle is complete. What happens next as it begins the migration phase of its life is entirely open to speculation. Presumably, once inside the host, another set of environmentalcues come into play, determining further whether the relationship will continue in a normal fashion or result in an aberrant infection. In the case of S. stercoralis, the larva undergoes a “fantastic voyage” throughout the body, eventually ending up in the small intestine, where it penetrates the epithelium and becomes a multiintracellular adult parasite (Fig. 2). In contrast, hookworms that infect dogs, Ancylostsoma caninum and Ancylostsoma braziliense, both of which are distant relatives of S. stercoralis, fail to complete their life cycle after penetration of our skin. The canine-specific larvae wander aimlessly in the subcutaneous tissues until they die, causing a disease referred to as “creeping eruption,” unable to advance to the next step in their life cycle that ordinarily would take place in the dog. The chemical attractant(s) that brings these two non-human hookworms to point of contact with us, in fact, may turn out to be similar to the one described here for Strongyloides. Penetration of skin may also be nonspecific, relying on recognition signals such as body heat. However, once inside a nonpermissive host, missing specific chemical environmental cues trigger their aberrant behavior. Ultimately, these worms die and are consumed by a granulomatous reaction involving eosinophils and IgE antibodies generated against the secretions of the migrating larvae. Work on dog and human hookworms have identified numerous nonspecific environmental cues similar in kind to those shown for S. stercoralis (e.g., heat, moisture, carbon dioxide in combination with moisture, and skin lipids; ref. 7), but in those studies, no single compound was identified such as the one described here.

Fig. 2.

Strongyloides in situ in the small intestine.

Urocanic acid leads Strongyloides to contact the skin of its host.

The neuronal arrangement of the third-stage larva of S. stercoralis has been determined (8), and a few of their functions have been described (9). Together with a systematic exploration of the complete repertoire of chemical and physical stimuli to which this worm can respond, it should be possible over the next several years to produce a detailed integrated neuronal map of this parasite that then could be extended to other related helminth pathogens, such as the hookworms. These investigations might lead to a new strategy for controlling parasitic infections, in which environmentally friendly compounds aimed at interfering with mechanisms of host-seeking are used, instead of continuing to attempt to develop new antiparasitic drugs and vaccines that eventually either become limited by the selection for drug resistance or are prohibitive in cost to administer to those most in need, as is already the case of most established vaccines. Developing effective, inexpensive compounds that interfere with the infection process based on the findings here is the “holy grail” of public health. Safer et al.'s work (4) gives hope for just such a breakthrough. If and when this finally occurs, a life-threatening multicellular infectious agent will have been successfully “caged.”