Detection of PrPSc in lymphoid tissues of lambs experimentally exposed to the scrapie agent

Histoblotting and immunohistochemistry were used to detect disease-associated prion protein (PrP(Sc)) in lymphoid tissues of lambs of known PrP genotype infected with the scrapie agent by stomach tube at the age of 2 months. The ileal and jejunal Peyer's patches and retropharyngeal and distal jejunal lymph nodes were studied 1 week, 5 weeks, 5 months and 11 months after inoculation. Other lymphoid tissues examined included superficial cervical lymph node, tonsil and spleen. PrP(Sc) was not detected in any tissue of any lamb at 1 week post-inoculation. At 5 weeks, PrP(Sc) was detected in tissues of lambs of susceptible PrP genotypes (AV(136)QQ(171) and VV(136)QQ(171)), but not lambs of other PrP genotypes (AA(136)QQ(171), AA(136)QR(171) and AV(136)QR(171)). PrP(Sc) was present in the germinal centres of tonsils, distal jejunal and retropharyngeal lymph nodes, and spleen. In the nodules of ileal and jejunal Peyer's patches, only occasional solitary cells showed the presence of PrP(Sc). At 5 months post-inoculation, increased accumulations of PrP(Sc) were detected in ileal and jejunal Peyer's patches, as well as in the retropharyngeal and distal jejunal lymph nodes of a single lamb inoculated with the agent from a sheep of the same susceptible PrP genotype. Eleven months after exposure to the scrapie agent, PrP(Sc) was detected in all lymphoid tissues examined from sheep of susceptible PrP genotypes. These studies show that PrP(Sc) was detectable in lymphoid tissues 5 weeks after exposure to the scrapie agent by stomach tube in lambs as young as 3 months of age and indicate that the PrP genotype is a significant factor for the rapid uptake and spread of the agent through lymphoid tissues.     

Immunohistochemical detection of prion protein in lymphoid tissues of sheep with natural scrapie.

The scrapie-associated form of the prion protein (PrPSc) accumulates in the brain and lymphoid tissues of sheep with scrapie. In order to assess whether detecting PrPSc in lymphoid tissue could be used as a diagnostic test for scrapie, we studied the localization and distribution of PrPSc in various lymphoid tissues collected at necropsy from 55 sheep with clinical scrapie. Samples collected from the spleen, palatine tonsil, ileum, and five different lymph nodes were immunohistochemically stained for PrPSc. PrPSc was found to be deposited in a reticular pattern in the center of both primary and secondary lymphoid follicles. In addition, granules of PrPSc were seen in the cytoplasm in macrophages associated with the lymphoid follicles. In 54 (98%) of the 55 scrapie-affected sheep, PrPSc was detected in the spleen, retropharyngeal lymph node, mesenteric lymph node, and the palatine tonsil. However, only in the palatine tonsils was PrPSc present in a consistently high percentage of the lymphoid follicles. PrP was not detected in any of the lymphoid tissues of 12 sheep that had no neurohistopathological signs of a scrapie infection. We conclude that the tonsils are the best-suited lymphoid tissue to be biopsied for the detection of PrPSc in the diagnosis of clinical scrapie in living sheep.     

Detection of PrP(Sc) in rectal biopsy and necropsy samples from sheep with experimental scrapie

Scrapie diagnosis is based on the demonstration of disease-associated prion protein (PrP(Sc)) in brain or, in the live animal, in readily accessible peripheral lymphoid tissue. Lymphatic tissues present at the rectoanal line were readily obtained from sheep without the need for anaesthesia. The presence of PrP(Sc) in such tissue was investigated in sheep infected orally with scrapie-infected brain material. The methods used consisted of immunohistochemistry and histoblotting on biopsy and post-mortem material. PrP(Sc) was detected in animals with PrP genotypes associated with high susceptibility to scrapie from 10 months after infection, i.e., from about the time of appearance of early clinical signs. In the rectal mucosa, PrP(Sc) was found in lymphoid follicles and in cells scattered in the lamina propria, often near and sometimes in the crypt epithelium. By Western blotting, PrP(Sc) was detected in rectal biopsy samples of sheep with the PrP genotype VRQ/VRQ, after electrophoresis of material equivalent to 8 mg of tissue. This study indicated that rectal biopsy samples should prove useful for the diagnosis of scrapie in sheep.     

Transportation of prion protein across the intestinal mucosa of scrapie-susceptible and scrapie-resistant sheep

To determine the mechanisms of intestinal transport of infection, and early pathogenesis, of sheep scrapie, isolated gut-loops were inoculated to ensure that significant concentrations of scrapie agent would come into direct contact with the relevant ileal structures (epithelial, lymphoreticular, and nervous). Gut loops were inoculated with a scrapie brain pool homogenate or normal brain or sucrose solution. After surgery, animals were necropsied at time points ranging from 15 min to 1 month and at clinical end point. Inoculum-associated prion protein (PrP) was detected by immunohistochemistry in villous lacteals and in sub-mucosal lymphatics from 15 min to 3.5 h post-challenge. It was also detected in association with dendritic-like cells in the draining lymph nodes at up to 24 h post-challenge. Replication of infection, as demonstrated by the accumulation of disease-associated forms of PrP in Peyer's patches, was detected at 30 days and sheep developed clinical signs of scrapie at 18-22 months post-challenge. These results indicate discrepancies between the routes of transportation of PrP from the inoculum and sites of de novo-generated disease-associated PrP subsequent to scrapie agent replication. When samples of homogenized inoculum were incubated with alimentary tract fluids in vitro, only trace amounts of protease-resistant PrP could be detected by western blotting, suggesting that the majority of both normal and abnormal PrP within the inoculum is readily digested by alimentary fluids.     

Short-term study of the uptake of PrP(Sc) by the Peyer's patches in hamsters after oral exposure to scrapie

The disease-associated prion protein (PrP(Sc)) has been detected in the ileal Peyer's patches of lambs as early as one week after oral exposure to scrapie. In hamsters, the earliest reported time of PrP(Sc) detection in the Peyer's patches after oral exposure to scrapie is 69 days post-infection. To evaluate the acute uptake of inoculum and to investigate whether the Peyer's patches constitute the primary site of entry for scrapie after oral exposure, hamsters were each exposed orally to 1 ml of a 10% brain homogenate from hamsters in the terminal stage of infection with the 263 K strain of the scrapie agent. PrP(Sc) was demonstrated in the Peyer's patches only a few days after exposure, i.e., much earlier than previously reported. This study supports the view that the Peyer's patches constitute at least one of the primary entry sites of PrP(Sc) after oral exposure to scrapie.     

Active surveillance for scrapie by third eyelid biopsy and genetic susceptibility testing of flocks of sheep in Wyoming

Control of scrapie, an ovine transmissible spongiform encephalopathy or prion disorder, has been hampered by the lack of conventional antemortem diagnostic tests. Currently, scrapie is diagnosed by postmortem examination of the brain and lymphoid tissues for PrP(Sc), the protein marker for this group of disorders. For live, asymptomatic sheep, diagnosis using tonsil or third-eyelid lymphoid tissue biopsy and PrP(Sc) assay has been described. To evaluate the feasibility and efficacy of third-eyelid testing for identification of infected flocks and individual infected sheep, 690 sheep from 22 flocks were sampled by third-eyelid lymphoid tissue biopsy and immunohistochemistry. Sheep were further evaluated for relative genetic susceptibility and potential contact exposure to scrapie. Third-eyelid testing yielded suitable samples for 80% of the sheep tested, with a mean of 18.1 lymphoid follicles (germinal centers) per histologic section. Three hundred eleven of the sheep were sampled through passive surveillance programs, in which only sheep with potential contact with an infected sheep at a lambing event were tested, regardless of their scrapie susceptibility genotype. In addition, 141 genetically susceptible sheep with no record of contact with an infected animal at a lambing event were sampled through a targeted active surveillance program. Ten PrP(Sc)-positive sheep were identified through the passive surveillance program, and an additional three PrP(Sc)-positive sheep, including two from flocks with no history of scrapie, were identified through the active surveillance program. All PrP(Sc)-positive sheep had the highly susceptible PrP genotype. Third-eyelid testing is a useful adjunct to flock monitoring programs, slaughter surveillance, and mandatory disease reporting in a comprehensive scrapie eradication and research program.     

Protease-resistant prion protein in brain and lymphoid organs of sheep within a naturally scrapie-infected flock

The hallmark of transmissible spongiform encephalopathies (TSE), such as scrapie in sheep, is the accumulation in tissues of an insoluble and protease resistant form (PrPres) of the cellular prion protein. In this study, we evaluated whether the diversity in both the clinical pattern and the PrP genotypes of scrapied sheep from the same flock was connected with different levels and/or glycoform patterns of the PrPres in the brain and lymphoid organs of the animals. Whereas the PrPres levels in spleen, lymph nodes and tonsils from sheep of different PrP genotypes and clinical status appeared comparable, they were highly variable in brain, particularly in the brain stem and the cerebellum. PrPres was only detected in sheep bearing at least one VRQ allele, including three asymptomatic sheep and the highest PrPres load was found in the cerebellum of VRQ/VRQ animals. All together, levels of PrPres in brain did not necessarily correlate with the severity of the clinical disease but might depend on the PrP genotype of the animals. Different brain regions from a given sheep displayed a similar glycopattern of PrPres, whereas the apparent molecular sizes of the unglycosylated and diglycosylated forms of the protein differed between brain and lymphoid tissues. We did not find any notifiable differences in the glycopattern of PrPres in brain from sheep of different PrP genotypes or different clinical status and this PrPres glycotype was also similar to that found in brain from four cattle BSE.     

Cellular immunity in Peyer''s patches of rats infected with Trichinella spiralis.

A rat model of Trichinella spiralis gut infection was used to observe the sequence of developing cellular immunity in Peyer's patches and other lymphoid tissues. Whereas cellular reactivity (lymphocyte blastogenesis) for worm antigens was evident in mesenteric lymph nodes draining the gastrointestinal tract within 3 days after infection, Peyer's patch lymphocytes developed maximal reactivity 2 to 3 weeks later at the same time as the spleen and other lymphoid tissues. Furthermore, the immune reactivity found in Peyer's patches was only transient. Thus, in this parasitic gut infection, the Peyer's patch lymphoid tissue does not appear to be the first site of cellular responsiveness but rather to acquire cellular reactivity only when other lymphoid elements in the infected host have also acquired similar antigen-induced reactivity.     

CD21-positive follicular dendritic cells: A possible source of PrPSc in lymph node macrophages of scrapie-infected sheep

Natural sheep scrapie is a prion disease characterized by the accumulation of PrP(Sc) in brain and lymphoid tissues. Previous studies suggested that lymph node macrophages and follicular dendritic cells (FDC) accumulate PrP(Sc). In this study, lymph nodes were analyzed for the presence of PrP(Sc) and macrophage or FDC markers using dual immunohistochemistry. A monoclonal antibody (mAb) to the C-terminus of PrP reacted with CD172a+ macrophages and CD21+ FDC processes in secondary follicles. However, a PrP N-terminus-specific mAb reacted with CD21+ FDC processes but not CD172a+ macrophages in secondary follicles. Neither the PrP N-terminus nor C-terminus-specific mAb reacted with CD172a+ macrophages in the medulla. These results indicate that lymph node follicular macrophages acquire PrP(Sc) by phagocytosis of CD21+ FDC processes. The results also suggest that follicular macrophages have proteases that process full-length PrP(Sc) to N-terminally truncated PrP(Sc).     

Role of the draining lymph node in scrapie agent transmission from the skin

Transmissible spongiform encephalopathies (TSEs) are neurodegenerative diseases that affect humans and animals. Diseases include scrapie in sheep and Creutzfeldt-Jakob disease in humans. Following peripheral exposure, TSE agents usually accumulate on follicular dendritic cells (FDCs) in lymphoid tissues before neuroinvasion. Studies in mice have shown that TSE exposure through scarified skin is an effective means of transmission. Following inoculation by this route TSE agent accumulation upon FDCs is likewise essential for the subsequent transmission of disease to the brain. However, which lymphoid tissues are crucial for TSE pathogenesis following inoculation via the skin was not known. Mice were therefore created that lacked the draining inguinal lymph node (ILN), but had functional FDCs in remaining lymphoid tissues such as the spleen. These mice were inoculated with the scrapie agent by skin scarification to allow the role of draining ILN in scrapie pathogenesis to be determined. We show that following inoculation with the scrapie agent by skin scarification, disease susceptibility was dramatically reduced in mice lacking the draining ILN. These data demonstrate that following inoculation by skin scarification, scrapie agent accumulation upon FDCs in the draining lymph node is critical for the efficient transmission of disease to the brain.     

Occurrence and distribution of infection-specific PrP in tissues of clinical scrapie cases and cull sheep from scrapie-affected farms in Shetland

The prion protein (PrP) genotypes of all cull sheep originating from four scrapie-affected farms in Shetland in 1998-1999 were determined and a representative sample of the different genotypes was selected for necropsy. Samples of brain and selected viscera were removed from 159 such sheep aged 2-11 years. These samples were examined immunohistochemically and by Western blotting for infection-specific forms of PrP. None of the sheep bearing the following genotypes showed any evidence of PrP accumulation in brain, intestine, selected lymph nodes or the cranial mesenteric ganglia: ARQ/ARQ (n = 41), ARQ/ARH (n = 12), ARH/ARH (n = 2), ARQ/ARR (n = 24), ARR/ARR (n= 2). In five of 71 sheep bearing a single VRQ allele, PrP accumulation was detected immunohistochemically in viscera or brain, or both. These results suggested that only a small proportion of susceptible sheep showed evidence of infection (accumulation of PrP) on the farms studied, and that even sheep of the most susceptible genotype (VRQ/VRQ) did not invariably develop disease in an infected environment. Furthermore, there was no evidence that, in sheep of semi-resistant or fully resistant genotypes, infection could be sequestered within the lymphoreticular system or peripheral nervous system and thereby provide a possible "carrier" source of infection. Rather, the data suggested that some sheep, possibly because they had been exposed to a relatively low infective dose, became infected and accumulated the infective agent over a protracted pre-clinical phase of the disease. Such sheep might be potentially infective for many years. In two VRQ/ARR genotype sheep, PrP was confined to the brain. Infection-specific PrP was also confined to the brain in two of 24 clinical cases of VRQ/ARQ scrapie. Thus, direct neuroinvasion, apparently without a prior phase of replication in the lymphoreticular system, occurred in a proportion of VRQ/ARQ sheep. Possibly it may occur in all sheep of the VRQ/ARR genotype. The factors responsible for direct neuroinvasion are not understood. However, it cannot be attributed to genotype alone.     

Diagnosis and PrP genotype target of scrapie in clinically healthy sheep of Massese breed in the framework of a scrapie eradication programme

Summary. The application of a selective culling programme in two scrapie affected flocks of Massese breed sheep is described. The genetic susceptibility of this breed and the sensitivity of different diagnostic methods in the pre-clinical diagnosis of scrapie were also investigated. Overall, 2,068 clinically healthy sheep underwent PrP genotyping, providing the basis for selective culling. The prevalence of scrapie infection was investigated in susceptible sheep by two independent diagnostic methods. All the sheep older than 18 months (n = 620) were tested by Prionics^® Check Western rapid test on the obex, with a prevalence of infection of 3.9%. Furthermore, 385 sheep underwent immunohistochemistry (IHC) on retropharyngeal lymph node (RPLN), with a prevalence of infection of 5.2%. Overall, 32 sheep were diagnosed with pre-clinical scrapie. Of these, 31 were positive by Western blot on the spleen, 29 by IHC on the RPLN and tonsil, 28 by IHC on the obex, 24 by rapid test, and only 18 by IHC on the third eyelid. All the scrapie positive sheep were of the ARQ/ARQ, ARQ/AHQ or ARQ/VRQ genotypes. No significant differences in scrapie prevalence were observed among these genotypes. The estimated risk of the three targeted alleles was also similar, suggesting that in this breed the VRQ allele was not at higher risk for scrapie, compared to the ARQ and AHQ alleles.     

Vacuolar lesion profile in sheep scrapie: factors influencing its variation and relationship to disease-specific PrP accumulation

Detailed neuropathological examination for vacuolar lesions was performed on the brains of 42 sheep with clinical signs compatible with scrapie. The sheep were grouped according to their breed (Poll-Dorset, Cheviot, Welsh Mountain, Shetland and Suffolk), their PrP genotype at codons 136, 154 and 171 (VRQ/VRQ, VRQ/ARQ, VRQ/ARR and ARQ/ARQ) and the type of infection (experimental infection with SSBP/1, or natural disease). Twenty-two neuroanatomical sites from seven brain regions were examined for vacuolation in the neuropil and five sites at the level of the obex were examined for intraneuronal vacuolation. In 36 sheep, immunohistochemical examination for disease-specific PrP (PrP(d)) accumulation had also been performed in the same brain regions in an earlier study. The magnitude of total neuropil vacuolation was highest in the naturally affected ARQ/ARQ Suffolk sheep and lowest in the experimentally infected VRQ/VRQ Cheviot sheep and VRQ/ARR Poll-Dorset sheep. The severity of neuropil vacuolation at nine of the 22 neuroanatomical sites examined was used to generate a vacuolar lesion profile, which showed variations between the different sheep groups. These variations could be attributed to both PrP genotype and sheep breed and also possibly to scrapie agent; there was, however, considerable individual variation in lesion profile within sheep groups. All groups showed a similar ratio of neuropil vacuolation to neuronal vacuolation at the level of the obex. Although a positive correlation between neuropil vacuolation and PrP(d) deposition was generally observed, it was low except for the astrocyte-associated pattern of PrP(d) accumulation. The study suggests that vacuolar lesion profiles in sheep are affected by several factors and, by comparison with lesion profiles in mice, are of no more than limited value for discriminating between scrapie strains.     

PrP protein is associated with follicular dendritic cells of spleens and lymph nodes in uninfected and scrapie-infected mice.

Abnormal forms of a host protein, PrP, accumulate in the central nervous system in scrapie-affected animals. Here, PrP protein was detected immunocytochemically in tissue sections of spleen, lymph node, Peyer's patches, thymus, and pancreas from uninfected mice and from mice infected with a range of mouse-passaged scrapie strains and bovine spongiform encephalopathy (BSE). In the spleen, lymph node and Peyer's patches, PrP-positive cells were identified as follicular dendritic cells (FDC) by their location, appearance, and immune complex trapping function, whereas in the thymus they appeared to be two types of stromal cells: interdigitating cells (IDC) and cortical epithelial cells. In pancreas, PrP-containing cells were confined to the islets of Langerhans. Although the distribution of PrP immunolabelling was the same in tissues from scrapie-affected and uninfected mice, there was evidence that PrP accumulated in abnormal forms in FDC of infected mice. If, as is likely, PrP is essential for agent replication, our results suggest that FDC are the site of scrapie and BSE replication in the spleen and lymph node.     

Histopathological findings in the combined immunity-deficiency syndrome

The histopathological appearances of the thymus, lymph nodes, spleen, and gut-associated lymphoid tissue (tonsil, Peyer's patches, and appendix) in cases of the combined immunity-deficiency syndrome are presented. The appearance of tissues remaining after foetal thymic transplants and the effects of such transplants on the morphology of lymph nodes are also discussed.Although thymic appearances are remarkably constant the picture in the lymph nodes and spleen may vary considerably. The tonsils, Peyer's patches, and appendix appear to constitute one lymphoid organ in man, at least with regard to the developmental arrest that may occur in this syndrome.     

CD38 is a marker of human lacteals

CD38 is a type II transmembrane glycoprotein involved in signaling and adhesion which is expressed mainly by immature hematopoietic cells and activated lymphoid cells. Central lymphatic channels of human small intestinal villi, the so-called lacteals, were coincidentally found to express CD38. Gastric and large intestinal mucosae, pancreas, liver, lung, nasal mucosa, kidney, thymus, palatine tonsil, Peyer's patches, appendix, and mesenteric lymph nodes, and rodent intestinal mucosa were subsequently examined for lymphatic expression of CD38. Cryosections prepared from biopsy or surgical resection specimens were immunostained with four different antibodies to CD38 combined with antibodies to von Willebrand factor and CD31 to differentiate lymphatics from blood vessels, or with antibody to lysosomal protein. Sections were evaluated by ordinary and confocal immunofluorescence microscopy. Jejunal cryosections were subjected to in situ hybridization for CD38. All CD38 antibodies decorated human lacteals, and some of these were positive for CD38 mRNA. Lymphatics draining Peyer's patches and appendix as well as afferent lymphatics of mesenteric lymph nodes expressed CD38 weakly. CD38 was not detected on lymphatics in other organs or in rodent lacteals. We propose that CD38 is a novel marker of human small intestinal lymphatic vessels.     

A comparative study of the microcirculation in the guinea-pig thymus, lymph nodes and Peyer's patches.

The circulatory patterns in three lymphoid tissues were compared macro- and microscopically. Indian ink or a viscous silicone rubber compound were used as contrast materials and serial 75 micronm sections examined. The findings were checked with the benzidine stain for red blood corpuscles in uninjected tissues. Thymic lobules showed a "through circulation"--capillaries penetrating the cortex from within outwards and draining into surface veins, comparable with the circulation in liver lobules. By contrast, lymph nodes and Peyer's patches showed a looped capillary distribution. The outer thymic cortex and primary follicles in lymph nodes contained the maximum capillary density, correlating with the highest mitotic index of thymocytes and lymphocytes in the respective tissues. The lowest vascular density in the thymus was in the medulla, particularly in Hassall's corpuscles, in secondary follicles of lympho nodes and Peyer's patches, which correlates with the storage capacity of antigens and antibodies at those sites. Postcapillary venules in lymph nodes and Peyer's patches showed a characteristic pattern with Indian ink. These venules were absent from the thymus.     

Molecular analysis of cases of Italian sheep scrapie and comparison with cases of bovine spongiform encephalopathy (BSE) and experimental BSE in sheep

Concerns have been raised about the possibility that the bovine spongiform encephalopathy (BSE) agent could have been transmitted to sheep populations via contaminated feedstuffs. The objective of our study was to investigate the suitability of molecular strain typing methods as a surveillance tool for studying scrapie strain variations and for differentiating PrP(Sc) from sheep scrapie, BSE, and sheep BSE. We studied 38 Italian sheep scrapie cases from 13 outbreaks, along with a British scrapie case, an experimental ovine BSE, and 3 BSE cases, by analyzing the glycoform patterns and the apparent molecular masses of the nonglycosylated forms of semipurified, proteinase-treated PrP(Sc). Both criteria were able to clearly differentiate sheep scrapie from BSE and ovine experimental BSE. PrP(Sc) from BSE and sheep BSE showed a higher glycoform ratio and a lower molecular mass of the nonglycosylated form compared to scrapie PrP(Sc). Scrapie cases displayed homogeneous PrP(Sc) features regardless of breed, flock, and geographic origin. The glycoform patterns observed varied with the antibody used, but either a monoclonal antibody (MAb) (F99/97.6.1) or a polyclonal antibody (P7-7) was able to distinguish scrapie from BSE PrP(Sc). While more extensive surveys are needed to further corroborate these findings, our results suggest that large-scale molecular screening of sheep populations for BSE surveillance may be eventually possible.     

Evidence for co-infection of ovine prion strains in classical scrapie isolates

The diversity of strains of ovine prions within classical scrapie isolates was investigated by transmission studies in wild type mice. To determine the maximum diversity of prion strains present in each ovine scrapie isolate examined, isolates from mice having the shortest and longest incubation times for terminal disease after primary inoculation were passaged serially. Serial passage of ARQ/ARQ scrapie isolates in RIII mice revealed the ME7 prion strain in mice with short incubation times for terminal prion disease and the 87A strain in those mice with long incubation times. Serial passage of VRQ/VRQ scrapie isolates in RIII mice led to emergence of the 221C prion strain in mice with short incubation times and a variant of the 221C strain in those mice with long incubation times. RIII mice with short incubation times had higher levels of total and proteinase K-resistant PrP(Sc) compared with those RIII mice with long incubation times, while mice with long incubation times had large aggregates and plaques of PrP(Sc). ME7 PrP(Sc) differed in stability compared with the 87A prion strain, while PrP(Sc) associated with 221C had similar stability to that of the 221C variant. Serial passage in VM mice led to identification of ME7 and 87V in the same scrapie isolate. The data show that different prion strains can emerge from the same ovine scrapie isolate following serial passage in wild type mice and that the transmission properties of these strains correlate with distinct patterns of PrP(Sc) deposition.     

Human tonsil intraepithelial B cells: a marginal zone-related subpopulation.

AIMS: To determine if intraepithelial B cells in reactive human palatine tonsils were similar to the marginal zone cells of the spleen and Peyer's patches. METHODS: Reactive human palatine tonsils were studied using conventional methods of light microscopy, electron microscopy, and a panel of monoclonal antibodies for leucocyte common antigens. RESULTS: Clinically important numbers of marginal zone-related B cells around the mantle zone were absent in lymphoid follicles, but in the cryptal epithelium there were abundant lymphoid cells with centrocyte-like nuclei and clear cytoplasm, intermingled with macrophages and plasma cells. The immunophenotype of these intraepithelial B cells was distinctive and similar to that found in the splenic marginal zone cells (IgM+, IgD-, CD23-, CD10-, CD35+, CD21+, bc12+, KB61+). CONCLUSIONS: Intraepithelial B cells in human tonsil could represent the counterpart of the marginal zone described in Peyer's patches. Their presence within the epithelium could reflect the destination for the malignant B cells in the lymphoepithelial lesion of mucosa associated lymphoid tissue (MALT) lymphomas. Human palatine tonsil lymphoid tissue has morphological, immunophenotypic, and pathological features similar to those of MALT.