3.1. Specificity and sensitivity of the different PCR primers and protocols
PCR results are summarized in the tables below comparing the specificity and sensitivity of different primers used for WzSV8, DHPV, RLB/NHPB, Spiroplasma, Microsporidia, IHHNV, EHP, YHV, TSV, MrNV/XSV, and DIV1 for the 3 regions sampled.
Both WzSV8 primers 170 F/R and PvSV (Lanza, 2022 personal communication and Cruz Flores et al. 2022) were more sensitive (77% and 84%, respectively) than the primer designed by BIOTEC (Thailand) (Srisala et al. 2023), with 64% prevalence (Table 3). These primers have been developed by American research teams using local isolates. We also unsuccessfully tested all 9 primers listed by Liu and coauthors (Liu et al. 2021) in 28 samples from region 1. After repeated negative results, we stopped using these primers.
Table 4 shows a comparison of 4 different primers used to detect DHPV plus MHBV. The nested method used by Phromjai et al. (2002) and Umesha et al. (2006) and a later modified version (Phromjai et al. 2002, Srisala et al. 2021) were more sensitive (14 and 12%, respectively) than other methods, including the seminested method reported as universal DHPV (5% prevalence) (Srisala et al. 2022) and the 1120F/Rfrom Pantoja & Lightner (2000) with 3% prevalence. MHBV was always negative and was included because of a previous report that its intranuclear inclusion bodies are somewhat similar to those of DHPV types (Gangnonngiw et al. 2022), while the type of DHPV reported from M. rosenbergii does not produce intranuclear inclusions.
Interestingly, the detection of WzSV8 and DHPV in wild broodstock in region 1 showed different patterns. On the one hand, WzSV8 170 F/R was less sensitive (17% prevalence) than 428F/R-168F/R from Srisala et al. (2023) (42% prevalence) and PvSV (Cruz-Flores et al. 2022) (75% prevalence). On the other hand, the modified version of DHPV of Phromjai et al. (2002) and Srisala et al. (2021) was more sensitive (33% vs 8%) than the primers adapted by Phromjai et al. (2002) and Umesha et al. (2006).
In the present study, Rickettsia-like bacteria were the most common type of intracellular bacteria detected, as RLB (Nunan et al. 2003a, b) or detected with a universal primer (Potts et al. 2020) (28% and 25%, respectively), with a significant prevalence in all three regions. The prevalence of NHPB was very low (1% overall) and was found in only one sample in region 3 (Table 5).
Nunan et al. (2004) were the first to report Spiroplasma as a pathogen to P. vannamei shrimp from samples with severe mortalities at a Colombian shrimp farm. Nunan et al. (2004) named this species S. penaei and designed a set of primers for its detection. Table 6 shows the prevalence of this pathogen as well as two other different primers reported for Spiroplasma-specific 16S rDNA, namely, F28/R5 (Bastian et al. 2004) and Pri 1/2 (Ding et al. 2007). Both primers have been reported to detect S. mirium as well as S.eriocheiris (Ding et al. 2007, Liang et al. 2010). Interestingly, in addition to S. penaei (24% prevalence), Pri 1/2 primers were also detected (10% prevalence) in regions 1 and 3. Ding et al. (2007) developed this set for detecting Spiroplasma in the Chinese mitten crab Eriocheir sinensis and the crayfish Procambarus clarkii. They found that both Pri 1/2 and F28/R5 detected it in some samples, a result that differs from those from the present study.
For microsporidia in the present study, we tested two sets of universal primers, 18f and 1492r, used by Sokolova et al. (2015) and (Vossbrinck. et al., 2004), which amplifies most of the small subunit rRNA of microsporidia. We included Pasharawipas et al. (1994) primers that are selective for A. penaei (Table 7). A. penaei primers were more sensitive than both universals tested (50% vs 36% and 19%). Microsporidia were detected in farms from three regions. Farm samples from region 3 had the highest prevalence, with 94% of the samples positive.
Table 8 shows a comparison of 4 primers commonly used for IHHNV (309 F/R, 389 F/R. 392 F/R and 77012F/77353R). IHHNV was only found in farm samples of region 1, and within these samples, only 15% could be considered as IHHNV complete virus, whereas 23% of the samples were EVEs positive. Both 392F/R and 389F/R were the most common primers, followed by 309F/R and 77012F/77353R.
Table 9 shows the results from 6 different primers used for detecting EHP. In general, most of the existing PCR detection methods target the EHP small subunit ribosomal RNA (SSU rRNA) gene (SSU-PCR) (Tourtip et al. 2009,Tangprasittipap et al. 2013, Tang et al. 2015,Tang et al. 2017). However, Jaroenlak et al. (2016) discovered that they can give false positive test results due to cross reactivity of the SSU-PCR primers with DNA from closely related microsporidia. To overcome this problem, a nested PCR method was developed for detection of the spore wall protein (SWP) gene of EHP (Jaroenlak et al. 2016). Only one of our samples from a farm in region 1 gave positive results using the primers targeting the SWP gene, equivalent to an average of 2% prevalence in this region and 1% overall. The other nested method (Tangprasittipap et al. 2013) gave a 2% prevalence, both in only one sample each from farms in regions 1 and 3. The method of Tang (Tang et al. 2015) gave 3% for overall prevalence from one sample each from different farm samples in the 3 regions. We did not find any positive reactors using the primers of Tang et al. (2017). The most abundant prevalence was found using Tourtip et al. (2009) primers with 10% overall prevalence (one sample positive in each of regions 1 and 2 but 10 in 18 samples positive from region 3).
The results for DIV1, YHV, TSV, YHV-GAV, PvNV, IMNV, MrNV and XSV were not included because they were very rare or absent.
3.2.PCR results by region
In general, WzSV8 was the dominant agent in all regions (89%) and was the only one found in all samples independent of region or size (Table 10). It was followed by Propionigenium and microsporidia (75% and 56%, respectively). However, Propionigenium is not considered a pathogen but was used in this study as a bio-indicator (Table 11). The prevalence of this anaerobic bacterium was 100% in both broodstock and wild animals from region 1. The only group of animals that had no positive test reaction were post larvae from region 2. The highest prevalence of microsporidia was in farm animals (94%) in region 3.
In Region 1, WzSV8, Propionigenium and microsporidia were present in all the samples (88%, 83% and 49%, respectively, as averages of all sites), and their presence in all samples, including wild animals, would suggest that these 3 microorganisms are endemic in P. vannamei. CMNV was present in all except wild animals with an average prevalence of 24%, being more abundant in postlarvae samples (50%). Spiroplasma was found only in the hatchery- and farm-level samples (29% and 38%, respectively). RLB, Vibrio and AHPND were found in only farm and broodstock samples (37%, 31% and 13%, respectively, as an average in two sites in region 1. DHPV and WSSV were found in farm animals (19% and 23%, respectively), DHPV (33%) in wild animals and WSSV (13%) in broodstock.
The IHHNV infectious form and its endogenous nonviral form were only found in farm animals (15% and 23%, respectively). By using the primers that target the EHP SWP gene as the most selective primer set, only 1 of 42 samples was positive (Table 9).
In Region 2, WzSV8 and RLB were present in both larvae (only one sample) and farm samples (100% and 29%, respectively). Spiroplasma was detected only in a single sample of larvae. For farm animals, the detection prevalence was 53%, 47%, 18%, 12% and 6%for Propionigenium, non-EHP-microsporidia, DHPV, Vibrio and IMNV, respectively (only found in one of 17 samples).
In Region 3, only animals from farms were received. The major microorganisms detected were microsporidia, WzSV8, RLB, Propionigenium and Spiroplasma at relatively high prevalences of 94%, 83%, 78%, 78% and 44%, respectively, while DHPV, DIV1 (5 of 18 animals were found positive), Vibrio, WSSV and EHP were detected at much lower prevalences of 28%, 28%, 22%, 6% and 6%, respectively (Tables 9 and 10). The high prevalence of EHP (56%) (table 9) using the Tourtip method (Tourtip et al. 2009) vs the other 4 methods (0-6%) is clear evidence that there was cross reaction with other microsporidia (Jaroenlak et al. 2016). It is interesting to note the high prevalence of Spiroplasma using primers other than those for S. penaei. If this is correct, it could mean that other species of Spiroplasma are already infecting P. vannamei in the region (Ding et al. 2007, Liang et al. 2010).
3.3. Miscellaneous PCR findings from broodstock specimens
Ovaries of four broodstock were collected for PCR analysis. We found that all of them were positive for WzSV8 and Propionigenium (copy number ranged from 36 to 1,000/sample). The detection of WzSV8 in the gonads by PCR coincides with the detection by histological sections and raises concern whether WsSV8 can affect these two completely different organs, but it also suggests that this virus may be transmitted both horizontally and vertically.
3.4. Histology results by frequency of lesions
It is interesting to note that viruses were the dominant microorganisms with pathogenic potential in all 3 regions. If we exclude any other pathological alteration besides viral inclusion bodies, 80% of all samples analyzed (124 samples) presented at least one type of viral inclusion (VIN), and 18% presented at least 2 types of VIN.
In all regions, WzSV8 was the most common virus observed, with the HP as the main target organ. Seventy-five percent of VIN was due to WzSV8. WzSV8 VIN was found in all tubule epithelial types except F-cells (Pic. 1), but focusing on E-cells facilitates rapid screening because of their location on the outer rim of the HP and because they rarely contain vacuoles. Lightner double inclusions (LDI) (Srisala et al. 2022, 2023) can be considered pathognomonic for WzSV8 because of their unique morphology of an additional (usually smaller) eosinophilic inclusion adjacent to the single, circular basophilic inclusions in vacuoles (Pic. 1). In some specimens rounded up, sloughed WzSV8-infected cells were noted in the lumen of tubules. Interestingly, WzSV8 VIN, including both basophilic and LDI, were also found frequently in the ovaries of some specimens (PCR positive for WzSV8), (Pic. 2) at both the developing and mature stages. In addition, WzSV8 VIN was occasionally found in the anterior midgut caecum (Pic. 2).
DHPV was the second most common virus lesion encountered, with 18% of shrimp examined and having VIN indicative of infection. DHPV VIN are characterized as basophilic VIN that are intact bodies (viral inclusion bodies or VIB) that can sometimes be seen, still intact, after being sloughed into the hepatopancreatic tubule lumen. During their developmental expansion in the nuclei of HP tubule epithelial cells or occasionally the anterior midgut caecum (AMC), the chromatin is marginated, and the intervening nucleolus is pushed aside to sometimes resemble a “bracket” to the expanding VIB. An example is shown here in the AMC (Pic. 3 and 4). DHPV had a higher tropism for the AMC based on the frequency of tissue positives for VIB, namely, 87% of the time in the anterior caecum, 26% in the hepatopancreas and 4% in the posterior caecum. This contrasts with Asia, where DHPV is very rarely reported from P. vannamei (mostly reported from P. monodon and only in its HP). This suggests a possible difference in strains of DHPV in Asia and the Americas.
From all samples, 48% had nematodes, gregarines or both. Larval nematodes (Pic. 5 and 6) were encysted in the wall of the foregut near the junction with the midgut. Gregarine trophozoites were within the lumen of the anterior midgut caecum and midgut intestine, while gametocytes were within the lumen of the posterior midgut caecum.
For the three regions, 26% of all shrimp analyzed had lymphoid organ spheroids, and more than 50% of the animals had pathological changes and/or diagnostic changes consistent with viral infection (WzSV8, HPV/DHPV, BP) in the HP. Three percent of shrimp had LOS but no indication of viral infection or host inflammatory responses in the HP (Pic. 7).
Histopathological alterations of the hepatopancreas were characterized by acute to chronic tubule epithelial necrosis and sloughing with host inflammatory responses around and into the affected tubules. This sometimes included melanization as well as large tubule distention and epithelial compression (atrophy-collapsed) along the basement membrane of the affected tubules. There was scant to absent eosinophilic stain uptake within the sinusoids of the areas of tubules lacking hemocytic cellular infiltration (Pics. 8-10).
The pathologies observed were consistent with acute to chronic toxicity of the hepatopancreas (due to Pir A/B exotoxins) as well as intratubular microbial infections by selected microbial agents (hepatopancreatic vibriosis and intracellular bacteria) (Pics 8-10).
Clusters of microsporidian spores were mainly observed within striated muscle sarcomeres, although examples were also noted within HP tubules and cells of the interstitium (Pic. 11).
3.5. Histology results by region
Region 1 had the highest range of identified pathogens,excluding microsporidia and encysted nematodes. The majority were within the HP/midgut with the exceptions of WSSV (not in the HP) and WzSV8 in the gonads (Pics. 1-10). WzSV8 prevalence ranged from 5% in larvae to 100% in wild animals. WzSV8-infected cells and LO spheroids were the most frequent histological anomalies: the average of all samples was 49% and 43% at low grade (scored at 1.3 to 1.5), respectively.
Although variable between individual shrimp reactive cells, nodules within LOS contained cytoplasmic vacuoles, karyorrhectic to pyknotic nuclei and/or suspected small RNA virus inclusions. The histopathology of wild animals (average weight 43 g) indicated endemic status of WzSV8 in the native population of P. vannamei in the region (Table 12). In farm and broodstock shrimp, the prevalence of WzSV8 was 44% and 47%, respectively. DHPV prevalence in region 1 samples was 12%, 1% and 30% for farm, broodstock and wild shrimp, respectively. Average lesions in the hepatopancreas and intestine, both at grade 1.5, were noted at 12% and 1.0%, respectively.
WSSV was detected only in farms and broodstock at a prevalence of 4.0 and 2.1% but at high grades of 3.7 and 3.3, respectively. BP was found only in a few farms and broodstock animals at a very low % (<2%, not included in the table). Microsporidia (2.5%) were found only in a few farm animals at high grade 3 (Pic. 11). Gregarines plus nematodes were noted in 25% of the farm and broodstock shrimp and 90% of the wild shrimp.
Region 2 had the lowest number of pathogens present. Similar to Region 1, WzSV8 VIN was predominant in 26% of the larvae samples at low grade 1. In farm animals, the prevalence of WzSV8 averaged 45% at grade 1 (Pic. 12). Systemic inflammatory responses were noted in 11% of the specimens (Pic. 13). These changes were hemocytic aggregations in hemocytic nodules, some of which were melanized, in the lymphoid organ, heart, sinusoids of the hepatopancreas, hematopoietic tissue and connective tissue of abdominal muscle and the stomach (Pic. 14 and 15). The prevalence of specimens showing LOS was 6%.
The region 3 virus prevalence was 31% and 4% for WzSV8 and DHPV, respectively (Pic. 16 and 17). The other viruses were not detected by histology in this region. Forty percent of the specimens had gregarines plus encysted nematode larvae associated with the digestive track (Pic. 18). Hepatopancreas necrotic tubules and prominent host inflammatory responses in the surrounding sinusoids were found. An 11% prevalence of abnormal HP was found in this region (Pic. 19).
3.6. Concurrence of histological changes
It is interesting to note that viral presence was widespread in all 3 regions. In Region 1, excluding any other pathological anomaly and only focusing on the presence of viral inclusions in farmed, broodstock or wild shrimp, 77%, 83% and 100% of the samples had at least 1 virus and 25%, 17% and 30% had at least 2 viruses, respectively. In region 2, 73% of the samples analyzed had at least one virus present, and in region 3, 80% of the samples had one virus present, and only 7% had 2.
In all regions, WzSV8 VIN was predominant. While WzSV8 VIN was always found mainly in the hepatopancreas, DHPV was found 87% of the time in the anterior caecum, 26% in the hepatopancreas and 4% in the posterior caecum. This contrasted with the type of DHPV that occurs in P. monodon in Asia, where its inclusions occur only in the HP and where DHPV is very rarely reported in farmed P. vannamei. These results may be the result of varietal strains of DHPV, from which at least 8 full genome sequences are available at GenBank (Srisala et al., 2021).
DHPV was the second most important viral pathogen present in regions 1 and 3. In Region 1, it was present at 12%, 1% and 30% in farms, broodstock and wild animals, respectively. In Region 3, the prevalence was 5%. In Region 1, we also found WSSV and BP in farm animals and broodstock only, with WSSV prevalence at 4.6% in farm animals and 0.5% in broodstock and with BP prevalence at 2.6% in farm animals and 2.5% in broodstock. While BP was found exclusively in the hepatopancreas, WSSV was found in the stomach, antennal gland, connective tissue, gills, epidermis, and gonads.
In Region 1, combined nematodes and gregarines were found in farm (26%), broodstock (28%) and wild animals (90%). In Region 3, 42% were found to have the same parasites. No parasites of this kind were found in region 2. These parasites were found mostly in the anterior and posterior parts of the intestine and the anterior midgut caecum.
3.7. Consistency between PCR and histology results
WzSV8, DHPV and WSSV were the only pathogens that were seen in histopathological samples (unequivocally identified by their viral inclusions) and at the same time detected by PCR. However, it is important to note that PCR and histology results in this report did not arise from the same animals but from random samples of the same population from the same pond. Other pathogens, such as Vibrio spp., AHPND-associated bacteria and RLB, were detected in high prevalence by PCR. Unfortunately, those observations could not be related directly to the histological damage seen in the hepatopancreas or other tissues. We could not confirm the identity of the bacteria in the lesions because we lacked tools such as in situ hybridization. For this reason, we could not identify the pathogens in the tissue sections, and we simply referred to these pathogens as bacteria.