The results presented in this manuscript describing transformation, sloughing and aggregation of hepatopancreatic microvilli into vermiform bodies superficially resembling gregarines was obtained over a period of 6 years in a piecemeal fashion as a series of initially independent, sideline observations made during the course of dedicated research projects on a variety of known shrimp pathogens ranging from viruses to bacteria, fungi and parasites. It was not until very recently that the connections between the independent observations were understood, allowing them to be linked together into a coherent whole. A major activity that helped us to gain understanding of the connections between our piecemeal observations was the intensive research that has been carried out since 2009 and particularly since 2011 on many hundreds of shrimp specimens studied in attempts to determine the cause of acute hepatopancreatic necrosis disease (AHPND), the most recent, serious shrimp pandemic to cause severe losses in Asian shrimp cultivation.
AHPND Outbreaks began in cultivated shrimp Penaeus (Penaeus) monodon and Penaeus (Litopenaeus) vannamei China in 2009 and thereafter spread progressively to Vietnam (2010), Malaysia (2011) and Thailand (2012), although the cause of the disease was not known at that time. Indeed, it was not until 2011 that a case definition for AHPND was first described (referred to as acute hepatopancreatic necrosis syndrome or AHPNS at the time) by D.V. Lightner from the University of Arizona at a seminar organized by the Vietnamese Department of Animal Health in Hanoi in June 2011 (unpublished). It was subsequently described in the Global Aquaculture Advocate magazine under the heading of early mortality syndrome (EMS) in the Jan/Feb issue of 2012. Later in the same year, a disease card was prepared by the Network for Aquaculture Centres in Asia Pacific (NACA) and made available at its website (www.enaca.org). Finally, the causative agent (i.e., novel isolates of Vibrio parahaemolyticus) was discovered in 2013. The unique diagnostic characteristic of the disease is massive, medial sloughing of shrimp hepatopancreatic (HP) tubule epithelial cells as a result of a currently unknown toxin(s) that originates from the causative bacteria colonizing the shrimp stomach.
Since confirmatory diagnosis of AHPND is still dependent on histological diagnosis of massive sloughing of HP cells, many hundreds of cephalothorax tissue sections of shrimp suspected of AHPND infections have been examined with a primary focus on the shrimp hepatopancreas. During the course of this work on AHPND, a variety of other hepatopancreatic pathogens have also been encountered and recorded but are not often reported. Among such anomalies there has been an increasing prevalence of vermiform bodies that superficially resemble gregarines within the lumens of hepatopancreatic (HP) tubules, at the HP-stomach-midgut junction and in the midgut of cultivated giant tiger shrimp (P. monodon) and whiteleg shrimp (P. vannamei). They sometimes occur in sufficient quantities to cause white fecal strings in a phenomenon called white feces syndrome (WFS) in pond reared shrimp from approximately 2 months of culture onwards, and they were originally described as gregarines that caused WFS. It has been estimated that Thai production losses due to WFS in 2010 were 10–15% based on decreased survival and smaller harvest sizes of shrimp from WFS ponds, and this estimate excluded the normal annual Thai losses to white spot disease.
Here we describe detailed studies of the vermiform bodies using light and electron microscopes and show that they are not independent organisms but formations consisting of aggregated, transformed microvilli (ATM) that have originated by sloughing from epithelial cells of the shrimp hepatopancreatic tubules. They then accumulate at the HP-midgut junction before being discharged within feces via the midgut.
Results
Field signs of white feces syndrome (WFS)
Gross signs of WFS in shrimp cultivation ponds included white to somewhat yellow, floating fecal strings that sometimes collected in mats or could also be found on feeding trays. Examination of shrimp from ponds exhibiting signs of WFS revealed that the dissected midgut junction and midgut were distended and filled with white to yellow-golden contents. When the contents of the gut or fecal strings were examined in squash mounts with the light microscope, they consisted of masses of vermiform bodies that superficially resembled gregarines.
A passive surveillance of WFS outbreaks in P. vannamei was carried out from 2009 to 2010 in the eastern and middle part of Thailand in 25 ponds from 13 farms to determine the relationship between WFS and vermiform bodies resembling gregarines. The results revealed that 96% (24/25) of the ponds exhibiting WFS contained shrimp specimens that presented vermiform bodies resembling gregarines. Severely affected ponds exhibited reduction in shrimp survival by 20–30 percent when compared to normal ponds. There was also a decrease in feed consumption and growth rates were reduced as revealed by average daily weight gain (ADG) for the whole crop operation of less than 0.1 g/day compared to 0.2 g/day in normal ponds. Feed conversion ratios (FCR) ranged from 1.7 to 2.5 when compared to 1.5 or less for normal ponds.
Light microscopy
Whole mounts of shrimp hepatopancreatic tissue at any life stage from post-larvae (PL) to broodstock of P. monodon and P. vannamei currently cultivated in Asia (either diseased or grossly normal) often revealed the presence of non-motile, vermiform bodies superficially resembling gregarines within the tubule lumens, the HP-midgut junction and the midgut. The average size was 39 (range 17 to 58) µm wide by 279 (50 to 517) µm long (N = 21) and was smaller but roughly proportional to the size of the tubules that contained them. They sometimes contained spherical bodies that resembled cysts. Besides these cyst-like inclusions, they had no apparent cellular or subcellular features (e.g., nuclei) and contained only what appeared to be interfolded membranes that could be visibly enhanced by staining with Rose Bengal. Smears of HP tissue from affected shrimp stained with hematoxylin and eosin made these bodies more clearly visible but still revealed no cellular or subcellular structures such as nuclei. However, examination of these two types of preparations clearly revealed why they were first referred to as gregarine-like entities (GLE).
Examination of living shrimp specimens of both P. monodon and P. vannamei from post larval to broodstock life stages often revealed the presence of GLE in small to massive numbers that varied from shrimp to shrimp specimen, even from the same pond or hatchery tank. The affected shrimp showed no gross signs of disease (including white fecal strings) resulting from their presence, even in high numbers.
HP tissue of affected shrimp fixed with Davidson's fixative and processed for normal histological examination of tissue sections stained with hematoxylin and eosin (H&E) did not give as good resolution of the GLE as did whole squash mounts or stained smears. Instead, they often appeared to be partially degraded by the preparation steps, so that their membrane contents were difficult to resolve or could not be distinguished easily from the normal chyme-like material that is often seen in the HP lumens of shrimp that are actively feeding. In some better preserved specimens, it was possible to prepare a series of photomicrographs suggestive of progressive aggregation and condensation of individual membranes into GLE that lacked recognizable cellular structures and accumulated at the center of the HP near the midgut junction to superficially resemble gregarines. The progression began as scattered membranes followed by stages of aggregation followed by condensation and accumulation at the HP center. For comparison, a photomicrograph of H&E stained tissue of cultivated P. monodon shows true gregarine trophozooites (probably Nematopsis sp. [5])clustered in the region of the HP/midgut junction. These are rarely found in cultivated shrimp in Thailand and compared to GLE, they are larger, stain more intensely and have prominent nuclei. In addition, they do not arise by a process of membrane aggregation and condensation.
Using light microscopy to examine semi-thin sections of HP tissue of affected shrimp fixed and embedded for transmission electron microscopy (TEM) and stained with toluidine blue, clearly revealed that the GLE consisted of a thin outer membrane that enclosed a complex of thicker interfolded membranes. These sometimes surrounded what was later found to be sloughed, whole B-cells. The tubules that contained the GLE also showed many epithelial cells with abnormally thin microvillar layers or denuded of microvilli. The latter showed signs of lysis. Free membrane-like structures were present in the tubule lumens in addition to the GRL.
Transmission electron microscopy (TEM)
TEM of affected shrimp tissues revealed that the GLE were surrounded by a thin, single-layer membrane that bore no resemblance to a bilaminar plasma membrane or to the complex outer layers of known protozoans, metazoans or gregarines (http://tolweb.org/Gregarina/124806; including a gregarine previously reported from Thailand. It enclosed a complex of thicker, interfolded membranes with a tubular substructure, and these occasionally surrounded whole B-cells sloughed from the HP tubule epithelium. The GLE otherwise contained no recognizable cellular contents such as nuclei, mitochondria, ribosomes, etc. The origin of the outer, single-layer membrane could not be determined, but the enclosed, interfolded membranes with a tubular substructure were found to originate from microvilli of HP tubule epithelial cells of the R and F types. The microvilli first became transformed into a partially collapsed state before they peeled off of the cells and sloughed into the tubule lumen where they aggregated to form GLE. The cells denuded of microvilli subsequently lysed, releasing their contents into the tubule lumen, often leaving a remnant containing the basal nucleus collapsed against the tubule basal membrane. Based on all the information from light microscopy to electron microscopy, these bodies were named aggregated transformed microvilli (ATM).
By TEM, the F and R cells with transformed microvilli did not show the presence of recognizable pathogens such as viruses, bacteria or parasites. However, in the E-cell region of the HP where the tubule epithelium was somewhat folded to form cript-like spaces with facing microvillar layers, minute, densely staining bodies of irregular shape and size could be seen by light microscopy and by TEM. These appeared to aggregate and be capable of passing through the microvillar layers to enter the subtending cellular cytoplasm. This association appeared to be accompanied by changes in the morphology of the microvilli, and cells that showed advanced microvillar transformation appeared to have large numbers of such inclusions. It was not clear whether they had increased in numbers by accumulation or by proliferation. They lacked subcellular structures that might indicate relationship to known viral, prokaryotic or eukaryotic organisms, and it could not be determined whether they were causal or incidental to ATM formation.
We examined the shrimp midgut only in squash mounts and in the portion of the midgut that occasionally appeared in saggital tissue sections of the whole cephalothorax region in H&E stained slides. We did not examine the midgut further with semithin sections or by TEM. However, in the H&E sections we found no evidence that ATM were formed in the midgut. Instead, they simply accumulated there as they were released from the HP.
Discussion
Before we had detailed results from electron microscopy revealing the nature of ATM origin, we were dependent on light microscopy results showing vermiform bodies that resembled gregarines. A search of the literature about gregarines at that time revealed that none of those previously reported from shrimp, including one previously reported from Thailand bore any resemblance to ATM. Specifically, they lacked motility (i.e., in squash mounts) and showed no organelles such as usually prominent nuclei. Subsequent results from electron microscopy confirmed the absence of nuclei and other ultrastructural features normally found in gregarines (e.g., complex pellicles, mitochondria, ribosomes, etc.). However, we did find a previous report by Johnson that described the occasional presence of cellophane-like aggregations of membranous material in the hepatopancreatic tubule lumens of crabs and proposed that they arose from the microvilli of the tubule epithelial cells. She did not give details of their formation and expressed the inability to explain their function or significance. She also expressed the inability to conclude whether they were the result of a normal process or of some kind of rare pathology. Our ATM from shrimp morphologically resembled the structures described by Johnson in crabs by light microscopy and especially by TEM. We could find no other published description of these entities in crustaceans.
When the occurrence of ATM is severe, it can lead to the formation of white fecal strings in shrimp, and if many shrimp in the same pond exhibit this phenomenon, it can lead to floating fecal strings that sometimes accumulate in floating mats (i.e., white feces syndrome or WFS). This usually occurs from 2 months of cultivation onwards, and it may be accompanied by high shrimp mortality. However, ATM sometimes occur together with shrimp hepatopancreatic diseases such as the AHPND, other types of vibriosis, and parasitemia with the microsporidian Enterocytozoon hepatopenaei. As a result, the cause of WFS in Vietnam was attributed also to E. hepatopenaei, but this was later shown to be very unlikely based on closer study of natural and laboratory infections of E. hepatopenaei in Thailand. Thus, it is certain that these severe cases of WFS result from massive ATM formation. However, the cause of ATM formation remains a mystery. Hopefully, understanding their nature and mode of formation will lead to more directed studies to determine their cause.
Overall, our work has shown that the formation of ATM results from transformation of microvilli followed by sloughing from the subtending cell and by subsequent lysis of that cell. These features indicate that ATM formation is a pathological process. The fact that ATM were not previously described in shrimp, despite their easy recognition in whole mounts and smears of HP tissue from living shrimp, argues that their previous occurrence was probably overlooked due to low prevalence, as previously reported by Johnson for crabs. However, they have recently increased in prevalence in Asia to the extent that they are now too prevalent to be overlooked.
It may be significant that the increase in prevalence of ATM has been coincidental with the increase in prevalence of AHPND outbreaks. Although this might suggest a possible causal association, there has also been a coincidental increase in prevalence of the hepatopancreatic microsporidian Enterocytozoon hepatopenaei with AHPND, and we now know that this is certainly not a causal relationship, since AHPND is caused by unrelated bacteria. So at least for E. hepatopenaei and AHPND bacteria, it seems likely that some of their increased prevalence has resulted from contamination of broostock and/or post-larvae (PL). This contention is supported by anecdotal evidence from widely separated Thai shrimp farmers who received portions of single batches of PL derived from SPF shrimp stocks but then experienced AHPND outbreaks more-or-less simultaneously. It is also supported by our discovery of endemic E. hepatopenaei in locally held broodstock and PL of SPF P. vannamei stocks that originated from the Americas where this microsporidian has never previously been reported.
Altogether the previous information suggests that the biosecurity measures in at least some shrimp hatcheries have not been sufficiently rigorous to exclude contamination by imported and/or local pathogens. Therefore, we must consider two possibilities with respect to ATM. Either they are caused by a new agent that has contaminated SPF stocks in a similar manner to AHPND and E. hepatopenaei, or that they constitute an alternate manifestation of an existing pathogen. For example, it has been reported that AHPND bacteria produce a potent toxin that can cause sloughing of hepatopancreatic tubule epithelial cells, and it may be asked whether the same toxin at low doses may cause the formation of ATM in the absence of cell sloughing. To test this latter possibility, the laboratory infection model recently described may be used with diluted toxin preparations from the causative bacteria. With respect to the existence of a new pathogen, comparative metagenomic analysis of shrimp with and without ATM may be the most useful.
Alternatively, the possible involvement of the minute, electron dense bodies described here to be associated with microvillar transformation may be further investigated. For example, it may be possible to separate them physically from tissue homogenates by differential centrifugation and/or filtration for further analysis and for laboratory challenge tests.
In conclusion, we have revealed by TEM that vermiform structures superficially resembling gregarines and commonly found now in the HP of Asian cultivated shrimp are not independent organisms but result from the transformation, sloughing and aggregation of microvilli from the HP tubule epithelial cells themselves. The denuded epithelial cells subsequently undergo lysis, indicating that the process has the potential for negative impact on shrimp growth and survival, and in very severe cases can lead to the phenomenon called white feces syndrome (WFS). Further investigation is needed to understand the cause of ATM and evaluate their impact on shrimp production.
July 2014
