Parasite manipulation of hosts

Encouraging transmission: upstream/downstream

In the case of free-living infective stages, encounter is not necessarily maximized by manipulating hosts, but parasite larvae can home in on either the host's environment (e.g. ticks crawling to the top of a grass stalk) or on the host itself (e.g. by detecting chemical signals given off by the host). Combes discusses the details of parasite localization of hosts in space and time for some particular systems.

In the case of non-free-living infective stages (particularly in the case of trophic transmission), parasites can try to manipulate either the upstream or the downstream host's behavior. However, they only have access to their own physiology and to the physiology of the upstream host (which they are currently inhabiting); if they want to manipulate the downstream host, they have to present it with a (possibly false) signal of some kind. I think that Combes's categorization is useful, but it also blurs the boundaries a little bit: if a trematode makes a killifish behave strangely, thus increasing its chance of parasitism by a bird, who is being manipulated? The killifish or the bird?

For example: consider the case of Dicrocoelium, where the behavior of ants is modified so that they crawl up grass stalks to improve their chances of being ingested "accidentally" by a cow and thus transmitting the parasite. The upstream host (ant) is having its physiology and behavior modified to bring it into the environment of the downstream host (cow), which behaves normally. The "signal" is gravity (ants move up stalks).

Categories of host manipulation

• Change in activity (up or down): reduction in speed/distance travelled/etc. (pomacentrid reef fish) or increased activity, exploration etc. (rats with Toxoplasmosis, mice with Trichinella): increased predation

  Many examples of changes in fish behavior

  Acanthocephalans in inverts (good to test difference in activity patterns among species rather than just increase/decrease, which can be caused by pathology): amphipods, cockroaches

  Vectors can be affected: fly less (mosquitoes with filaria, Plasmodium fly less) or bite more, or change host preferences

• Conspicuous behavior:

  Height-seeking behavior: fish, ants. Side effect of pathology (e.g. hypoxia in fish)?

  Photophilia (light-seeking)

  Changes in color (loss of camouflage)

  Changes in size

• Changes in social behavior: castration, changes in mating behavior (host or parasite or compensation?), changes in dominance; do parasites drive host social behavior (group size, etc.)?

Mechanisms of host manipulation

• Organ disruption or damage:
  • Sensory organs (can increase or decrease transmission: Onchocerca volulus)
  • Gonads (castrators)
  • Central nervous system (CNS): rabies. Much CNS destruction is just nasty and not apparently adaptive: e.g. syphilis, prion disease, Parelaphostrongylus tenuis in moose. CNS pathologies, or ear pathologies, caused by trematodes may also be implicated in some dolphin strandings (they may interfere with echolocation).
  • muscles: e.g. lemmings (Dicrostonyx richardsoni) increase exploratory activity, mice with Trichinella pseudospiralis travel farther. These could also be compensatory reactions to changes in nutrient status (see below).
• Changes in nutritional or metabolic status: the host may increase activity to compensate for nutrient losses to parasites, or decreasing it because it is starving, or change metabolic rate to try to kill the parasite. Changes in nutritional status come from:
  • damage to assimilation organs (e.g. gut destruction)
  • anorexia (host response? more common in poorly-fed animals)
  • change in metabolic processes: e.g. malarial fever increases basal metabolic rates by 40% (although this example is clearly caused by host resistance, not by parasite manipulation)
• Interference with control systems (esp neuroendocrine or growth factor)
  • e.g. Gammarus, Dicrocoelium.

But is it really manipulation?

Probably the best way to answer these questions is simply to look at the changes in parasite and host fitness. In order to determine whether a change in host physiology and/or behavior deserves to be called "manipulation", we have to classify its effects on host and parasite fitness. Here is one such possible classification:

Parasite fitness	Host fitness	Explanation
+ (transmission)	-	Parasite manipulation
+ (survivorship)	0/-	Parasite site selection
- (survivorship)	+	Host behavioral resistance
- (transmission)	?	Host inclusive-fitness reactions?
0/-	-	Host pathology?
0/+	+	Host compensation

Testing hypotheses

• it's not usually possible to "turn off" host manipulation, although in cases where we know the actual biochemistry this might be possible;
• it's difficult to set up a sufficiently complete artificial environment to allow transmission in the lab, particularly for heteroxenous parasites;
• transmission tends to be sensitive to environment in a way that is hard to replicate in the lab (harder, for example, than measuring physiology of a single life stage in the lab).

Other possibilities include:

• epidemiological or observational studies in the field (see Janice Moore's starling/pillbug/acanthocephalan study, or the seal study reported in Combes, or the comparisons done by Lafferty and Morris [Combes p. 257] of predation on infected vs. uninfected fish) -- these kinds of studies give you accurate information, but not detailed information, about what's happening in natural systems
• phylogenetically controlled studies (e.g. Moore and Gotelli 1996), which measure behaviors in a range of related species to test whether they represent "adaptations" or mere phylogenetic constraints.

More loosely, there are many qualitative arguments we can use to argue that behaviors represent adaptations:

• check to make sure that the observed reaction is actually most consistent with a transmission-increasing adaptation. For example:
  • increased predation rates of parasitized hosts (assumed to be an adaptation for transmission to the next host) is not necessarily by the right host (Brassard et al 1982).
  • Parasites may select particular host organs (that have strong effects on host behavior) for reasons other than influencing host behavior. For example, parasites in host CNS tissues are often isolated from host defenses: `immunological privilege' (Dunsmore et al 1983). Another example is the lens of the eye (Szidat, 1969).
  • if a change in a control system is a parasite adaptation, then any changes in behavior should be postponed until the parasite is actually in an infective stage, ready to infect the next host in the life cycle.
• Mechanistic explanations, really nailing down the biochemical or other changes leading to behavioral change, are also powerful. For example, parasitized Gammarus have modified escape responses. Injecting Gammarus with serotonin produces similar results; octopamine can block the results. Similarly, S. mansoni leads to higher opioid levels in hamsters, although it is not clear whether this is a host response or a parasite manipulation. In some cases (growth hormone in rodents), parasites actually directly produce host hormones, which makes it pretty clear that it is a case of parasite manipulation.
• Intuitively, behavioral changes that are complex and ones that we have a hard time explaining as simple pathogenic effects are the ones that we usually ascribe to parasite manipulation. But how can we quantify these feelings?

Costs of manipulation

The costs of host manipulation may lead to kin/group selection of parasites: e.g. costs of producing host hormones. Kin/group selection must be especially strong in parasites of the central nervous system, because typically the individual parasites that are in the CNS doing the manipulation (latching on to neurons or whatever) are not successfully transmitted to the next host in the cycle: they provide an opportunity for their neighbors, but don't themselves get to hitch a ride.

Dicrocoelium dendriticum is one example of a CNS parasite where one "manipulator" benefits all the other parasites in the host. Combes gives the example of Microphallus and Macrotremata, where there are potential "free riders" both within and between species.

There is also a recent paper (S. P. Brown, Proc Roy Soc B) which gives a game-theoretic treatment of whether it's worth a parasite's effort to invest in host manipulation, on the basis of how many individuals it shares the host with and how related they are. The main prediction is that there can be a group size/relatedness threshold below which manipulation is not worth it for the parasite. Brown suggests that we look for variation in manipulation as a function of parasite number. Q:how might hormonal and "direct" control of behavior differ in the individual and group pressures they put on parasites?