Just before twilight, a herd of large sauropods grazed sedately on a stand of kauri trees. With their long necks, the mature adults could easily reach up into the understory and pull down whole branches laden with succulent new foliage. The forest was filled with the sounds of eating: the thrashing of large animals through the underbrush, the cracking and rending of branches torn from the trunks, and the crashing thumps of these giants. Other, more subtle noises emanated from the herd, the rumbling from stomachs as the consumed vegetation was ground by gizzard stones and fermented in ballooning bellies, and some contented vocalization between members of the herd. The air was scented by the pungent odor of resin oozing from the damaged araucari-ans. Suddenly masses of biting flies rising from their resting places on the bark of the trees engulfed the herd. Included among them were female sand flies looking for a source of blood to ensure egg development.
After settling on a sauropod, one female fed repeatedly, acquiring not only blood for her eggs but also something much more significant—pathogenic microorganisms. The insect was interrupted in the middle of what was to be its final meal; perhaps a therapod appeared and the startled sauropod responded by scraping against some vegetation, forcing the sand fly to quickly flee. Her escape route led to the trunk of a resin-coated araucarian tree. Once in the sticky trap, she couldn't pull free. The fly's straining, desperate movements attracted the attention of a small predator patrolling the bark, who nipped open a miniscule hole in the end of her abdomen, deftly pulled out the reproductive system, and devoured the protein-rich eggs. Some of the gut contents of the entrapped insect spilled out onto the fresh resin as life ebbed away. She lay on one side in a drop of spilled blood, disemboweled, head and mouthparts clearly visible, wings outstretched—a sudden and unexpected victim of double jeopardy. An additional resin flow entombed the small female fly (color plate 8C).
Long before our search for Cretaceous diseases began, events as described above took place in a Burmese forest 100 million years ago. The last few seconds in the life of this fly were extremely important, for they would set the stage for the very first discovery of a Cretaceous insect-vectored disease.
One of many theories presented on the demise of the dinosaurs states that they succumbed to diseases.25 While intriguing, there was no evidence that terrestrial vertebrate pathogens existed in the Mesozoic—until recently. This is the story of how they were discovered.
We have been interested in the history of infectious organisms for decades. Conversation after conversation has centered on how to find pathogens in the fossil record, when and how did a disease arise, and where did the ancestor of these organisms first appear. Our belief has always been that amber represents the most ideal place to look for answers because of its remarkable preservative qualities. Delicate structures such as bacteria, pollen, and even cells with nuclei and mitochondria have been preserved. So if one believes that agents of disease can indeed be preserved in amber, then what kind of pathogens should one expect to find?
The answer is an arthropod-borne pathogen because amber is the best-known repository of fossil arthropods. So, insects in amber became the starting point in our search for Cretaceous diseases. This quest is where the role of a scientist becomes much like that of a detective: searching for suspects, looking for clues, applying deductive reasoning, and hoping for a bit of luck.
In the preliminary investigation, a scientist would screen arthro pods that are already recognized as vectors of disease. For example, we know that mosquitoes carry malaria, yellow fever, and West Nile virus, tick bites are responsible for Lyme disease, and blackflies transmit nematodes causing river blindness. To increase the chances of finding a pathogen, the vectors one examines should be abundant, the organism must be detectable with the light microscope, and the infection should be confined to a specific location in the arthropod. Applying these parameters started the game.
Some 100 million years after becoming entrapped in resin, a remarkable treasure found its way to our laboratory, one of several biting insects recovered in Burmese amber that met our criteria for possibly containing pathogens. When the piece arrived, I carefully placed the specimen, which was a sand fly, on a microscope slide and began peering through the eyepieces, hoping but not really optimistic. Such painstaking and frustrating work— looking for evidence of disease in minute fossil bloodsucking insects! What were the chances of finding anything like a microscopic pathogen inside an insect less than a millimeter in length? Searching for something so small, something measured in microns, in a specimen unprepared by conventional fixation or un-enhanced by diagnostic stains was daunting, tedious work. Then you still had to consider whether the microorganism would even be recognizable.
The odds against such a discovery appear astronomical. First, a sand fly must have fed on an infected vertebrate, ingested pathogens along with blood, and then end up being preserved in amber. Furthermore, this particular piece would have to be one of the few dug out of the ground, polished, and selected among all others to be shipped out of Burma. And finally, after passing through the hands of several dealers and collectors, the specimen would eventually have to end up under my microscope.
To make matters even more complicated, the very act of examining a fossil fly for internal pathogens is beset with difficulties. To begin with, the fly has to be fairly transparent in order to examine inside the body cavity. In most cases, the integument re mains opaque or only translucent, making it impossible to obtain a clear view of the internal structures. Even if the specimen is reasonably transparent, small organisms like pathogens can stick together, appearing as dark globs, or simply be masked by the similarly colored background. The ability to distinguish pathogens in amber is only possible if they were particularly well fixed by naturally occurring chemicals in the resin, which is a rare event.252 253
That first hopeful glance through the microscope revealed that the abdomen was partially filled with a dark substance—could that really be blood? I thought maybe, just maybe. Switching to medium power, the surprisingly clear head region sprang into view, exposing the proboscis. In the middle of this feeding tube there was a spherical object, only 4 microns in diameter, with a dark body inside. At still greater magnification, the specimen revealed several more spherical objects, one in the center and others toward the edge of the proboscis (fig. 34). What were they? I began snapping photos. Astonishingly, the pictures revealed a short, filiform structure emerging from one of the spherical ob-jects—a flagellum! This was exciting evidence that we had found a protozoan and it could be a pathogen.
The next day was spent reexamining the specimen. Closer inspection told us the piece had to be repolished in order to see the structures better. This procedure is beset with danger because the risk of damaging the specimen increases with every polishing. The evidence already showed that this fly was extremely valuable scientifically. After several nerve-racking hours spent tediously hand polishing that small gem, we could now focus attention on the contents of the abdomen that did in fact contain a partially digested blood meal. The fortuitous removal of the ovaries by the predator enabled us to examine the midgut, which was packed with numerous long coiled bodies orientated in a myriad of positions. They were so twisted and intertwined that it was difficult to find one that was not covered by at least a portion of another, but they were definitely flagellated protozoa (fig. 35). Inside their bodies could be seen the large dark nucleus and
smaller kinetoplast. Ancient trypanosomatids seen for the first time by man—the first evidence of vector-borne diseases. An amazing discovery!
Of course, these findings had to be validated by making comparisons with the various developmental stages of present-day trypanosomatids in sand flies. After reading numerous scientific papers, it became obvious that the organisms most closely related to the fossils were a group of trypanosomatids known as Leishmania. Could Leishmania have existed 100 mya? When present-day sand flies feed on a vertebrate infected with Leishmania, they acquire small, round, non-flagellated stages called amastig-otes, which are formed in the tissues of the vertebrate. These amazingly resemble the round bodies first seen in the proboscis of the fossil. In the insect, the amastigotes then elongate and develop flagella, which provide them with mobility. These stages, the promastigotes, multiply by simple division and become quite numerous within the insect's alimentary tract. Many types occur, including oval forms called procyclics (fig. 36). Eventually some of the promastigotes become elongated and convert to nectomo-
nad stages. After a period of proliferation, the final paramastig-ote stage is formed. These short, stubby flagellated cells migrate to the head of the fly and are transferred to a vertebrate during the insect's next meal. Amastigotes, nectomonads, and para-mastigotes were all seen in the fossil sand fly254 (figs. 24, 34).
Convincing photographic prints of the various stages were needed for publication, so into the darkroom we went. Night after night, intensive work finally produced a set of prints that demonstrated all of the characters needed to align these microorganisms with the leishmanial pathogens of vertebrates.
Now we had to obtain the next bit of evidence in our case for insect-borne disease in the Cretaceous. With the presence of vertebrate Leishmania parasites established, how can the host be determined? Two genera of Old World sand flies carry Leishmania to mammals and reptiles today. Those in the genus Phlebotomus prefer mammals and differ morphologically from members of Sergentomyia that dine on reptiles. So we turned to the sand fly for clues that would indicate the identity of the vertebrate host. Described in an extinct genus because it differed from all extant types,255 the fossil nevertheless possessed some characters found in extant species of Sergentomyia. One was the position of the hairs on the tips of the abdominal segments, which are usually erect in Phlebotomus and decumbent in Sergentomyia. Although only one of these remained on the fossil (many hairs and scales of insects in amber are lost as the trapped arthropods attempt to escape), that lone hair was decumbent. Shaky evidence, but luckily, even if hairs weren't found, their sockets were still present and decumbent hairs have elongate sockets, which was the condition of the sockets in the specimen.
Another identifying character dealt with the wing venation, and this fly's venation definitely aligned it with the reptile feeders. This was all very interesting and while showing that the sand fly possessed characters of present-day reptile feeders, there still wasn't enough evidence to conclude that she fed on Cretaceous reptiles. Perhaps, just by some wild chance, there were some blood cells remaining from her last meal. Vertebrate blood cells are normally broken down within four to five hours after feeding,142 but we had already concluded from the blood volume remaining in the abdomen of the fossil that only the early stages of blood digestion had begun before death,255 so some of the vertebrate blood cells might still be intact.
We turned back to the fossil for more answers. Where else could one expect vertebrate blood cells to occur? The abdominal midgut had already been thoroughly examined when the various stages of the flagellates were photographed. That left only the thoracic midgut, a region that is usually obscured by thick cuticle and dense flight muscles. And as expected, this area appeared to be completely opaque. A seemingly hopeless endeavor, but still worth a try. After several hours of orientating the fly in different planes, a miniscule spot of light suddenly opened a view into the thoracic midgut lumen. Illuminated in this small expanse was a group of nucleated vertebrate blood cells. Another breathtaking find!
Taking and processing several rolls of film eventually led to satisfactory prints of these enigmatic nucleated cells, and the process of identification began. Measurements were taken and the literature on vertebrate blood cell size and morphology was consulted. The nucleated blood cells in the fossil midgut lumen were oval, ranging from 10 to 15 |im in length. Mammals were eliminated as a source because most of their blood cells are discoidal and enucleated. However, oval nucleated blood cells occur in birds, amphibians, and reptiles.256 How could we narrow the search further? Amphibians have large blood cells, ranging from 18 to 67 |im in greatest dimension (most are over 26 |im), and since there are also no cases of Leishmania or sand fly vectored trypanosomatids in amphibians today,183 we could strike them off the list of suspects, leaving birds and reptiles. Both reptiles and birds have blood cells overlapping in size with those in the fossil.256 However, there are no present-day cases of Leishmania or other sand fly-transmitted trypanosomatid infections in birds,257 which eliminated them from further consideration and left only one group, the reptiles.
We know that there are numerous recorded cases of try-panosomatid infections in lizards and snakes, all of which are known or suspected of being transmitted by sand flies.183258259 Also, one of the complete nucleated blood cells in the fossil insect is almost identical in shape and size to a lizard proerythro-cyte.183 All of these facts taken together, from the characters of the vector fly to the morphology and size of the blood cells, led to the verdict that the fossil sand fly had been feeding on a reptile.
In Leishmania infections, the amastigotes are produced in blood cells of vertebrates, where they appear as small, dark areas inside lighter areas known as parasitophorous vacuoles. The most astonishing thing is that some of the fossil blood cells actually contained these vacuoles with developing amastigotes inside (fig. 37). They bore a striking resemblance to developing Sauro-leishmania amastigotes in the blood cells of present-day lizards.183 260 So here were infected reptilian blood cells inside this little sand fly. The chances of finding evidence of such pathogens among the fossil remains of any vertebrate would be miniscule. That is why amber is a treasure trove with so many secrets from the past.
What an intricate case of detective work this turned out to be. The implications of this find did not fully sink in until we began to wonder what types of reptiles might have provided the meal for our incredible sand fly. What were the dominant reptiles, or should we say the dominant vertebrates at that time—none other than dinosaurs! Could this fossil sand fly have been feeding on a dinosaur? Why not? Certainly they were widespread and since reptile-seeking sand flies today prey on a variety of lizards and snakes, why wouldn't Cretaceous forms dine on dinosaurs? Suddenly those cells took on a new light: they could be the first infected dinosaur blood cells seen by man. And much more significant, might they contain agents that caused, or at least contributed to, the demise of the dinosaurs? We can't say for sure, but we know that sand flies were widespread in the Cretaceous, probably with a global distribution. With many of the continents connected, pandemics could easily have occurred. This one fos-
sil, with its immense wealth of scientific information, opened up a large view to the past, one that revealed the stark reality of infections, death, and possible extinctions.
While finding a single infected vector establishes the presence of a pathogen at a particular place and time, this alone cannot indicate the prevalence of infection. That is why we were quite excited to discover additional infected sand flies in Burmese amber obtained over the next few months. In fact, out of 21 female sand flies that have been examined, 10 were found with trypanoso-matids. This discovery confirmed our original suspicion that the incidence of reptilian leishmaniasis was extremely high in that location 100 mya. The chances of finding an infected present-day sand fly in regions with natural leishmanial infections are very low. What would then be the incidence of infection based on the sand flies found so far in Burmese amber? Incredibly high! Perhaps most or all of the resident sand flies were feeding on infected vertebrates, indicating an epidemic among the vertebrate hosts, possibly including dinosaurs.
If one disease agent could be found, what about others? Our interest was whetted and we began to investigate biting midges as possible vectors of vertebrate pathogens. Our perseverance was rewarded when we discovered an ancient malarial organism in a Burmese amber ceratopogonid (color plate 9A). Inside the body cavity of the fossil fly were developing oocysts (fig. 38) and sporozoites of a Haemoproteus-like pathogen, all completely preserved (fig. 23). Again using morphological characters of the biting midge, we determined that it probably fed on a large vertebrate, and the predominant large vertebrates at that time were dinosaurs.261
While protozoa are visible with the light microscope, viruses are not. But maybe we could find some insect-pathogenic viruses with a marker visible with the light microscope, like proteina-ceous polyhedral bodies that encapsulate the virions. They would establish the presence of certain virus groups. Again, persistence paid off. The midgut of a biting midge in Burmese amber contained several hundred polyhedra (color plates 9C, 9D) of
Figure 38. Developing oocysts of Paleohaemoproteus burmacis in a Burmese amber biting midge.167
a cytoplasmic polyhedrosis virus.172 As occurs in this type of present-day virus infections, the polyhedra were localized in the midgut of the host. Cytoplasmic polyhedrosis viruses are members of the viral family Reoviridae, which also includes some vertebrate pathogenic arboviruses. So while arboviruses have not been recovered from fossil arthropods, the polyhedra of insect parasitic viruses that could be the precursors of some types of arboviruses have been found. Today, cytoplasmic polyhedro-sis viruses infect biting midges, mosquitoes, and sand flies, all of which also transmit arboviruses to vertebrates. The viral-infected Burmese amber biting midge in all likelihood, based on various morphological features, fed on vertebrates.172
Looking at additional amber fossils has told us something about other insect diseases in the Cretaceous. We have found a putative nuclear polyhedrosis virus in a sand fly, a fungal pathogen of a scale insect (color plate 13D), mite parasites attached to insects (color plates 13B, 14C), and a nematode parasite of a biting midge (color plate 13C).337 Finding a pathogenic fossil organism only tells us that it was present at a particular point in time. How long it existed before that can only be determined when older fossils are discovered. Logic tells us that many diseases were well established long before our samples appeared. Just how long remains to be seen.
Was this article helpful?