Of caterpillars and horses: recent scientific progress on the cause of Mare Reproductive Loss Syndrome

An abortigenic disease, now known as Mare Reproductive Loss Syndrome (MRLS), significantly impacted the horse industry in the central Ohio Valley in late April and early May, 2001 and 2002. In 2001, approximately 25% of all pregnant mares aborted within several weeks (over 3,000 mares lost pregnancies), and abortion rates exceeded 60% on some farms. MRLS struck hard and without warning, was caused by something in the environment, was not transmittable between animals, and was not associated with any known abortigenic agent or disease.

It is a newly recognized disease, and it cost the state of Kentucky ~$330 million in 2001 alone (Thalheimer and Lawrence, 2001). Initial research in the summer and fall of 2001 was targeted toward specific agents, such as mycotoxins (Newman, 2003), fescue toxicosis (Schultz and Bush, 2003), cyanide (Harkins et al., 2003), mandelonitrile (Camargo et al., 2002), weather (Priddy et al., 2003), ionic imbalance and poison hemlock (Powell, 2002), among others. To date, none of these factors has been demonstrated to reproduce MRLS. A comprehensive review of the history of MRLS and the initial experiments is provided in Powell et al., 2003.

An epidemiologic survey conducted by Drs. Roberta Dwyer, David Powell and co-workers revealed a temporal correlation between MRLS and presence of eastern tent caterpillars (ETC; Malacosoma americanum) on horse farms (Dwyer et al., 2003).

A statistical correlation does not necessarily mean that the caterpillars caused the abortions, so several groups of scientists from around the country designed experiments to determine the role ETC play in MRLS.

The experiments summarized here are the results of collaborative efforts by Dr. Karen McDowell (Veterinary Science, University of Kentucky), Dr. Bruce Webb (Entomology, University of Kentucky), Dr. Neil Williams and Dr. Mike Donahue (Livestock Disease Diagnostic Center, University of Kentucky), and Dr. Kyle Newman (Venture Laboratories, Lexington, Kentucky). The studies have followed a deductive approach, with the results of each experiment defining the questions posed in the next experiment. Details of the first four experiments can be found in Webb et al. (2004). For all of these experiments, pregnant mares were housed communally on grass paddocks and supplemented with hay as needed, with water available ad libitum.

Pregnancies were monitored by manual palpation and real-time ultrasonography weekly during the first experiment, and daily in all subsequent experiments, beginning prior to the first day of treatment and continuing for at least 30 days. Aborted fetuses and placentas, as well as blood samples from the mares, were taken to the Livestock Disease Diagnostic Center (LDDC) for necropsy, histopathology, bacteriology, serology, and virology.


Eastern tent caterpillars cause mare abortions

In April and May, 2002 (the first opportunity to perform experiments with ETC after the 2001 MRLS outbreak), two experiments were conducted to determine if ETC could actually cause fetal loss in pregnant mares. For 20 days (first experiment) or for 10 days (second experiment), for 6 hrs each day, pregnant mares were placed on grass pasture in 16 x 16 ft pens, and ETC were released into the grass in the pens. Each day all pens were moved to a new area of pasture so that the mares would have fresh pasture to graze while in the pens.

In the first experiment, 18 of 29 mares exposed to ETC and/or nest materials lost their pregnancies. In the second experiment, three of seven mares exposed to ETC lost their pregnancies (P<0.05), while none of six mares exposed to ETC frass (excrement) and none of the seven control mares (not exposed to ETC, ETC frass or nest materials) aborted. Signs of abortion were consistent with MRLS, e.g. few to no premonitory signs of impending abortion in the mares, increased echogenicity of fetal fluids prior to fetal death and abortion, and presence of actinobacilli or non-beta-hemolytic streptococci in fetal fluids and fetal tissues. In the second experiment, blood samples were collected from the mares daily for hematology, and every third day for clinical chemistry and serum progesterone and estradiol concentrations. None of these measurements were good predictors of which mares were affected by MRLS. Serum concentrations of both hormones declined associated with abortion, but the decline was after changes in fetal fluid echogenicity were apparent. Results of the first experiment were released to the public on April 26, 2002, before MRLS abortions were reported in the area, and in sufficient time for farms to take steps to reduce exposure of pregnant mares to ETC.


Effects of eliminating pasture grass, and freezing or autoclaving ETC

In the next experiment (August 2002), the effect of pasture was removed by mixing ETC treatments with sweetfeed, and feeding the mares from feed buckets.

Additionally, the effects on abortigenic activity of freezing larvae or exposing them to high heat by autoclaving were tested. Finally, whether another hairy caterpillar, gypsy moth caterpillar (GMC, Lymantria dispar), could cause abortions in pregnant mares was evaluated. Pregnant mares, ~40-120 days gestation, were fed ETC larvae that had been collected during April and May, 2002 and stored frozen (n= 5 mares), frozen ETC larvae that had been autoclaved and refrozen (n=5 mares), or frozen GMC (n=4 mares) (50 g insect/mare/day for 8 days). Blood samples were collected daily from each mare.

Three mares fed frozen ETC aborted (P<0.05) with signs consistent with MRLS, and none of the mares fed autoclaved ETC aborted. One mare fed frozen GMC aborted. Signs of abortion in the ETC-fed mares were consistent with MRLS, but the abortion in the GMC-fed mare was not. The fetus from the GMCfed mare did not grow at a normal rate and there was not an increase in fetal fluid echogenicity prior to abortion, as is typical of MRLS. This experiment demonstrated that the abortigenic activity of ETC was stable to freezing and was destroyed by autoclaving. Whether the abortion in the mare fed GMC was due to treatment or would have occurred regardless of treatment cannot be determined from this experiment.


The abortigenic activity is present in the caterpillar exoskeleton, and cannot be extracted with aqueous or organic solutions

Two experiments, performed in May and June, 2003, were designed to determine which physical portion of the insect was abortigenic, and whether the activity could be extracted with aqueous or organic solutions.

ETC larvae were collected in central Kentucky in April and May of 2003 and stored frozen at -80oC.

Pregnant mares (~45-100 days gestation) were fed individually from feed buckets, with treatments (representing 50 g ETC larvae/mare/day for 10 days) mixed in sweetfeed.

In the first experiment, 35 mares were randomly divided into seven treatment groups of five mares each. For the first two treatments, mares were fed ETC and served as positive controls, or fed saline and served as negative controls. For the next three treatments, mares were fed ETC that had been carefully dissected into three portions, the exoskeleton, including setae (caterpillar hairs), the gut, or the remainder of the internal insect tissues.

Each of the three portions of the insect constituted a separate treatment fed to mares. For the final two treatments, ETC were homogenized in saline and then separated by size (larger than 0.45 microns or smaller than 0.45 microns), and mares were fed either the larger sized fraction or the smaller sized fraction.

Fetal losses occurred in all five mares fed ETC, in three mares fed ETC exoskeleton (P<0.05), and in one mare fed the larger sized fraction of the homogenized insects. No losses occurred in the negative control (saline) group, in mares fed other ETC tissues (gut or internal tissues), or in mares fed the smaller sized fraction of the homogenized insects.

Increased echogenicity of fetal fluids prior to fetal death and bacteriologic findings in fetal tissues (nonbeta- hemolytic streptococci or actinobacilli) were consistent with MRLS as the syndrome is recognized in the field. This experiment demonstrated that the abortigenic activity was associated with the exoskeleton of the insect and that it could not be extracted with an aqueous solution.

The second experiment was designed to further fractionate the ETC exoskeleton, in order to learn more about the physical and/or chemical nature of the abortigenic activity. Twenty-five pregnant mares were randomly divided into five treatment groups of five mares each. For the first three treatments, mares were fed dissected ETC exoskeleton, (positive controls), corn oil (negative controls), or ETC exoskeleton that had been crushed with a mortar and pestle to disrupt its physical integrity. For the remaining two treatments, lipids were extracted from the ETC exoskeleton with an organic solvent, methylene chloride. The methylene chloride was then evaporated through corn oil, transferring the extracted lipids to the corn oil. The two treatments consisted of the corn oil containing the extracted lipids or the exoskeleton after it had gone through the organic extraction process.

Fetal loss occurred in three mares fed ETC exoskeleton (positive control) (P<0.05), in one mare fed powdered exoskeleton, and in two mares fed exoskeleton after the organic extraction. Again, increased echogenicity of fetal fluids prior to fetal death and presence of non-beta-hemolytic streptococci or actinobacilli in fetal tissues and fluids indicated that the losses were due to MRLS.

Taken together, these two experiments demonstrate that the abortigenic activity of ETC is associated only with the exoskeleton of the caterpillar, it is reduced when the physical integrity of the exoskeleton is disrupted, and it is not easily extracted with either aqueous or organic solutions.


ETC causes abortions in pregnant gilts in a manner consistent with MRLS

Natural occurrence of MRLS and controlled experiments have been reported in horses, and thus far it was not known if ETC could cause pregnancy loss in any non-equid animal. In this experiment, pigs were chosen because pigs and horses share similarities in fetal development and placentation, both are monogastrics, and pigs have a shorter gestational period, ~112-115 days compared with ~340 days for horses. Additionally, if MRLS could be induced in a litter bearing species, then perhaps different fetuses would be affected to different degrees, and thereby provide valuable information on the progression of the disease.

ETC larvae were collected in central Kentucky in April and May of 2003 and stored frozen at -80oC.

Ten pregnant gilts, at 54-56 days of gestation at the beginning of the project, were randomly assigned to receive ETC (40 g ETC larvae/gilt/day for 10 days, mixed with their normal gestational ration; n=5 gilts) or to serve as controls (normal gestational ration only; n=5 gilts). Gilts were then randomly paired, with one ETC-fed gilt and one control gilt in each pair.

Two of five gilts fed ETC aborted their entire litters (P=0.4444). Those two gilts and their paired controls were euthanized the day following fetal loss. The remaining three ETC-fed gilts and their control pairs were still pregnant at the time of euthanasia, 29 days after the first ETC treatment. Of particular interest was the identification of caterpillar setae (hairs) embedded in the alimentary tract of all gilts fed ETC but in none of the control gilts (P=0.0079). The setae were surrounded by eosinophilic microgranulomatous lesions and were randomly scattered throughout the alimentary tract of each gilt. Caterpillar setae were not found in other gilt tissues examined, and no other significant macroscopic or microscopic changes or disease processes were identified in the gilts or the fetuses. This study demonstrated that domestic pigs are a good model for the study of MRLS. It also provided the first indication that ETC can cause fetal death and abortion in a non-equid species, and provided the first indication that ETC cause physical and potentially harmful lesions in the digestive tract of mammals. We have since confirmed that ETC cause similar lesions in the alimentary tract of mares.


Where do we go from here?

The evidence is overwhelming that ETC cause MRLS. At present, the only known way to prevent the disease is to avoid contact between pregnant mares and ETC.

On some farms that is an adequate solution; on others it is not. There are many unanswered questions, from both practical and scientific perspectives. Some of these are:

* What is the minimum level of exposure necessary to cause abortions (how many caterpillars does it take and for how long)?
 * When several mares are exposed to the same levels of ETC, why do some abort while some do not (what makes one mare more/less susceptible than the next)?
* What role does the mare’s immune system play?
* Does the abortigenic factor/ agent persist in the environment after the several weeks in the spring when ETC are present?
* Can other caterpillars/insects cause the disease?
* Can mares be treated before exposure to prevent the disease, by immunization, by feed supplements, or by some other means?
* Can mares be effectively treated after exposure?

The answers to these questions, and the new knowledge gained in the process of asking, will certainly further our understanding and management of MRLS. Such studies are also likely to provide new insights into the of other fetal/placental diseases, and into the physiology of normal pregnancy in horses.


References